18771 MICROFINANCE HANDBOOK SUSTAINABLE BANKING with the POOR MICROFINANCE HANDBOOK An Institutional and Financial Perspective Joanna Ledgerwood THE WORLD BANK WASHINGTON, D.C. Copyright © 1999 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing December 1998 Second printing July 1999 Third printing July 2000 The findings, interpretations, and conclusions expressed in this paper are entirely those of the author and should not be attributed in any manner to the World Bank, to its affiliated organiza- tions, or to members of its Board of Executive Directors of the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any terri- tory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission prompt- ly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. Joanna Ledgerwood is consultant to the Sustainable Banking with the Poor Project, World Bank, for Micro Finance International, Toronto, Canada. Library of Congress Cataloging-in-Publication Data Ledgerwood, Joanna. Microfinance handbook: an institutional and financial perspective/Joanna Ledgerwood. p. cm.--(Sustainable banking with the poor) Includes bibliographical references and index. ISBN 0-8213-4306-8 1. Microfinance. 2. Financial institutions. I. Title. II. Series. HG178.2.L43 1998 98­21754 332.1--dc21 CIP Contents Foreword xv Preface xvi Introduction 1 PART I--ISSUES TO CONSIDER WHEN PROVIDING MICROFINANCE 9 Chapter 1 Understanding the Country Context 11 Suppliers of Financial Intermediation Services 12 Existing Microfinance Providers 14 What Role Do Donors Play in Microfinance? 16 Financial Sector Policies and Legal Enforcement 17 Interest Rate Policies 18 Government Mandates for Sectoral Credit Allocations 19 Legal Enforcement of Financial Contracts 19 Financial Sector Regulation and Supervision 20 When Should MFIs Be Subject to Regulation? 21 Considerations When Regulating MFIs 23 Country Approaches to Regulating MFIs 25 Economic and Social Policy Environment 26 Economic and Political Stability 26 Poverty Levels 28 Investment in Infrastructure and Human Resource Development 28 Government View of the Microenterprise Sector 29 Appendix 1. Risks in the Microfinance Industry 30 Sources and Further Reading 31 Chapter 2 The Target Market and Impact Analysis 33 Objectives of the Microfinance Institution 33 Direct and Indirect Targeting 34 The Importance of Adequate Cash Flow and the Capacity to Service Debt 35 v vi CONTENTS Minimal Equity Requirement 36 Moral Hazard 36 Market Size 36 Identifying the Target Market 37 Characteristics of the Population Group 37 Types of Microenterprises 42 Impact Analysis 46 Kinds of Impacts 47 What Kinds of Impacts Have We Seen with Microfinance? 48 Impact Proxies 49 Client-Oriented Impact Analysis 49 When Should Impact Be Assessed? 52 Methods of Impact Assessment 53 Fundamental Characteristics of Qualitative Approaches 53 Fundamental Characteristics of Quantitative Approaches 54 Comparisons of Quantitative and Qualitative Approaches 56 Integrating Methodologies 56 The Choice of Unit of Analysis 56 Appendix 1. Quantitative Impact Assessment 58 Sources and Further Reading 59 Chapter 3 Products and Services 63 The Systems Framework 64 Microfinance Institutions--Minimalist or Integrated? 65 Financial Intermediation 66 Credit 66 Savings 71 Insurance 74 Credit Cards and Smart Cards 74 Payment Services 75 Social Intermediation 76 Enterprise Development Services 78 Social Services 81 Appendix 1. Microfinance Approaches 82 Appendix 2. Matching Enterprise Development Services to Demand 86 Sources and Further Reading 90 Chapter 4 The Institution 93 The Importance of Institutions 93 Attributes of a Good Institution 94 The Importance of Partner Institutions 94 Institutional Types 97 Formal Financial Institutions 97 Semiformal Financial Institutions 101 Informal Financial Providers 104 CONTENTS vii Institutional Growth and Transformation 106 Expansion Within an Existing Structure 106 Creating an Apex Institution 106 Creating a Formal Financial Intermediary 109 Governance and Ownership 110 Accessing Capital Markets 113 Institutional Capacity Building 117 Appendix 1. MFI Operational Review 118 Appendix 2. Manual for Elaboration of a Business Plan 123 Sources and Further Reading 128 PART II--DESIGNING AND MONITORING FINANCIAL PRODUCTS AND SERVICES 131 Chapter 5 Designing Lending Products 133 Cash Patterns, Loan Terms, and Payment Frequency 133 Client Cash Patterns and Loan Amounts 133 How Does the Loan Term Affect the Borrower's Ability to Repay? 134 Frequency of Loan Payments 136 Working Capital and Fixed Asset Loans 136 Loan Collateral 137 Collateral Substitutes 137 Alternative Forms of Collateral 138 Loan Pricing 138 Calculating Interest Rates 140 How Do Fees or Service Charges Affect the Borrower and the MFI? 142 Cross-Subsidization of Loans 143 Calculating Effective Rates 143 Estimating the Effective Rate 144 Calculating the Effective Interest Rate with Compulsory Savings or Other Loan Variables 146 Calculating the Effective Interest Rate with Varying Cash Flows 147 How Does the Effective Cost for the Borrower Differ from the Effective Yield to the Lender? 148 Appendix 1. How Can an MFI Set a Sustainable Rate on Its Loans? 149 Appendix 2. Calculating an Effective Interest Rate Using the Internal Rate of Return Method 150 Appendix 3. Calculating the Effective Rate with Varying Cash Flows 152 Sources and Further Reading 153 Chapter 6 Designing Savings Products 155 Demand for Savings Services 156 Is There an Enabling Environment? 157 Legal Requirements for Offering Voluntary Savings Services 157 Deposit Insurance 158 Does the MFI Have the Necessary Institutional Capacity to Mobilize Savings? 160 Ownership and Governance 160 Organizational Structure 160 Human Resources 161 Marketing 162 Infrastructure 163 viii CONTENTS Security and Internal Controls 163 Management Information Systems 163 Risk Management and Treasury 163 Sequencing the Introduction of Savings Services 164 Types of Savings Products for Microentrepreneurs 164 Liquid Accounts 165 Semiliquid Accounts 165 Fixed-Term Deposits 166 Costs of Mobilizing Voluntary Savings 166 Pricing Savings Products 167 Sources and Further Reading 168 Chapter 7 Management Information Systems 169 An Overview of Issues Related to Management Information Systems 170 Three Areas of Management Information Systems 171 Accounting Systems 171 Credit and Savings Monitoring Systems 172 Client Impact Tracking Systems 178 Installing a Management Information System 178 Institutional Assessment 178 Configuration 178 Software Modifications 179 Testing 179 Data Transfer 179 Training 180 Parallel Operations 180 Ongoing Support and Maintenance 180 Appendix 1. Overview of Commercial Management Information System Software Packages 180 Appendix 2. Criteria for Evaluating Loan Tracking Software 183 Sources and Further Reading 183 PART III--MEASURING PERFORMANCE AND MANAGING VIABILITY 185 Chapter 8 Adjusting Financial Statements 187 Accounting Adjustments 188 Accounting for Loan Losses 188 Accounting for Depreciation of Fixed Assets 192 Accounting for Accrued Interest and Accrued Interest Expense 193 Adjusting for Subsidies and Inflation 194 Accounting for Subsidies 195 Accounting for Inflation 197 Restating Financial Statements in Constant Currency Terms 199 Appendix 1. Sample Financial Statements Adjusted for Subsidies 200 Appendix 2. Sample Financial Statements Adjusted for Inflation 202 Sources and Further Reading 204 CONTENTS ix Chapter 9 Performance Indicators 205 Portfolio Quality 206 Repayment Rates 206 Portfolio Quality Ratios 207 Loan Loss Ratios 211 Productivity and Efficiency Ratios 212 Productivity Ratios 212 Efficiency Ratios 213 Financial Viability 215 Financial Spread 216 Two Levels of Self-Sufficiency 216 Subsidy Dependence Index 218 Profitability Ratios 220 Return on Assets Ratio 221 Return on Business Ratio 222 Return on Equity Ratio 223 Leverage and Capital Adequacy 223 Leverage 224 Capital Adequacy Standards 224 Scale and Depth of Outreach Indicators 225 Performance Standards and Variations 227 Appendix 1. Sample Balance Sheet 233 Appendix 2. Sample Income Statement 234 Appendix 3. Sample Portfolio Report 235 Appendix 4. Adjusted Sample Financial Statements (Combined) 236 Appendix 5. Analyzing an MFI's Return on Assets 238 Sources and Further Reading 241 Chapter 10 Performance Management 243 Delinquency Management 243 The Effect of Delinquency on an MFI's Profitability 244 Controlling Delinquency 245 Rescheduling or Refinancing Loans 246 Productivity and Efficiency Management 248 Improving Staff Productivity 248 Managing Costs 251 Risk Management 254 Asset and Liability Management 254 Operating Risk Management 258 Appendix 1. Gap Analysis 260 Sources and Further Reading 261 GLOSSARY 263 INDEX 271 x CONTENTS Boxes 1.1 Formal Sector Suppliers in Rural Mexico 14 1.2 Do Microfinance Clients Need Subsidized Interest Rates? 15 1.3 Credit Institutions as a Political Tool: Debt Foregiveness in India 15 1.4 Microfinance in Indonesia 16 1.5 Multilateral Development Banks' Strategies for Microfinance 18 1.6 The Consultative Group to Assist the Poorest 19 1.7 Usury Laws in West Africa 19 1.8 Alexandria Business Association: Legal Sanctions 20 1.9 Regulating MFIs: The Case of Finansol 22 1.10 Enhancing the Effectiveness of Oversight 23 1.11 Private Financial Funds in Bolivia 25 1.12 Nonbank Financial Institutions in Ghana 26 1.13 Microfinance in Areas of Political Unrest 27 1.14 The Albanian Development Fund 28 1.15 Tax Laws in Argentina 29 2.1 Targeted Credit Activities 34 2.2 U.S. Agency for International Development Findings on Female Borrowers 38 2.3 The Kenya Rural Enterprise Programme 40 2.4 The Influence of Ethnicity and Language in Microfinance 41 2.5 Islamic Banking 42 2.6 The Association for the Development of Microenterprises' Work with Existing Microenterprises 43 2.7 Foundation for Enterprise Finance and Development's Approach to Sector Development 46 2.8 The Impact of Finance on the Rural Economy of India 47 2.9 PRODEM's Impact and Market Analysis Project 50 3.1 Principles of Financially Viable Lending to Poor Entrepreneurs 67 3.2 Individual Loans at Fédération des Caisses d'Epargne et de Crédit Agricole Mutuel, Benin 69 3.3 The Association for the Development of Microenterprises 69 3.4 Rotating Savings and Credit Associations 70 3.5 Repayment Instability in Burkina Faso 71 3.6 The Role of Groups in Financial Intermediation 72 3.7 Caisses Villageoises, Pays Dogon, Mali 73 3.8 Self-Employed Women's Association Insurance 75 3.9 The Association for the Development of Microenterprises' MasterCard 75 3.10 Swazi Business Growth Trust Smart Cards 76 3.11 Demand for Payment Services at the Fédération des Caisses d'Epargne et de Crédit Agricole Mutuel, Benin 76 3.12 Village Banks: An Example of a Parallel System 77 3.13 Social Intermediation in a Social Fund in Benin 78 3.14 Business Skills Training 79 3.15 Direct and Indirect Business Advisory Services 79 3.16 Cattle Dealers in Mali 79 3.17 Who Offers Social Services? 81 3.18 Freedom From Hunger--Providing Nutrition Training with Microfinance 81 A3.1.1 Training for Replicators in West Africa 83 A3.2.1 Policy Dialogue in El Salvador 90 CONTENTS xi 4.1 Freedom From Hunger: Partnering with Local Institutions 96 4.2 Types of Financial Institutions 97 4.3 Development Banks 98 4.4 Tulay sa Pag-Unlad's Transformation into a Private Development Bank 99 4.5 The Savings Bank of Madagascar 99 4.6 Caja Social: A Colombian Commercial Bank Reaching the Poor 100 4.7 Caja de Ahorro y Prestamo Los Andes 101 4.8 Accion Comunitaria del Peru 101 4.9 The Rehabilitation of a Credit Union in Benin 102 4.10 CARE Guatemala: The Women's Village Banking Program 104 4.11 The Use of Self-Help Groups in Nepal 105 4.12 Using the Nongovernmental Organization as a Strategic Step in Expansion 107 4.13 Catholic Relief Services: Using the Apex Model for Expansion 108 4.14 Transformation from a Nongovernmental Organization to Financiera Calpía 110 4.15 Catholic Relief Services' Guatemala Development Bank 110 4.16 BancoADEMI Ownership Structure 112 4.17 Guarantee Scheme in Sri Lanka 114 4.18 Key Measures for Accessing Commercial Financing 114 4.19 Accessing Capital Markets by Issuing Financial Paper 115 4.20 ProFund--an Equity Investment Fund for Latin America 115 4.21 The Calvert Group--a Screened Mutual Fund 116 4.22 DEVCAP--a Shared-Return Fund 116 4.23 Credit Unions Retooled 118 A4.2.1 Suggested Business Plan Format 125 A4.2.2 Estimating the Market: An Example from Ecuador 126 5.1 The Association for the Development of Microenterprises' Collateral Requirements 139 5.2 Cross-Subsidization of Loans 143 6.1 Deposit Collectors in India 156 6.2 Savings Mobilization at Bank Rakyat Indonesia 158 6.3 Deposit Insurance in India 159 6.4 Security of Deposits in the Bank for Agriculture and Agricultural Cooperatives, Thailand 159 6.5 Ownership and Governance at the Bank for Agriculture and Agricultural Cooperatives, Thailand 161 6.6 Using Local Human Resources in Caisses Villageoises d'Epargne et de Crédit Autogérées, Mali 162 6.7 The Sequencing of Voluntary Savings Mobilization 164 6.8 Bank Rakyat Indonesia Savings Instruments 165 6.9 The Choice of Savings Products in Pays Dogon, Mali 166 6.10 Administrative Costs for Banco Caja Social, Colombia 167 7.1 Determining the Information Needs of an MFI 170 7.2 Microfinance Institutions with Developed "In-House" Systems 174 7.3 Improving Reporting Formats: Experience of the Workers Bank of Jamaica 176 7.4 Management Information Systems at the Association for the Development of Microenterprises, Dominican Republic 177 7.5 Framework for a Management Information System at Freedom From Hunger 178 7.6 Running Parallel Operations in the Fédération des Caisses d'Epargne et de Crédit Agricole Mutuel, Benin 180 8.1 The Impact of Failure to Write Off Bad Debt 191 8.2 Cash and Accrual Accounting for Microfinance Institutions 194 xii CONTENTS 9.1 Repayment Rates Compared with Portfolio Quality Ratios 208 9.2 The Effect of Write-off Policies 211 9.3 Computation of the Subsidy Dependence Index 220 9.4 CARE Guatemala's Subsidy Dependence Index 221 9.5 Outreach Indicators 226 9.6 Depth of Outreach Diamonds 228 9.7 CAMEL System, ACCION 229 9.8 Financial Ratio Analysis for Microfinance Institutions, Small Enterprise and Promotion Network 230 9.9 PEARLS System, World Council of Credit Unions 231 9.10 Tracking Performance through Indicators, Consultative Group to Assist the Poorest 232 A9.5.1 Analyzing a Financial Institution's Return on Assets 238 10.1 Fifteen Steps to Take in a Delinquency Crisis 246 10.2 Unusual Gains in Efficiency at Women's World Banking, Cali, Colombia 248 10.3 Performance Incentives at the Association for the Development of Microenterprises, Dominican Republic 250 10.4 Performance Incentive Schemes at Tulay Sa Pag-Unlad Inc., the Philippines 251 10.5 Credit Officer Reports at the Association for the Development of Microenterprises 252 10.6 Transfer Pricing at Bank Rakyat Indonesia and Grameen Bank in Bangladesh 253 10.7 Standardization at Grameen Bank 254 10.8 Institution Vulnerability to Fraud 259 10.9 Fraud Control at Mennonite Economic Development Association 260 Figures 1 Relationship between Level of Analysis and Technical Complexity in this Book 5 1.1 Understanding the Country Context 11 2.1 Client Characteristics 38 2.2 Types of Microenterprises 42 3.1 Minimalist and Integrated Approaches to Microfinance 65 3.2 Group Social Intermediation 78 A3.2.1 The Entrepreneur's Context 86 Tables 1.1 Providers of Financial Intermediation Services 13 1.2 Private Institutions in Microenterprise Development 17 2.1 Enterprise Sector Credit Characteristics 45 4.1 Key Characteristics of a Strong Microfinance Institution 95 4.2 What Is at Stake for Microfinance Owners? 112 A4.2.1 Financial Cost as a Weighted Average 128 5.1 Examples of Loan Uses 137 5.2 Declining Balance Method 140 5.3 Flat Method 141 5.4 Effective Rate Estimate, Declining Balance 145 5.5 Effective Rate Estimate, Flat Method 145 5.6 Change in Loan Fee and Loan Term Effect 146 5.7 Variables Summary 147 A5.3.1 Internal Rate of Return with Varying Cash Flows (Grace Period) 152 A5.3.2 Internal Rate of Return with Varying Cash Flows (Lump Sum) 152 8.1 Sample Portfolio Report 188 CONTENTS xiii 8.2 Sample Loan Loss Reserve Calculation, December 31, 1995 190 9.1 Sample Portfolio Report with Aging of Arrears 209 9.2 Sample Portfolio Report with Aging of Portfolio at Risk 209 9.3 Calculating Portfolio at Risk 210 9.4 Calculating Portfolio Quality Ratios 212 9.5 Calculating Productivity Ratios 213 9.6 Calculating Efficiency Ratios 215 9.7 Calculating Viability Ratios 219 9.8 Effect of Leverage on Return on Equity 224 A9.5.1 Breakdown of a Microfinance Institution's Profit Margin and Asset Utilization 239 A9.5.2 Analysis of ADEMI's Return on Assets 240 10.1 Cost of Delinquency, Example 1 244 10.2 Cost of Delinquency, Example 2 245 10.3 On-Time and Late or No Payments 247 10.4 Branch Cash Flow Forecasts 256 Foreword T he Sustainable Banking with the Poor Project sive discussions with practitioners, policymakers, and (SBP) began several years ago as a rather unusual donors. SBP also co-produced with the Consultative joint effort by a regional technical department, Group to Assist the Poorest the "Microfinance Practical Asia Technical, and a central department, Agriculture and Guide," today a regular source of practical reference for Natural Resources. Through the years and the institution- World Bank Group task team leaders and staff working al changes, the project has maintained the support of the on microfinance operations or project components that Asia Region and the Rural Development Department in involve the provision of financial services to the poor. the Environmentally and Socially Sustainable This volume, Microfinance Handbook: An Development vice presidency and the endorsement of the Institutional and Financial Perspective, represents in a Finance, Private Sector and Infrastructure Network. way a culmination of SBP work in pursuit of its princi- An applied research and dissemination project, SBP pal aim. A comprehensive source for donors, policymak- aims at improving the ability of donors, governments, ers, and practitioners, the handbook first covers the and practitioners to design and implement policies and policy, legal, and regulatory issues relevant to micro- programs to build sustainable financial institutions that finance development and subsequently treats rigorously effectively reach the poor. While its main audience has and in depth the key elements in the process of building been World Bank Group staff, the project has con- sustainable financial institutions with effective outreach tributed significantly to the knowledge base in the to the poor. microfinance and rural finance fields at large. The handbook focuses on the institutional and finan- More than 20 case studies of microfinance institutions cial aspects of microfinance. Although impact analysis is have been published and disseminated worldwide, many reviewed briefly, a thorough discussion of poverty target- of them in two languages; a seminar series at the World ing and of the poverty alleviation effects of microfinance is Bank has held more than 30 sessions on microfinance and left to other studies and publications. rural finance good practice and new developments; three We are confident that this handbook will contribute regional conferences have disseminated SBP products in significantly to the improvement of policies and prac- Asia and Africa, while fostering and benefiting from exten- tices in the microfinance field. Mieko Nishimizu Ian Johnson Vice President Vice President South Asia Region Environmentally and Socially Sustainable Development Jean-Michel Severino Masood Ahmed Vice President Acting Vice President East Asia and Pacific Region Finance, Private Sector and Infrastructure xv Preface T he Microfinance Handbook, one of the major Group to Assist the Poorest. I am grateful to Mike products of the World Bank's Sustainable Goldberg and Gregory Chen for their initial contribu- Banking with the Poor Project, gathers and pre- tions. Thomas Dichter wrote the sections on impact sents up-to-date knowledge directly or indirectly con- analysis and enterprise development services and provid- tributed by leading experts in the field of microfinance. ed comments on other chapters. Tony Sheldon wrote It is intended as a comprehensive source for donors, poli- the chapter on management information systems and cymakers, and practitioners, as it covers in depth matters provided comments on the first draft. Reinhardt pertaining to the regulatory and policy framework and Schmidt contributed substantially to the chapter on the essential components of institutional capacity build- institutions. ing--product design, performance measuring and moni- I wish to thank especially Cecile Fruman for her ini- toring, and management of microfinance institutions. tial and ongoing contributions, comments, and support. The handbook was developed with contributions I also appreciate the support and flexibility of Carlos E. from other parts of the World Bank, including the Cuevas, which were vital in keeping the process going. Consultative Group to Assist the Poorest. It also benefit- Thanks are extended to Julia Paxton and Stephanie ed from the experience of a wide range of practitioners Charitonenko Church for their contributions as well as and donors from such organizations as ACCION to the many people who reviewed various chapters and International, Calmeadow, CARE, Women's World provided comments, including Jacob Yaron, Lynn Banking, the Small Enterprise Education and Promotion Bennett, Mohini Malhotra, McDonald Benjamin, Network, the MicroFinance Network, the U.S. Agency Jeffrey Poyo, Jennifer Harold, Bikki Randhawa, Joyita for International Development, Deutsche Gesellschaft Mukherjee, Jennifer Isern, and Joakim Vincze. für Technische Zusammenarbeit, Caisse Française de I would also like to thank Laura Gomez for her con- Développement, and the Inter-American Development tinued support in formatting the handbook and manag- Bank. ing its many iterations. Special thanks are due to the outside sponsors of the The book was edited, designed, and typeset by Com- Sustainable Banking with the Poor Project--the Swiss munications Development Inc. Agency for Development and Cooperation, the Royal Ministry of Foreign Affairs of Norway, and the Ford ***** Foundation--for their patience, encouragement, and support. The Sustainable Banking with the Poor Project team is I would like to thank the many individuals who con- led by Lynn Bennett and Jacob Yaron, task managers. tributed substantially to this handbook. A first draft was Carlos E. Cuevas is the technical manager and Cecile prepared by staff members at both the Sustainable Fruman is the associate manager. Laura Gomez is the Banking with the Poor Project and the Consultative administrative assistant. xvi Introduction I t has been estimated that there are 500 million econom- Financial services generally include savings and credit; ically active poor people in the world operating however, some microfinance organizations also provide microenterprises and small businesses (Women's World insurance and payment services. In addition to financial Banking 1995). Most of them do not have access to ade- intermediation, many MFIs provide social intermediation quate financial services. To meet this substantial demand for services such as group formation, development of self- financial services by low-income microentrepreneurs, confidence, and training in financial literacy and manage- microfinance practitioners and donors alike must adopt a ment capabilities among members of a group. Thus the long-term perspective. The purpose of this handbook is to definition of microfinance often includes both financial bring together in a single source guiding principles and tools intermediation and social intermediation. Microfinance is that will promote sustainable microfinance and create viable not simply banking, it is a development tool. institutions. Microfinance activities usually involve: The goal of this book is to provide a comprehensive I Small loans, typically for working capital source for the design, implementation, evaluation, and I Informal appraisal of borrowers and investments management of microfinance activities.1 I Collateral substitutes, such as group guarantees or The handbook takes a global perspective, drawing on compulsory savings lessons learned from the experiences of microfinance I Access to repeat and larger loans, based on repayment practitioners, donors, and others throughout the world. performance It offers readers relevant information that will help them I Streamlined loan disbursement and monitoring to make informed and effective decisions suited to their I Secure savings products. specific environment and objectives. Although some MFIs provide enterprise development services, such as skills training and marketing, and social services, such as literacy training and health care, these Microfinance Defined are not generally included in the definition of microfi- nance. (However, enterprise development services and Microfinance has evolved as an economic development social services are discussed briefly in chapter 3, as some approach intended to benefit low-income women and MFIs provide these services.) men. The term refers to the provision of financial services MFIs can be nongovernmental organizations to low-income clients, including the self-employed. (NGOs), savings and loan cooperatives, credit unions, 1. The term "microfinance activity" is used throughout to describe the operations of a microfinance institution, a microfinance pro- ject, or a microfinance component of a project. When referring to an organization providing microfinance services, whether regulat- ed or unregulated, the term "microfinance institution" (MFI) is used. 1 2 MICROFINANCE HANDBOOK government banks, commercial banks, or nonbank At the same time, local NGOs began to look for a financial institutions. Microfinance clients are typically more long-term approach than the unsustainable income- self-employed, low-income entrepreneurs in both urban generation approaches to community development. In and rural areas. Clients are often traders, street vendors, Asia Dr. Mohammed Yunus of Bangladesh led the way small farmers, service providers (hairdressers, rickshaw with a pilot group lending scheme for landless people. drivers), and artisans and small producers, such as black- This later became the Grameen Bank, which now serves smiths and seamstresses. Usually their activities provide more than 2.4 million clients (94 percent of them a stable source of income (often from more than one women) and is a model for many countries. In Latin activity). Although they are poor, they are generally not America ACCION International supported the develop- considered to be the "poorest of the poor." ment of solidarity group lending to urban vendors, and Moneylenders, pawnbrokers, and rotating savings and Fundación Carvajal developed a successful credit and credit associations are informal microfinance providers training system for individual microentrepreneurs. and important sources of financial intermediation but Changes were also occurring in the formal financial they are not discussed in detail in this handbook. Rather, sector. Bank Rakyat Indonesia, a state-owned, rural the focus is on more formal MFIs. bank, moved away from providing subsidized credit and took an institutional approach that operated on market principles. In particular, Bank Rakyat Indonesia devel- Background oped a transparent set of incentives for its borrowers Microfinance arose in the 1980s as a response to (small farmers) and staff, rewarding on-time loan repay- doubts and research findings about state delivery of ment and relying on voluntary savings mobilization as a subsidized credit to poor farmers. In the 1970s govern- source of funds. ment agencies were the predominant method of pro- Since the 1980s the field of microfinance has grown viding productive credit to those with no previous substantially. Donors actively support and encourage access to credit facilities--people who had been forced microfinance activities, focusing on MFIs that are com- to pay usurious interest rates or were subject to rent- mitted to achieving substantial outreach and financial seeking behavior. Governments and international sustainability. Today the focus is on providing financial donors assumed that the poor required cheap credit services only, whereas the 1970s and much of the 1980s and saw this as a way of promoting agricultural pro- were characterized by an integrated package of credit and duction by small landholders. In addition to providing training--which required subsidies. Most recently, subsidized agricultural credit, donors set up credit microfinance NGOs (including PRODEM/BancoSol in unions inspired by the Raiffeisen model developed in Bolivia, K-REP in Kenya, and ADEMI/BancoADEMI Germany in 1864. The focus of these cooperative in the Dominican Republic) have begun transforming financial institutions was mostly on savings mobiliza- into formal financial institutions that recognize the need tion in rural areas in an attempt to "teach poor farmers to provide savings services to their clients and to access how to save." market funding sources, rather than rely on donor funds. Beginning in the mid-1980s, the subsidized, targeted This recognition of the need to achieve financial sustain- credit model supported by many donors was the object ability has led to the current "financial systems" of steady criticism, because most programs accumulated approach to microfinance. This approach is characterized large loan losses and required frequent recapitalization to by the following beliefs: continue operating. It became more and more evident I Subsidized credit undermines development. that market-based solutions were required. This led to a I Poor people can pay interest rates high enough to new approach that considered microfinance as an inte- cover transaction costs and the consequences of the gral part of the overall financial system. Emphasis shifted imperfect information markets in which lenders from the rapid disbursement of subsidized loans to target operate. populations toward the building up of local, sustainable I The goal of sustainability (cost recovery and eventu- institutions to serve the poor. ally profit) is the key not only to institutional perma- INTRODUCTION 3 nence in lending, but also to making the lending I Sources of funds to finance loan portfolios differed by institution more focused and efficient. type of institution. NGOs relied heavily on donor I Because loan sizes to poor people are small, MFIs funding or concessional funds for the majority of their must achieve sufficient scale if they are to become lending. Banks, savings banks, and credit unions fund- sustainable. ed their loan portfolios with client and member I Measurable enterprise growth, as well as impacts on deposits and commercial loans. poverty, cannot be demonstrated easily or accurately; I NGOs offered the smallest loan sizes and relatively outreach and repayment rates can be proxies for impact. more social services than banks, savings banks, or One of the main assumptions in the above view is credit unions. that many poor people actively want productive credit I Credit unions and banks are leaders in serving large and that they can absorb and use it. But as the field of numbers of clients with small deposit accounts. microfinance has evolved, research has increasingly The study also found that basic accounting capacities found that in many situations poor people want secure and reporting varied widely among institutions, in many savings facilities and consumption loans just as much as cases revealing an inability to report plausible cost and productive credit and in some cases instead of produc- arrears data. This shortcoming, notably among NGOs, tive credit. MFIs are beginning to respond to these highlights the need to place greater emphasis on financial demands by providing voluntary savings services and monitoring and reporting using standardized practices (a other types of loans. primary purpose of this handbook). Overall, the findings suggest that favorable macroeconomic conditions, man- aged growth, deposit mobilization, and cost control, in Size of the Microfinance Industry combination, are among the key factors that contribute to the success and sustainability of many microfinance During 1995 and 1996 the Sustainable Banking with institutions. the Poor Project compiled a worldwide inventory of MFIs. The list included nearly 1,000 institutions that provided microfinance services, reached at least 1,000 Why is Microfinance Growing? clients, and had operated for a minimum of three years. From this inventory, more than 200 institutions Microfinance is growing for several reasons: responded to a two-page questionnaire covering basic 1. The promise of reaching the poor. Microfinance activi- institutional characteristics. ties can support income generation for enterprises According to the survey results, by September 1995 operated by low-income households. about US$7 billion in outstanding loans had been pro- 2. The promise of financial sustainability. Microfinance vided to more than 13 million individuals and groups. activities can help to build financially self-sufficient, In addition, more than US$19 billion had been mobi- subsidy-free, often locally managed institutions. lized in 45 million active deposit accounts. 3. The potential to build on traditional systems. Micro- The general conclusions of the inventory were: finance activities sometimes mimic traditional systems I Commercial and savings banks were responsible for (such as rotating savings and credit associations). They the largest share of the outstanding loan balance and provide the same services in similar ways, but with deposit balance. greater flexibility, at a more affordable price to I Credit unions represented 11 percent of the total microenterprises and on a more sustainable basis. This number of loans in the sample and 13 percent of the can make microfinance services very attractive to a outstanding loan balance. large number of low-income clients. I NGOs made up more than half of the sample, but 4. The contribution of microfinance to strengthening and they accounted for only 9 percent of the total num- expanding existing formal financial systems. Micro- ber of outstanding loans and 4 percent of the out- finance activities can strengthen existing formal standing loan balance. financial institutions, such as savings and loan coop- 4 MICROFINANCE HANDBOOK eratives, credit union networks, commercial banks, I Others develop neither the financial management sys- and even state-run financial institutions, by expand- tems nor the skills required to run a successful operation. ing their markets for both savings and credit--and, I Replication of successful models has at times proved potentially, their profitability. difficult, due to differences in social contexts and lack 5. The growing number of success stories. There is an of local adaptation. increasing number of well-documented, innovative Ultimately, most of the dilemmas and problems success stories in settings as diverse as rural encountered in microfinance have to do with how clear the Bangladesh, urban Bolivia, and rural Mali. This is in organization is about its principal goals. Does an MFI pro- stark contrast to the records of state-run specialized vide microfinance to lighten the heavy burdens of poverty? financial institutions, which have received large Or to encourage economic growth? Or to help poor amounts of funding over the past few decades but women develop confidence and become empowered with- have failed in terms of both financial sustainability in their families? And so on. In a sense, goals are a matter and outreach to the poor. of choice; and in development, an organization can choose 6. The availability of better financial products as a result one or many goals--provided its constituents, governance of experimentation and innovation. The innovations structure, and funding are all in line with those goals. that have shown the most promise are solving the problem of lack of collateral by using group-based and character-based approaches; solving problems of About This Book repayment discipline through high frequency of repayment collection, the use of social and peer This handbook was written for microfinance practition- pressure, and the promise of higher repeat loans; ers, donors, and the wider readership of academics, con- solving problems of transaction costs by moving sultants, and others interested in microfinance design, some of these costs down to the group level and by implementation, evaluation, and management. It offers a increasing outreach; designing staff incentives to one-stop guide that covers most topics in enough detail achieve greater outreach and high loan repayment; that most readers will not need to refer to other sources. and providing savings services that meet the needs of At first glance it might seem that practitioners and small savers. donors have very different needs and objectives and thus could not possibly benefit equally from one book. The objectives of many donors who support microfi- What Are the Risks of Microfinance? nance activities are to reduce poverty and empower spe- cific segments of the population (for example, women, Sound microfinance activities based on best practices play indigenous peoples). Their primary concerns traditional- a decisive role in providing the poor with access to finan- ly have been the amount of funds they are able to dis- cial services through sustainable institutions. However, burse and the timely receipt of requested (and, it is there have been many more failures than successes: hoped, successful) performance indicators. Donors are I Some MFIs target a segment of the population that rarely concerned with understanding the details of has no access to business opportunities because of lack microfinance. Rather, it is often enough for them to of markets, inputs, and demand. Productive credit is believe that microfinance simply works. of no use to such people without other inputs. Practitioners need to know how actually to operate a I Many MFIs never reach either the minimal scale or microfinance institution. Their objectives are to meet the efficiency necessary to cover costs. the needs of their clients and to continue to operate in I Many MFIs face nonsupportive policy frameworks and the long term. daunting physical, social, and economic challenges. Given the purpose of this handbook, it seems to be I Some MFIs fail to manage their funds adequately directed more toward practitioners than toward donors. enough to meet future cash needs and, as a result, However, donors are beginning to realize that MFI they confront a liquidity problem. capacity is a more binding constraint than the availability INTRODUCTION 5 of funds, making it essential for donors and practitioners technical. Part II, "Designing and Monitoring Financial to operate from the same perspective if they are to meet Products and Services," narrows its focus to the provi- effectively the substantial need for financial services. In sion of financial intermediation, taking a more technical fact, if we look briefly at the evolution of microfinance, it approach and moving progressively toward more specific is apparent that both donors and practitioners need to (or micro) issues. Part III, "Measuring Performance and understand how microfinance institutions operate. Managing Viability," is the most technical part of the When microfinance first emerged as a development tool, handbook, focusing primarily on assessing the financial both donors and practitioners focused on the cumulative viability of MFIs. (These relationships are shown in fig- amount of loans disbursed, with no concern for how well the ure 1.) loans suited borrower needs and little concern about whether Part I addresses the broader considerations of microfi- or not loans were repaid. Donors were rewarded for disburs- nance activities, including the supply of and demand for ing funds, and practitioners were rewarded for on-lending financial services, the products and services that an MFI those funds to as many people (preferably women) as possi- might offer, and the institutions and institutional issues ble. Neither was particularly accountable for the long-term involved. Part I is the least technical part of the hand- sustainability of the microfinance institution or for the long- book; it requires no formal background in microfinance term effect on borrowers or beneficiaries. or financial theory. Its perspective is more macro than Now that the field of microfinance is more mature, it is that of parts II and III, which provide more detailed becoming clear that effective, efficient, and sustainable "how-to" discussions and are specifically focused on the institutions are needed to provide financial services well provision of financial services only. Part I will be of most suited to the demands of low-income clients. Both donors interest to donors and those considering providing micro- and practitioners are beginning to be held accountable for finance. Practitioners may also benefit from part I if they results. The focus is no longer solely on quantity--on the are considering redefining their market, changing their amount disbursed--but on the quality of operations. This institutional structure, offering additional services, or view is based on the notion that borrowers will buy micro- implementing different service delivery methods. finance products if they value the service; that is, if the Part II, which addresses more specific issues in the product is right for them. Borrowers are now being treated design of financial services (both lending and savings as clients rather than beneficiaries. Thus if they are to be effective and truly meet their development objectives, Figure 1. Relationship between Level of Analysis and donors must support MFIs that are "doing it right." To Technical Complexity in this Book do so, they need to understand how to both recognize and evaluate a good microfinance provider. As it turns out, both donors and practitioners are real- Macro ly on the same side--their joint goal is to make available Part I Issues to Consider When appropriate services to low-income clients. Therefore, it Providing Microfinance falls to donors and practitioners themselves to define best practices and to advocate policies that will encourage Level of analysis growth, consistency, and accountability in the field. The Part II Designing and Monitoring intent of this handbook is to provide a basis of common Financial Products and Services understanding among all stakeholders. Part III Measuring Performance Organization of the Book and Managing Viability The handbook has three parts. Part I, "Issues in Micro- Micro finance Provision," takes a macroeconomc perspective Low High toward general microfinance issues and is primarily non- Technical complexity 6 MICROFINANCE HANDBOOK products) and the development of management informa- ACCION Publications Department tion systems, will be of most interest to practitioners who 733 15th Street NW, Suite 700 are developing, modifying, or refining their financial prod- Washington D.C. 20005 ucts or systems, and donors or consultants who are evalu- Phone: 1 202 393 5113, Fax 1 202 393 5115 ating microfinance organizations and the appropriateness Web: www.accion.org of the products and services that they provide. Part II E-mail: publications@accion.org incorporates some basic financial theory and, accordingly, readers should have a basic understanding of financial Calmeadow Resource Center management. Readers should also be familiar with using a 365 Bay Street, Suite 600 financial calculator, computer spreadsheet, or both. Toronto, Canada M5H 2V1 Part III provides tools for evaluating the financial health Phone: 1 416 362 9670, Fax: 1 416 362 0769 of an MFI and a means of managing operational issues. Web: www.calmeadow.com The material focuses primarily on financial intermediation E-mail: resource@calmeadow.com (that is, credit and savings services, based on the assump- tion that the main activity of an MFI is the provision of CGAP Secretariat financial services). Social intermediation and enterprise Room G4-115, The World Bank development services are not addressed directly. However, 1818 H Street NW basic financial theory that is relevant to the financial man- Washington D.C. 20433 agement of MFIs delivering these services is provided as Phone: 1 202 473 9594, Fax: 1 202 522 3744 well. The material also underscores the importance of the Web: www.worldbank.org/html/cgap/cgap.html interrelationship between serving clients well and moving E-mail: CProject@Worldbank.org towards institutional and financial self-sufficiency. These two goals serve each other; neither is sufficient on its own. Micro Finance Network As the most technical section of the handbook, part 733 15th Street NW, Suite 700 III will be of particular interest to practitioners and con- Washington D.C. 20005 sultants. Donors will also benefit from part III if they Phone: 1 202 347 2953, Fax 1 202 347 2959 want to understand how MFIs should be adjusting their Web: www.bellanet.org/partners/mfn financial statements and calculating performance indica- E-mail: mfn@sysnet.net tors. While the technical information is fairly basic, some understanding of financial statements and finan- PACT Publications cial analysis is required. 777 United Nations Plaza The overall purpose of part III is to improve the New York, NY 10017 level of financial understanding and management in Phone: 1 212 697 6222, Fax: 1 212 692 9748 MFI operations. As donors come to understand both Web: www.pactpub.com the complexity of microfinance and that it can be deliv- E-mail: books@pactpub.org ered in a financially sustainable manner, knowledge of the more technical aspects of microfinance will become Part I--Issues in Microfinance Provision increasingly important to them in deciding whether to support institutions and programs. Chapter 1­The Country Context provides a framework Each chapter is designed to be used alone or in con- for analyzing contextual factors. It focuses on issues that junction with other chapters, depending on the specific affect the supply of microfinance, including the finan- needs of the reader. A list of sources and additional cial sector, financial sector policies and legal enforce- reading material is provided at the end of each chapter. ment, financial sector regulation and supervision, and Many of the publications listed at the end of each economic and social policies. chapter can be accessed through the following organiza- Chapter 2­The Target Market and Impact Analysis tions, either by mail or through their Web sites: looks at the demand for financial services among low- INTRODUCTION 7 income populations and presents ways of identifying a financial results to reflect more accurately the financial target market based on client characteristics and the position of an MFI. types of enterprises they operate. It also discusses impact Chapter 9­Performance Indicators details how to mea- analysis and how the desired impact affects an MFI's sure and evaluate the financial performance of the MFI, choice of target market. focusing on ratio analysis to determine how successful is Chapter 3­Products and Services considers the various the institution's performance and which areas could be services that low-income entrepreneurs might demand, improved. In addition, it provides various outreach indi- including financial and social intermediation, enterprise cators that can be monitored. development, and social services. An overview of well- Chapter 10­Performance Management presents ways known microfinance approaches is presented in the in which to improve the financial and resource manage- appendix. ment of microfinance institutions. It discusses delin- Chapter 4­The Institution discusses the various types quency management, staff productivity and incentives, of institutions that can effectively provide and manage and risk management, including asset and liability man- the provision of microfinance activities. It addresses agement. issues such as legal structures, governance, and institu- tional capacity, and also provides information on access- ing capital markets for funding. A Necessary Caveat Part II--Designing and Monitoring Financial Products Microfinance has recently become the favorite interven- and Services tion for development institutions, due to its unique potential for poverty reduction and financial sustainabil- Chapter 5­Designing Lending Products provides informa- ity. However, contrary to what some may claim, micro- tion on how to design or modify lending products for finance is not a panacea for poverty alleviation. In fact, a microentrepreneurs both to meet their needs and to poorly designed microfinance activity can make things ensure financial sustainability of the MFI. worse by disrupting informal markets that have reliably Chapter 6­Designing Savings Products provides infor- provided financial services to poor households over the mation on the legal requirements to provide savings, past couple of centuries, albeit at a high cost. types of savings products, and operational considerations There are many situations in which microfinance is for providing savings, including pricing. This chapter neither the most important nor the most feasible activi- focuses on the provision of voluntary savings; it does not ty for a donor or other agency to support. Infra- address forced or compulsory savings often associated structure, health, education, and other social services are with lending products. critical to balanced economic development; and each in Chapter 7­Management Information Systems (MIS) its own way contributes to a better environment for presents an overview of effective MIS, including microfinance activities in the future. Care should be accounting systems, loan-tracking systems, and client- taken to ensure that the provision of microfinance is impact tracking systems. It also provides a brief discus- truly demand driven, rather than simply a means to sat- sion on the process of installing an MIS and a summary isfy donors' agendas. evaluation of existing software packages. Part III--Measuring Performance and Managing Viability Sources and Further Reading Chapter 8­Adjusting Financial Statements presents the adjustments to financial statements that are required to Gonzalez-Vega, Claudio, and Douglas H. Graham. 1995. account for loan losses, depreciation, accrued interest, "State-Owned Agricultural Development Banks: Lessons inflation, and subsidies. Adjustments are presented in and Opportunities for Microfinance." Occasional Paper two groups: standard entries that should be included in 2245. Ohio State University, Department of Agricultural the financial statements and adjustments that restate Economics, Columbus, Ohio. 8 MICROFINANCE HANDBOOK Mutua, Kimanthi, Pittayapol Nataradol, and Maria Otero. Institutions." Sustainable Banking with the Poor, World 1996. "The View from the Field: Perspectives from Bank, Washington, D.C. Managers of Microfinance Institutions." In Lynn Bennett Women's World Banking Global Policy Forum. 1995. and Carlos Cuevas, eds., Journal of International "The Missing Links: Financial Systems That Work for Development 8 (2): 195­210. the Majority." Women's World Banking (April). New Paxton, Julia. 1996. "A Worldwide Inventory of Microfinance York. Part I­ Issues to Consider When Providing Microfinance CHAPTER ONE Understanding the Country Context T he overall political and economic environment of I What forms of financial sector regulation exist, and a country affects how microfinance is provided. are MFIs subject to these regulations? Government economic and social polices, as well I What economic and social policies affect the provi- as the development level of the financial sector, influence sion of financial services and the ability of micro- microfinance organizations in the delivery of financial ser- entrepreneurs to operate? vices to the poor. Understanding these factors and their As can be seen in figure 1.1, contextual factors affect effect on microfinance is called assessing the country con- how suppliers of financial intermediation reach their text. This process asks the following questions: clients. I Who are the suppliers of financial services? What This chapter uses a macroeconomic approach to place products and services do they supply? What role do microfinance in the overall context of a country and so governments and donors play in providing financial make clear how important macro-level policy and regula- services to the poor? tion are for developing microfinance providers and I How do existing financial sector policies affect the microenterprises. Practitioners and donors need to exam- provision of financial services, including interest rate ine the financial system to locate needs and opportunities policies, government mandates for sectoral credit allo- for providing microfinance services. Analyzing the country cation, and legal enforcement policies? context reveals whether changes in policy or in the legal Figure 1.1 Understanding the Country Context Contextual factors Clients 1. Financial sector policies and Suppliers of legal environment I Women financial intermediation I Interest rate restrictions I Microentrepreneurs I Government mandates I Small farmers I Formal sector institutions I Financial contract enforcement I Landless and smallholders I Semi-formal sector I Resettled persons institutions 2. Financial sector regulation and I Indigenous persons I Informal sector supervision I Low-income persons in institutions 3. Economic and social policy remote or subsistence areas IEconomic stability IPoverty levels IGovernment policies 11 12 MICROFINANCE HANDBOOK framework are needed to allow more efficient markets to providers that can be found in each group. The dividing emerge. lines are not absolute, and regulatory structures vary. For example, credit unions may fall in the formal sector in one country and in the semiformal sector in another. Suppliers of Financial Intermediation Services Formal financial institutions are chartered by the gov- ernment and are subject to banking regulations and The first step in understanding the context in which a supervision. They include public and private banks, microfinance provider operates is to determine who insurance firms, and finance companies. When these makes up the financial system. institutions serve smaller businesses or farmers (as is often "The financial system (or financial sector, or finan- the case with public sector financial institutions), there is cial infrastructure) includes all savings and financ- potential for them to move into the microfinance sector. ing opportunities and the financial institutions that Within the formal sector, private institutions generally provide savings and financing opportunities, as well focus on urban areas, whereas many public institutions as the valid norms and modes of behavior related to provide services in both urban and rural areas. Loans these institutions and their operations. Financial from private sector institutions are often large individual markets are the markets--supply, demand, and the amounts and are usually allocated to large, established, coordination thereof--for the services provided by private, and government-owned enterprises in modern the financial institutions to the nonfinancial sectors industrial sectors. Private formal sector institutions typi- of the economy." (Krahnen and Schmidt 1994, 3) cally mobilize the greatest amount of deposits from the To analyze a country's financial system, it is necessary general public. Public sector rural institutions often pro- to look at both the demand for and the supply of finan- vide agricultural loans as a means of developing the rural cial services. This section focuses on the supply of finan- sector. Funding sources include government-distributed cial services; demand will be examined in chapter 2. and foreign capital, with savings and deposits as sec- Understanding a country's financial system allows ondary sources. Processing of transactions entails microfinance providers to identify areas in which services detailed paperwork and bureaucratic procedures that or products for certain client groups are inadequate or result in high transaction costs, reinforcing the bias nonexistent. It can also identify institutional gaps and the toward relatively large loans. Box 1.1 describes the oper- potential for partnerships between different types of ations and problems of rural financial institutions in institutions to reach the poor cost effectively. Mexico. Intermediaries that provide financial services range Semiformal institutions are not regulated by banking from highly formal institutions to informal moneylen- authorities but are usually licensed and supervised by ders. Understanding the size, growth, number, gover- other government agencies. Examples are credit unions nance, and supervision of these institutions is an and cooperative banks, which are often supervised by a important part of assessing how the financial system bureau in charge of cooperatives. (NGOs are sometimes works. Financial intermediation varies with the services considered part of the semiformal sector, because they and products provided and depends to some extent on are often legally registered entities that are subject to the type of institution providing the services. Not all some form of supervision or reporting requirements.) markets have access to the same services and products. These financial institutions, which vary greatly in size, Determining which financial services are currently being typically serve midrange clients associated by a profes- provided to the differing markets is important in identi- sion or geographic location and emphasize deposit fying unserved or underserved clients. mobilization. Financial systems can generally be divided into the Semiformal institutions provide products and services formal, semiformal, and informal sectors. The distinction that fall somewhere between those offered by formal sec- between formal and informal is based primarily on tor and informal sector institutions. The design of their whether there is a legal infrastructure that provides loan and savings products often borrows characteristics recourse to lenders and protection to depositors. Table from both sectors. In many countries semiformal insti- 1.1 outlines the wide range of public and private tutions often receive donor or government support UNDERSTANDING THE COUNTRY CONTEXT 13 through technical assistance or subsidies for their sanctions within a family, a village, or a religious com- operations. munity substitute for legal enforcement. Credit terms Informal financial intermediaries operate outside are typically adapted to the client's situation. The total the structure of government regulation and supervi- amount lent, as well as the number and the frequency of sion. They include local moneylenders, pawnbrokers, installments, is fitted to the borrower's expected cash self-help groups, and NGOs, as well as the savings of flow. Little if any paperwork is involved in applying for family members who contribute to the microenter- a loan (Krahnen and Schmidt 1994, 32). prise. Often they do not comply with common book- Depending on historical factors and the country's keeping standards and are not reflected in official level of economic development, various combinations of statistics on the depth and breadth of the national these suppliers are present in the financial sectors of financial sector. Knowing where and how these finan- developing countries. Research increasingly shows that cial sources operate helps determine what services are the financial sector is complex and that there are sub- in demand. stantial flows of funds between subsectors. Identifying These institutions concentrate on the informal the suppliers of financial services in a given country or sector--on loans and deposits for small firms and house- region leads to a greater understanding of the financial holds. Often loans are granted without formal collateral system and also reveals gaps for a microfinance provider on the basis of familiarity with the borrower. Social to address. Table 1.1 Providers of Financial Intermediation Services Formal sector Semiformal sector Informal sector Central bank Savings and credit cooperatives Savings associations Banks Multipurpose cooperatives Combined savings and credit Commercial banks associations--rotating savings and Merchant banks Credit unions credit associations and Savings banks variants Rural banks Banques populaires Postal savings banks Informal financial firms Labor banks Cooperative quasi-banks Indigenous bankers Cooperative banks Finance companies Employee savings funds Investment companies Development banks State-owned Village banks Nonregistered self-help groups Private Development projects Individual moneylenders Other nonbank institutions Commercial Finance companies Registered self-help groups and savings clubs Noncommercial Term-lending institutions (friends, neighbors, relatives) Nongovernmental organizations (NGOs) Building societies and credit unions Traders and shopkeepers Contractual savings institutions NGOs Pension funds Insurance companies Markets Stocks Bonds Source: Food and Agriculture Organization 1995, 5. 14 MICROFINANCE HANDBOOK Box 1.1 Formal Sector Suppliers in Rural Mexico IN 1994 AND 1995 A WORLD BANK TEAM CONDUCTED AN As a consequence of the inefficiency of rural financial mar- economic and sector study in Mexico to examine the effi- kets, the funds did not flow to the best uses, hindering rural ciency and fairness of rural financial markets. The team con- development in Mexico. ducted case studies of 96 nonbank lenders, plus a household The study convinced the government of Mexico to ask survey of 800 rural entrepreneurs and a review of the regula- the World Bank team to propose measures that would help tory framework for the uniones de crédito (credit unions) and develop rural financial markets. The first measure proposed the sociedades de ahorro y préstamo (savings and loan compa- was to expand the distribution network, introduce adequate nies). All forms of formal and informal savings and loan ser- financial products and technologies, and carry out organiza- vices were analyzed. tional reforms, including the introduction of incentives and The study provided strong evidence about the poor per- internal controls. The second was to improve environmental formance of rural financial markets in Mexico. These mar- conditions by developing better policies and regulations on a kets proved to be shallow (only 45 percent of rural broad range of issues, such as secured transactions and the entrepreneurs received a credit transaction during 1992-94), regulation of intermediaries (mostly nonbanks) and of relat- segmented, noncompetitive, inefficient, and inequitable ed markets such as insurance. (showing strong biases against disadvantaged individuals). A pilot operation was designed to set up a network of The consequence was a weak supply of credit. The main rea- bank branches in localities of fewer than 20,000 inhabitants sons for this situation were underdeveloped institutional in which no formal financial intermediary was present. The infrastructure (two-thirds of the municipalities have no bank first step was to convince commercial banks that there is a offices), inappropriate banking technologies for small mar- business opportunity in rural financial markets. The message kets, and attenuated property rights. Past government inter- was that profitability is possible if five conditions are met. vention characterized by debt forgiveness and subsidized First, banks must have a good understanding of their market interest rates had contributed to the bad image of rural and offer simple financial products, such as deposits with finance. Few bankers from the private sector truly believed short maturities and small minimum balances, as well as there was a business opportunity in this field. On the short-term credit lines. Second, they must keep their fixed demand side, 75 percent of the households interviewed had costs low, with minimum investments and two to four never requested a loan from the formal financial sector. employees. Third, they must lend on the basis of a client's Rural entrepreneurs felt this would be too risky, given the character and reputation. Fourth, they must develop con- excessive amount of collateral required and the noninstitu- ducive incentive schemes, encouraging profit sharing and tional techniques used to enforce contracts. Many entrepre- efficiency wages. Finally, their internal control mechanisms neurs admitted that they were afraid of formal institutions must be credible, and there must be a credible threat of dis- and that they were unwilling to pay high transaction costs. missal for nonperforming staff. Source: Chaves and Sanchez 1997. Existing Microfinance Providers demand, oversupplying, saturating or distorting the market). Microfinance providers are found in both the public and the private sectors. To identify market gaps when THE EFFECT OF GOVERNMENT PROGRAMS ON PRIVATE PRO- providing or considering providing financial services to VIDERS. Depending on their approach, government-run microentrepreneurs, it is important to determine who microfinance programs can either contribute to or be detri- the existing providers are and how well the needs of mental to successful microfinance activities. Governments the market are being met. Donors can determine who that operate subsidized, inefficient microfinance programs is active in microfinance and who might require sup- through health, social services, or other nonfinancial state- port or funding. In areas where there is no microfi- run ministries or departments (or through a state-run bank- nance activity, practitioners can determine who their ing system) negatively influence the provision of sustainable competitors are and the effects they have on the mar- microfinance services (see box 1.2). Governments often ketplace (such as developing awareness, increasing have little or no experience with implementing microfi- UNDERSTANDING THE COUNTRY CONTEXT 15 Box 1.2 Do Microfinance Clients Need Subsidized Box 1.3 Credit Institutions as a Political Tool: Interest Rates? Debt Forgiveness in India MICROFINANCE CLIENTS TEND TO BORROW THE SAME RURAL FINANCIAL INSTITUTIONS THAT ARE ASSOCIATED WITH amount even if the interest rate increases, indicating that, governments often become the target of politicians. The within a certain range, they are not interest rate sensitive. influential clientele of these institutions makes them particu- In fact, people are often willing to pay higher rates for larly attractive political targets. India's government-appointed better service. Continued and reliable access to credit and Agricultural Credit Review Committee reported in 1989: savings services is what is most needed. During the election years, and even at other times, Subsidized lending programs provide a limited volume there is considerable propaganda from political of cheap loans. When these are scarce and desirable, the platforms for postponement of loan recovery or loans tend to be allocated predominantly to a local elite pressure on the credit institutions to grant exten- with the influence to obtain them, bypassing those who sions to avoid or delay the enforcement process of need smaller loans (which can usually be obtained com- recovery. In the course of our field visits, it was mercially only from informal lenders at far higher interest often reported that political factors were responsible rates). In addition, there is substantial evidence from for widespread defaults on the ostensible plea of developing countries worldwide that subsidized rural crop failures in various regions. credit programs result in high arrears, generate losses both The "willful" defaulters are, in general, socially for the financial institutions administering the programs and politically important people whose example oth- and for the government or donor agencies, and depress ers are likely to follow; and in the present democratic institutional savings and, consequently, the development set-up, the credit agencies' bureaucracy is reluctant to of profitable, viable rural financial institutions. touch the influential rural elite who wield much for- Microfinance institutions that receive subsidized fund- mal and informal influence and considerable power. ing are less likely to effectively manage their financial per- Farmers' agitation in many parts of the country can formance, since they have little or no incentive to become take a virulent form, and banners are put up in many sustainable. Subsidized interest rates create excess demand villages declaring that no bank officer should enter that may result in a form of rationing through private the village for loan recovery purposes. This dampens transactions between clients and credit officers. the enthusiasm of even the conscientious members of the bank staff working in rural areas in recovery Source: Robinson 1994. efforts. The general climate, therefore, is becoming increasingly hostile to recoveries. nance programs and no incentive to maintain long-term Source: Yaron, Benjamin, and Piprek 1997, 102. sustainability. Also, government microfinance programs are often perceived as social welfare, as opposed to economic development efforts. Some government programs grow too private sector microfinance organizations and that govern- large, without the necessary institutional base, and fail to ments should not lend funds directly to the poor. Others coordinate efforts with local NGOs or self-help groups. argue that the government should provide financial ser- Governments that forgive existing debts of the poor vices to microentrepreneurs but must do so on a commer- to state banks can have a tremendous effect on private cial basis to provide continued access to microfinance and sector MFIs, whose borrowers may mistakenly under- to avoid distorting the financial markets. Some advan- stand that their loans need not be repaid either. In gen- tages of government involvement include the capacity to eral, both donors and practitioners should determine the disseminate the program widely and obtain political sup- consequences of existing or past credit subsidies or debt port, the ability to address broader policy and regulatory forgiveness (see box 1.3). concerns, and the capacity to obtain a significant amount Much discussion surrounds the involvement of gov- of funds (Stearns and Otero 1990). A good example of a ernment in the provision of microfinance. Some people successfully run government operation is the Bank Rakyat argue that the government's role is to create an enabling Indonesia, a profitable state bank in Indonesia that serves environment for the success of both microenterprises and low-income clients (box 1.4). 16 MICROFINANCE HANDBOOK Box 1.4 Microfinance in Indonesia THE BANK RAKYAT INDONESIA, A STATE-OWNED BANK, RAN center, and its management has a free hand in determining a program of directed subsidized credit for rice farmers until its interest rates and other operating policies. In the 1980s 1983. The unit desa system was established in 1984 as a the Indonesian government implemented financial sector separate profit center within the bank. reforms that deregulated certain interest rates and abolished The unit desa system is a nationwide network of small interest rate ceilings. As the unit desa system struggled to village banks. The founding objectives were to replace become financially self-sustainable, market forces drove its directed agricultural credit with broad-based credit for any decisionmaking and it became increasingly subject to com- type of rural economic activity, to replace subsidized credit petition from other financial institutions. with positive on-lending rates with spreads sufficient to Management has a high degree of autonomy and full cover all financial and operational intermediation costs, and accountability for the performance of the unit desa system. to provide a full range of financial services (savings as well as This accountability is pushed down the line; every unit is credit) to the rural population. All these objectives were met responsible for its own lending decisions and profits. within just a few years. The system's phenomenal success in Monetary staff incentives and the prospects of promotion savings mobilization is its distinguishing achievement. reinforce individual accountability. Reforms to further Although the unit desa system forms an integral part of enhance the unit desa system's autonomy are currently being the Bank Rakyat Indonesia, it operates as a separate profit studied. Source: Yaron, Benjamin, and Piprek 1997. THE EFFECT OF PRIVATE MICROFINANCE PROVIDERS ON What Role Do Donors Play in Microfinance? OTHER SUPPLIERS. Private sector MFIs are often indige- nous groups or NGOs operated by local leaders in their Donors' interest in microfinance has increased substan- communities. They are frequently supported by interna- tially over the past few years. Virtually all donors, tional donors and international NGOs that provide including local, bilateral, and multilateral government technical assistance or funding, particularly in the start- donors and local and international NGOs, support up phase. Some private MFIs still subsidize interest rates microfinance activities in some way, providing one or or deliver subsidized services; others create self-sufficient more of the following services: operations and rely less and less on external donor I Grants for institutional capacity building funds. A few are beginning to access funding through I Grants to cover operating shortfalls commercial banks and international money markets. In I Grants for loan capital or equity addition to NGOs, both banks and nonbank financial I Concessional loans to fund on-lending intermediaries may provide financial services to the I Lines of credit microsector. I Guarantees for commercial funds Private sector MFIs that have operated in or are cur- I Technical assistance. rently operating in the country or region influence the Because donors are the primary funders of microfi- success of new providers by establishing client (or bene- nance activities (since most MFIs do not collect savings ficiary) expectations. For example, in Bangladesh the and are not yet financially viable enough to access com- Grameen Bank is so well known within the village pop- mercial funding), the approach they take to microfi- ulation that other microfinance providers routinely nance and the requirements they set for MFIs to access adopt the Grameen lending model, and when Grameen funding can greatly affect the development of the field Bank changes its interest rate or adds a new product, of microfinance. Most donors have moved away from clients of other MFIs demand the same. Microfinance subsidized lending and are focusing more on capacity organizations should be aware of the services offered by building and the provision of loan capital. However, other MFIs and the effects they may have on the effec- there is a variety of microfinance providers and a variety tive delivery of financial services (see table 1.2). of approaches taken by donors. In fact, most MFIs work UNDERSTANDING THE COUNTRY CONTEXT 17 Table 1.2 Private Institutions in Microenterprise Development Advantages Disadvantages Strong methodology based on experience Limited institutional capacity for scaling-up programs Committed, trained staff Low level of technical expertise, especially in financial and Capacity to reach grassroots associations information systems Little or no corruption Frequent lack of sufficient resources to expand programs Responsive and nonbureaucratic approach Isolated manner of operation Motivation to scale-up programs and aim for Little interaction or knowledge about government activity self-sufficiency in informal sector Capacity to associate among many to form a coordinated Limited vision that maintains small programs effort Source: Stearns and Otero 1990. with more than one donor, often developing separate activities. Donors should coordinate their efforts to cre- products to meet each donor's requirements. ate a consistent strategy toward microfinance based on While it is true that donors should avoid duplicating market segmentation and the comparative advantages of and (especially) contradicting each other's efforts, it is dif- each. A lack of coordination can quickly undermine the ficult to ensure that donors coordinate their efforts and efforts of good microfinance providers, as evidenced by create perfectly consistent strategies based on market seg- the many cases where donors (and governments) have mentation and comparative advantages. Particularly distorted the entire microfinance market by subsidizing because microfinance is perceived by some as a panacea interest rates and consequently making it difficult for for poverty alleviation, donors may want to take on the other microfinance providers to compete. same kinds of activities, especially lending to the poorest Donors can also provide an excellent learning of the poor. Meanwhile, areas in which donors may have source for microfinance providers. Many donors have real advantages, such as capacity building and policy dia- worked with specific institutions, in particular regions, logue, are greatly in need of resources. Indeed, a danger of and with certain approaches. Sharing these experiences widespread donor interest in microfinance is an oversup- with practitioners and other donors is extremely valu- ply of funds for on-lending in the face of limited institu- able. In many countries, donors and practitioners have tional capacity to take on these funds. set up informal networks to share their experiences, Some donors, such as the multilateral organizations, influence policies, and set industry standards. Box 1.6 may find their comparative advantage is in influencing describes an important initiative in this direction. policy reform and supporting the efforts of finance min- istries to strengthen supervisory bodies and create poli- cies that lead to a stable macroeconomy and financial Financial Sector Policies and Legal sector (see box 1.5). Donors can also be instrumental in Enforcement directing governments to various poverty alleviation ini- tiatives that further the development of microenterpris- After determining who is active in microfinance and es, such as infrastructure development and land transfer what the financial system looks like, it is necessary to programs. examine financial sector policies and the legal environ- It is helpful for both practitioners and donors to ment as they pertain to microfinance. Financial sector know what other donors are doing. Practitioners need to policy considerations include: know which donors support an approach consistent I Interest rate policies with their own, so that accessing funding does not mean I Government-mandated credit allocations changing the philosophy of their organization. Donors I Legal enforcement of contractual obligations and the need to avoid duplicating or counteracting each others' ability to seize pledged assets. 18 MICROFINANCE HANDBOOK Box 1.5 Multilateral Development Banks' Strategies for Microfinance BUILDING NATIONAL FINANCIAL SYSTEMS THAT SERVE THE ISafeguarding funds. Existing MFIs are stretching their lim- poorest segments of a society requires three essential compo- its to expand services. MFIs need incentives from donors nents: a countrywide, community-based financial services to maintain high standards of financial prudence while infrastructure; linkages between the grassroots infrastructure they expand their services. of the informal economy and the formal infrastructure of the I Maintaining a focus on the poor. MFIs are also striving to financial markets; and a favorable regulatory environment meet donor pressures for rapid financial sustainability. that allows microfinance institutions and microentrepreneurs This has led some to abandon the poorest of the poor. to flourish. The strategy of multilateral development banks Donors need to provide incentives for MFIs to guard should thus comprise three elements: capacity building, for- against upward pressure on their portfolio and keep their mal financial market linkages, and policy reform. The strate- focus on the poor--including the very poor--while striv- gy needs to be implemented at all levels in the given ing for high levels of financial sustainability. country--at the levels of the community, financial interme- I Promoting sectorwide performance standards. Given the diaries, and government regulatory bodies. large number of new institutional entrants to the sector, Multilateral development banks should develop their MFIs and donors should work together to promote con- internal capacity to implement strategy by instituting effec- sistent performance standards. tive project approval processes, coordinating with other I Building a cadre of microfinance technicians and training donors, training their staffs in microfinance program design, centers. Perhaps the most critical obstacle to expanding the and providing them with the requisite technical support and microfinance sector today is the lack of knowledgeable financing instruments to implement successful microfinance microfinance technicians to help institutions develop their programs. Multilateral development banks should be cautious management systems. Donors could help ease the shortage about flooding the sector with financial resources until there by supporting training and technical assistance centers. is appropriate institutional capacity to absorb them. I Promoting sustained linkages to commercial capital. International donors, including multilateral development Multilateral development banks need to provide funding banks, should focus on the following interventions: packages that encourage MFIs to move quickly toward I Expanding institutional capacity. The demand for commercial sources of financing, to avoid becoming microfinance is beginning to generate a flood of donor dependent on donated funds. financing for the sector. MFIs need grant funding to I Facilitating an enabling regulatory environment. MFIs and build their capacity before they can handle large vol- their clients face a daunting array of regulatory constraints, umes of financing. They require technical assistance to which the multilateral development banks can address in make the transition from start-up projects to financial- their policy reform work. MFIs are less inclined to engage in ly sustainable institutions. Some need further assis- policy advocacy work because they lack both the resources tance in making the transition to a formal financial and the experience. Donors need to encourage governments intermediary. to engage these implementing agencies in policy dialogue. Source: Yanovitch and Macray 1996. Interest Rate Policies there is reason to question the appropriateness of inter- est rate limits in any form, financial institutions that can Given the cost structure of microfinance, interest rate demonstrate services to the poor at a reasonable cost restrictions usually undermine an institution's ability to should receive exemptions in countries where usury laws operate efficiently and competitively (Rock and Otero are in effect. 1997, 23). Typically, restrictions do not achieve their MFIs need to price their loan products to allow for public policy purpose of protecting the most vulnerable full cost recovery. MFIs operating in countries that sectors of the population. Instead, they drive informal impose usury laws often have to establish pricing mecha- lenders underground, so that poor borrowers fail to ben- nisms that exceed the usury laws, particularly if they are efit from the intended low-cost financial services. While to become registered formal entities (see box 1.7). UNDERSTANDING THE COUNTRY CONTEXT 19 Box 1.6 The Consultative Group to Assist the Poorest Box 1.7 Usury Laws in West Africa THE CONSULTATIVE GROUP TO ASSIST THE POOREST ALL FINANCIAL INSTITUTIONS OPERATING IN THE EIGHT (CGAP)is the product of an agreement reached by donor countries that make up the West African Economic and agencies at the 1993 Conference on Hunger. CGAP was Monetary Union are subject to a usury law providing that launched in June 1995 by 9 donor agencies; its member- the maximum interest rate charged to a borrower should ship has since increased to 26. Its objectives are to not exceed double the discount rate of the union's central strengthen donor coordination in microfinance, to dis- bank. In 1996 the usury rate fluctuated around 13 per- seminate microfinance best practices to policymakers and cent. practitioners, to mainstream World Bank activity in this Microfinance institutions, NGOs, and donors were area, to create an enabling environment for MFIs, to sup- active in working with the central bank to persuade it to port MFIs that deliver credit or savings services to the grant an exemption that would enable microfinance insti- very poor on a financially sustainable basis, and to help tutions to charge high enough interest rates to reach established providers of microfinance to assist others to financial sustainability. Their advocacy efforts were suc- start such services. CGAP's goal is to expand the level of cessful, and the central bank is currently in the process of resources reaching the poorest of the economically active revising the usury law. Two usury rates that would no poor. longer be linked to the discount rate have been proposed: The group's secretariat administers a core fund, sup- one for commercial banks (18 percent) and one for ported by a World Bank contribution of US$30 million, microfinance institutions (27 percent). The flexibility of from which it disburses grants to microfinance institu- the central bank demonstrates an understanding of the tions that meet eligibility criteria. A consultative group important role that microfinance institutions are playing composed of member donors who make a minimum con- in West Africa. tribution of US$2 million pool their funds with the Source: Contributed by Cecile Fruman, Sustainable Banking with World Bank and establish policies criteria, operating pro- the Poor Project, World Bank. cedures, and guidelines for the secretariat. This group also reviews the secretariat's performance. A policy advisory group of practitioners from leading MFIs around the world advises the consultative group and the secretariat on these allocated funds must be aware of imposed condi- strategies for providing support to the poor. tions that may affect their operational sustainability-- Source: CGAP 1996a. for example, below-market interest rates. MFIs operating in countries in which the govern- ment has mandated sectoral credit allocations need to be Government Mandates for Sectoral Credit Allocations aware of these mandates, both as possible funding sources and as potential excess supply (often at discount- In many countries, governments mandate that formal ed rates) in the market. Rather than mandating credit financial sector institutions provide a certain percentage allocations, governments should be encouraged to focus of their portfolio or a certain volume of their assets to their policies on increasing outreach to the poor by cre- the informal or poorer segments of society or to certain ating enabling regulatory environments and building economic sectors. Special windows are created in com- institutional capacity. mercial banks or rediscounted lines of credit are provid- ed. For the most part, sectoral allocations do not work Legal Enforcement of Financial Contracts well because there are no incentives for commercial banks to participate. Many prefer to pay a penalty rather In some developing countries, the legal framework is than meet their obligations. unclear or does not allow for effective enforcement of While some MFIs may benefit from government financial contracts. Although the majority of microfi- credit allocations through accessing funding from com- nance providers does not require collateral equal to the mercial banks, sectoral mandates almost always distort value of the loans disbursed and hence is not concerned the market. MFIs participating with banks to access with its ability to legally repossess a client's assets, there 20 MICROFINANCE HANDBOOK are many instances in which laws relating to financial transactions influence the behavior of borrowers. Well- Box 1.8 Alexandria Business Association: defined property rights and good contract law help to Legal Sanctions minimize the costs of accomplishing both exchange THE ALEXANDRIA BUSINESS ASSOCIATION IN EGYPT BEGAN transactions and production transactions. Minimizing operations in 1983 offering loans to individual business transaction costs frees up resources that can be used to owners in amounts ranging from US$300 to US$15,000. increase overall welfare (USAID 1995). When a loan is received, the borrower signs postdated Some legal systems allow lenders to formally charge checks for each installment. These checks are enforceable borrowers when they fail to repay a loan, which can lead IOUs or promissory notes that use carefully crafted word- to imprisonment or fines. Usually just the threat of jail or ing that is recognized by the legal system. In addition to a a visit from police can work as an effective deterrent for signed check for each installment, one check is signed for borrowers who are considering defaulting on a loan. It is the whole amount of the loan. If a client defaults, the useful for microfinance providers to determine the vari- Alexandria Business Association can take the client to ous legal sanctions available when clients do not adhere court and claim the whole amount prior to the maturity to their agreements and the ability and effectiveness of date of the loan. In Egypt it is against the law to bounce a check; the the courts to enforce financial contracts. penalty may be imprisonment for three months to three The Alexandria Business Association provides a good years. If a client has not paid the installment by the end of example of the usefulness of an effective legal environ- the month due, the association's lawyers record the case in ment when providing microfinance (box 1.8). court. A letter from the court that lists the penalties for nonrepayment according to the law is delivered to the borrower. If payment is still not forthcoming, the lawyer Financial Sector Regulation and Supervision will take one of two routes: send a memo to the police and guide them to the borrower, or--if the lawyer is con- One of the most important issues in microfinance today is cerned about bribery of the police by the borrower-- the regulation and supervision of MFIs. As mentioned, direct recourse to the courts to have the borrower taken most informal and semiformal organizations providing into custody within two weeks. The lawyer attends the financial services to microenterprises do not fall under the court proceedings, and the process becomes a negotiation between the borrower and the judge. government regulations that are applied to banks and For association clients, legal pursuit can and does lead other formal financial institutions. Many nonbank MFIs, to prison. The legal director of Alexandria Business especially NGOs, operate on the fringes of existing regu- Association says that about 30 people have been jailed lations, especially with regard to deposit mobilization. In since the organization began operations. some instances they do so with the knowledge of the Source: Dichter 1997. authorities, who, for political reasons or simply for lack of time and resources, do not interfere. In other instances these nonbank MFIs simply avoid dealing with the issues MFIs. The information is most useful for govern- and proceed with deposit mobilization by calling it some- ments that are considering regulating the microfinance thing else. All parties involved in microfinance in a partic- sector. It is also helpful for both practitioners and ular country need to understand the dynamic of these donors to understand what is involved if and when an legally ambiguous operations. One important danger is MFI becomes regulated so that they know how the that as more bank and nonbank MFIs begin operating, MFI will be affected. Furthermore, if donors and authorities who have been disposed to liberal interpreta- practitioners are aware of the issues involved, they can tions of the regulations will be forced to invoke a much potentially influence government decisions regarding stricter construction of the laws, thus tipping the balance regulation of the sector and propose self-regulatory unfavorably from the point of view of those engaged in measures. microfinance. Financial regulation refers to the body of principles, The following discussion focuses on issues that rules, standards, and compliance procedures that apply to must be considered when regulating and supervising financial institutions. Financial supervision involves the UNDERSTANDING THE COUNTRY CONTEXT 21 examination and monitoring of organizations for compli- monly occurs when donors push "programs" and target ance with financial regulation. credit, rather than focus on meeting existing demand for Prudential regulation and supervision are designed to financial services. (Chavez and Gonzalez-Vega 1995): MFIs should also be regulated when they reach the I Avoid a banking crisis and maintain the integrity of size at which their failure would have consequences that the payments system reach far beyond owners and creditors. Finansol in I Protect depositors Colombia is a good example of the importance of regu- I Encourage financial sector competition and efficiency. lating growing financial intermediaries (box 1.9). To create an environment that is conducive to finan- The Finansol case highlights important lessons cial intermediation, governments and policymakers must regarding the regulation of MFIs created out of NGOs. ensure that financial regulation does not result in finan- "An NGO (nongovernmental organization) is not cial repression--in regulations that distort financial mar- inherently bound to the same standards of eco- kets and reduce the efficiency of financial institutions. nomic performance or financial prudence that may Examples of financial repression are imposed interest rate be reasonably expected in the business sector. ceilings, subsidized credit, and tax structures that dis- While this concern is not insurmountable, the courage investment in microfinance. Governments must NGO parent must ensure that it sets standards also ensure that supervisory bodies have the authority appropriate to the financial sector in dealing with and the capacity to implement the regulatory standards its regulated microfinance subsidiary. From the (Chavez and Gonzalez-Vega 1994). Finansol experience, it is apparent that NGOs Despite some success stories, MFIs probably reach should not be owners of MFIs unless they are less than 5 percent of the potential clients in the world operated separately, and the only financial link is today. Serving this market will require access to funding ownership. This relationship must be transparent, far beyond what donors and governments can provide. and any financial exchange must include proper Thus many MFIs want to expand their outreach by rais- transfer pricing. Also, NGOs should not fully con- ing funds from commercial sources, including deposits. trol the board and management of an MFI. It Most MFIs are significantly different from commer- would be appropriate for regulators to monitor the cial banks in institutional structure. Furthermore, man- financial health of an NGO when it is the majority aging a microloan portfolio differs in important ways owner of a regulated financial entity and in some from managing a conventional bank portfolio. MFIs cases to require consolidated financial reporting be and microloan portfolios cannot be safely funded with presented so that an accurate assessment of the commercial sources, especially public deposits, unless financial health of the regulated financial interme- appropriate regulation and supervision regimes are diary may be determined." (Barenbach and developed. Churchill 1997, 47) Even MFIs that do not mobilize deposits may benefit Furthermore the Finansol experience shows how reg- from entry rules that ensure the adoption of accepted good ulatory restrictions such as usury laws and limits on asset practices, as well as adequate recordkeeping and reporting. growth can greatly affect the success of an MFI. It also highlights the importance of having owners with some- When Should MFIs Be Subject to Regulation? thing at risk who are able to monitor their investment and who can help if there is a crisis. And finally, it MFIs should be regulated if and when they mobilize demonstrates the importance of flexibility on the part of deposits from the public. Individual depositors cannot the superintendency. be expected to monitor the financial health of an MFI "When regulation is warranted, it requires coherent and necessarily rely on the state to do this. MFIs should prudential guidelines that will further the growth of also be regulated when standards of good practice are the microfinance sector while protecting the inter- clearly needed, whether because there are no practicing est of small savers and supporting the integrity of organizations or institutions or because existing practi- the financial sector as a whole." (Barenbach and tioners are not operating effectively. The latter com- Churchill 1997, 1) 22 MICROFINANCE HANDBOOK Box 1.9 Regulating MFIs: The Case of Finansol FINANSOL EMERGED FROM CORPOSOL, A NONGOVERNMENTAL I Transparency. The lack of separation--operationally, organization (NGO) dedicated to providing services to financially, and culturally--between the new financial microentrepreneurs in Colombia's informal sector. In 1992 intermediary and the parent nongovernmental organiza- Corposol determined that it needed access to more capital tion led to a confusion of purpose, a lack of indepen- than its donors were able to provide through grants and guar- dence, inadequate management, and a lack of the antee funds in order to finance its impressive growth. It elect- financial information needed to assess the performance ed to create Finansol by buying the license of an existing of either the operations or the loan portfolio. finance company in which to book the loans. The license did Alarmed by Finansol's growth rate and the deteriorating not allow Finansol to accept general deposits from the public, quality of loans, the bank superintendency increased its however. The credit officers remained with Corposol, which supervision. In December 1995 the superintendency required also carried out other activities, including training. Finansol and Corposol to sever all operational ties. Because Finansol inherited an excellent loan portfolio, a proven Finansol was regulated, it was forced to establish provisions lending methodology, and consistent operational profitabil- on its portfolio. In 1996 Finansol provisioned for more than ity. Nevertheless, several factors led to the deterioration of its US$6 million, which greatly affected its profitability. The financial position. effect of the losses eroded Finansol's net worth to such an I Usury law. To comply with the usury law and cover its full extent that it violated the superintendency's rule that a regu- operating costs, Finansol charged a training fee with each lated financial institution may not lose more than 50 percent loan disbursement. This resulted in incentives for Corposol of its starting net worth in a given fiscal year. In May 1996 to disburse loans but not necessarily to collect loans. Finansol's equity position fell below this limit, creating a situ- I Rapid expansion of untested new products. This occurred ation in which the superintendency could have intervened. without the commensurate capacity to manage the The potential ramifications of intervention were significant. diversification. It could have sent a signal to the market that Finansol's crisis was I Deviation from basic banking principles. Corposol bor- escalating, rather than reaching a resolution. Not only could this rowed short and lent long. It also refinanced a large por- have caused Finansol's collapse due to lack of market confidence, tion of its nonperforming portfolio. but it could have affected the broader Colombian financial mar- I Management response to banking regulations. Colombian ket as well, since Finansol had issued paper amounting to regulators limited asset growth of regulated financial approximately US$30 million. Accordingly, averting this crisis institutions to 2.2 percent per month. As a nongovern- was of utmost concern to all parties, and intense efforts to launch mental organization, Corposol was not affected, so it a recapitalization plan were initiated to forestall intervention. retained a portion of the loans to permit the combined In December 1996 Finansol was successfully recapital- portfolio to grow at a faster rate. ized, and Corposol, the NGO, was forced into liquidation. Source: Churchill 1997. RISK FACTORS OF MFIS. While both MFIs and commer- risk profile. An appropriate approach to regulating cial banks are vulnerable to liquidity problems brought on MFIs must be based on an understanding of the differ- by a mismatch of maturities, term structure, and curren- ent risks and of the country's legal and institutional cies, the risk features of MFIs differ significantly from framework. those of commercial banks. This is due primarily to the MFIs' client base (low-income, assetless clients requiring MEANS OF REGULATING MFIS. Regulatory approaches small loans), lending models (small, unsecured loans for range from self-regulation, in which the industry devel- short terms based on character or group guarantees), and ops its own supervisory and governance bodies (such as ownership structure (capitalized by donors rather than credit cooperatives) to full regulation (either under exist- commercial investors/owners). For a detailed discussion of ing laws or through the establishment of specialized laws the risks specific to MFIs, see the appendix to this chapter. specific to MFIs). Another alternative is for MFIs to be Regulations designed for commercial banks are usu- regulated by the regulatory authorities, while contracting ally not suitable for MFIs because of MFIs' different a third party to perform the supervisory functions. UNDERSTANDING THE COUNTRY CONTEXT 23 The challenge facing regulators as they consider appropriate regulatory approaches to the microfinance Box 1.10 Enhancing the Effectiveness of Oversight sector is complicated by the great diversity of MFIs in institutional type, scale of operations, and level of pro- I Fraud protection. Experience shows that the altruistic fessionalism. To be effective, the regulatory approach origin of most MFIs does not exempt them from the in a given country must be consistent with the overall risk of serious fraud. The first line of defense against such fraud is internal audit procedures carried out by financial sector framework and must consider the vari- the MFI itself. Superintendencies should develop the ety of institutional types. Flexibility is crucial when capacity to review such procedures, perhaps through regulating MFIs, because many techniques and prac- mechanisms such as random client-level monitoring of tices are still experimental when a country is beginning loans. to seriously provide microfinance services (see box I Insolvency. Regulators must have the authority and will 1.10). to protect the system by imposing sanctions on insol- vent institutions. Having several licensed microfinance Considerations When Regulating MFIs providers can reduce the pressure to bail out failures. I Internal superintendency structures. As the microfinance It is important for regulators to establish minimum stan- industry develops in individual countries, superinten- dards for MFIs while at the same time remaining flexible dencies will need to organize themselves to regulate it. and innovative. At a minimum, when regulating and Some are creating specialized MFI departments. A less costly alternative might be to contract out reviews to supervising MFIs, five issues need to be considered experts familiar with MFI operations. (adapted from CGAP 1996b): I Minimum capital requirements Source: CGAP 1996b. I Capital adequacy I Liquidity requirements I Asset quality CAPITAL ADEQUACY. Capital adequacy refers to the level I Portfolio diversification. of capital in an organization that is available to cover its risk. Conventional capital adequacy concepts assume M INIMUM CAPITAL REQUIREMENTS . Minimum capital that it is clear what constitutes equity and debt. In non- requirements are set for all organizations entering the bank MFIs, it is necessary to distinguish equity and debt financial sector. This means that financial organizations by asking such questions as: Do donated funds consti- wanting to formalize must have a minimum amount of tute equity? Do concessional funds provided by donors capital to support their activities (stated as a currency constitute debt and therefore affect the leverage of the amount rather than as a percentage of assets). "Capital" MFI? These questions must be answered before deter- refers to the amount of equity an institution holds. (It mining the capital adequacy of an MFI. can also include some subordinated debt, depending on All financial institutions are required to have a mini- the specific rules of the regulatory board.) Since MFIs mum amount of capital relative to the value of their rely primarily on donor funds for capital contributions, assets. This means that in the event of loss of assets, the they may not have sufficient capital to meet these rules, organization would have sufficient funds of its own which can limit the formalization of microfinance orga- (rather than borrowed from depositors) to cover the loss. nizations. While regulators should be encouraged to set Capital adequacy standards refer to the percentage of minimum capital standards in accordance with their assets that is funded by debt. Stated differently, capital objectives of encouraging competition and low-risk adequacy standards refer to a maximum level of debt behavior, it is important for them to recognize that MFI versus equity (degree of leverage) that a financial institu- owners may consider social objectives over profit maxi- tion can have. (Assets are funded either by debt or equi- mization goals and that minimum capital requirements ty. Capital adequacy standards limit the proportion of may not be as powerful an inducement to sound gover- assets that can be funded with debt.) Current interna- nance as would generally be the case with standard com- tional standards outlined in the Basle Accord provide a mercial banking institutions. maximum leverage ratio of 12 to 1 or, stated the other 24 MICROFINANCE HANDBOOK way around, minimum capital of 8 percent of risk- on the stability of the market, regulators may find it pru- weighted assets. dent to set relatively high liquidity standards for MFIs, Assets are risk-weighted between 0 and 100 percent. taking into account the added costs that this implies. For example, an unsecured loan to an unknown entity is of high risk and would therefore likely be risk-weighted ASSET QUALITY. Asset quality represents the risk to earn- at 100 percent. A fully secured loan to the government ings derived from loans made by the organization. In of an industrial country poses little risk and would be other words, it measures the degree of risk that some of risk-weighted at a much lower rate. Once an institu- the loan portfolio will not be repaid. tion's assets (including off-balance-sheet items) are risk- Most MFIs do not require formal collateral and weighted, the total amount of capital required can be instead base their loan decisions on character, group determined. Obviously, the lower the risk, the lower the solidarity, and past repayment history. In many coun- risk-weighting, the lower the total assets, and, therefore, tries, bank regulations limit the percentage of the port- the lower the amount of required capital. Because capi- folio that may be extended as nonsecured loans. The tal is generally a more expensive source of funds than types of security available from microentrepreneurs are debt, banks want to have lower capital requirements and not usually recognized. Instead of collateral, regulators consequently try to have lower risk assets. The result is may consider the best indicators of asset quality to be beneficial for both the banks and the regulators because past performance of the portfolio and the current rate banks are discouraged from high-risk lending. of portfolio at risk (defined as the balance of outstand- Most MFIs are not fully leveraged due to their inabili- ing loans which have an amount past due). Bank regu- ty to borrow funds on the basis of their performance and lators should accept flexible definitions for loan security debt capacity. The greater part of the typical MFI's assets that enhance repayment incentives. Sufficient experi- is funded with donor contributions that are generally ence is now documented to demonstrate that such considered capital. Thus almost all MFIs easily meet cap- incentives can result in very low delinquencies and per- ital adequacy standards. For example, BancoSol, one of mit the delivery of financial services to sectors of the the most leveraged MFIs as of year-end 1995, had a risk- economy that do not have access to conventional weighted capital adequacy of 17 percent, much higher collateral. than is required under the Basle Accord. Strict loan-loss-provisioning policies should be There are strong arguments against allowing MFIs to enforced, commensurate with loan maturities, risk- leverage their equity capital (borrow a greater proportion weighting that considers the quality of security, and stan- of funds) as aggressively as commercial banks. As most dard methods that can be easily and consistently countries are relatively inexperienced with microfinance replicated. Regulators should encourage the adoption of (there is little empirical data about MFI performance), standard loan classification procedures to limit manage- regulators may wish to begin cautiously in fixing lever- ment discretion. A focus on staff incentives for dealing age ratios. It is suggested that an initial capital-asset ratio with arrears is encouraged. be no lower than about 20 percent for MFIs, subject to Regulations like those for commercial banks that downward adjustment as the institution and the indus- require elaborate loan documentation or involved credit try gain experience. approval procedures are not appropriate for MFIs, because the financial viability of MFIs depends on mini- LIQUIDITY REQUIREMENTS. Liquidity refers to the amount mizing the processing cost of microloans. Case-by-case of available cash (or near-cash) relative to the MFI's loan reviews are impractical. demand for cash. MFIs are exposed to high levels of liq- uidity risk: seasonal factors influence many of their P ORTFOLIO DIVERSIFICATION . Portfolio diversification clients; MFIs tend to depend on donors, whose funding refers to financial institutions' need to ensure that can be unpredictable; and their nondonor liabilities tend they have not concentrated their portfolio in one geo- to be short term. If the organization is operating in a sta- graphic sector or one market segment. Unlike other ble financial market, it may be able to deal with liquidity financial institutions, most MFIs have highly special- risk through short-term borrowings. However, depending ized portfolios that consist solely of short-term work- UNDERSTANDING THE COUNTRY CONTEXT 25 ing capital loans to informal sector clients. Although savings (box 1.11 describes Bolivia's initiative). In an MFI may have thousands of loans to diverse indus- South Africa and the Philippines local MFIs have tries, externalities could affect the entire market. formed associations as a first step toward educating Regulators should encourage MFIs to diversify their regulators and promoting policy dialogue. In loan portfolios through the development of guidelines Indonesia and Peru a third party has been engaged to that limit sector concentration. perform supervisory functions, either in place of or along with the bank superintendency (Berenbach and Country Approaches to Regulating MFIs Churchill 1997). In all countries that are considering establishing Although most countries have not yet established an new types of institutions, one of the main issues is the approach to regulating MFIs, Bolivia, Peru, and the limited capacity of bank superintendencies to super- West African Economic and Monetary Union have vise an increasingly large number of small and highly determined that a new type of legal institution was decentralized institutions. This is the case in Ghana required for MFIs that wanted to begin mobilizing (box 1.12). Box 1.11 Private Financial Funds in Bolivia THE BOLIVIAN GOVERNMENT RECENTLY CREATED A NEW Financial operations. Private financial funds hold broad type of institution, private financial funds, to "channel operational powers allowing for financial leasing, traditional resources to finance the activities of small businesses and lending, certificates of deposit to guarantee performance, and microenterprises in the productive and commercial sectors, factoring. This diversity of credit instruments will make it as well as to make loans to individuals for soft durable goods possible to adapt credit supply to the specific needs of small purchases. As an additional means of facilitating access to borrowers in both the productive and commercial sectors, credit for individuals, [private financial funds] may also including wage earners. engage in small-scale consumer credit operations" (from the Private financial funds may provide for their clients finan- law setting up the funds). cial services such as drafts and payment orders, foreign Incorporation. Private financial funds will be organized exchange operations, and the purchase and sale of foreign as corporations--an ideal form for financial intermedia- currencies. In addition, they may receive savings and time tion because of the legal stability of a commercial entity deposits and contract obligations with second-tier banking in civil society, and because it allows for timely increases institutions pursuant to the limits set by these intermediaries. or replenishment of equity when required by the Limits. To avoid risks that jeopardize their main purpose Superintendency of Banks and Financial Institutions. and are incompatible with the large amount of capital they Minimum operating equity. To incorporate and function involve, private financial funds are restricted from several private financial funds require the capital equivalent to at types of banking operations. These include capture of least US$1 million. This is significantly less than the US$3.2 demand deposits, foreign trade operations, trust operations million required for incorporating a bank. This start-up cap- and other charges of fiduciary duty, investments in enter- ital, together with a strict and prudential framework that prise capital, participation in the underwriting and place- establishes credit limits lower than those established for ment of securities, and mutual fund management. banks and a prohibition on making loans to the private A single loan may not exceed 3 percent of the net worth financial fund's own shareholders and managers, represents a of the private financial fund; credits with personal guarantees reasonable combination of equity backing and spreading of may not exceed 1 percent of the fund's net worth; a private credit risks. financial fund may not maintain a credit relationship with Private financial funds should maintain net worth equiv- an institution of the national financial system for more than alent to at least 10 percent of their assets and contingencies 20 percent of its net worth; and shareholders, statutory audi- weighted on the basis of risk; institutions already in opera- tors, directors and managers, and individuals or entities asso- tion shall have three years to bring their equity levels into ciated with a private financial fund may not obtain loans line with this new legal requirement. from the institution. Source: Loubiere 1995. 26 MICROFINANCE HANDBOOK Economic and Social Policy Environment Box 1.12 Nonbank Financial Institutions in Ghana After examining the supply of microfinance services, finan- IN 1993 THE GOVERNMENT OF GHANA ENACTED THE cial sector policies and the legal environment, and regula- Nonbank Financial Institutions Law to create an alterna- tion and supervision issues, the final area to examine tive financial market that would cater to the needs of the within the country context is the economic and social poli- broad sector of the Ghanaian economy that was not cy environment. Economic and social policies influence being served by traditional financial institutions. This both the ability of an MFI to effectively provide financial was a bold and laudable attempt to introduce diversity in the financial markets by creating an enabling environ- services and the types of activities microenterprises under- ment for local private participation. take. For example, economic policies that affect the rate of However, the prudential regulations that accompa- inflation in a country, the growth of the economy, or the nied the law contained serious mismatches. In Ghana degree of openness to market forces all influence the people regularly pay informal savings collectors to pro- required interest rate of loans as well as the ability of vide a daily savings service. Instead of receiving interest microentrepreneurs to successfully operate their businesses on their deposits, customers pay a commission to have and thus utilize financial services. The government's their savings collected and kept safe. As a consequence, investment in infrastructure development, the scale and 60 percent of Ghana's money supply is outside the for- depth of poverty in a country, and access to social services mal banking sector. However, the original regulations also affect the way in which a microfinance organization that governed savings and loans institutions limited the operates. If because of poor roads microentrepreneurs are frequency of deposits to once a week. To address this issue, regulated institutions lobbied for changes in the not able to reach markets, access health care services, or legislation to permit unlimited client deposits. send their children to school, their activities are affected In addition, there was concern by some nonbank and so, too, is their use of financial services. financial institutions that the minimum capital require- ment--US$100,000--was too low. After advocacy Economic and Political Stability efforts by the financial sector, the capital requirement is being reconsidered. High reserve ratios were also man- Volatility in all markets affects the risk and hence the choic- dated by the superintendency, requiring nonbank finan- es made by business owners and financial institutions. In cial institutions to place 57 percent of their deposits in general, the stability of financial and other markets makes first and secondary reserves, leaving the institution with microenterprises and consequently microfinance services too little to invest in productive activities. more viable (Yaron, Benjamin, and Piprek 1997). Another issue with the banking regulation involves MFIs should understand how various economic con- the reporting format. The superintendency is trying to supervise nine separate categories of financial institu- siderations and policies affect the poor. For most coun- tions with an omnibus prudential guideline and report- tries, there are numerous reports highlighting the ing format. Since mortgage companies, commercial economic situation and the various policies that affect banks, and savings and loan institutions, for example, the economy. Reports produced by the World Bank, provide very different services, they cannot report the governments, and research institutions should be gath- same information. To respond to this issue, nonbank ered to develop an overall assessment of the economic financial institutions have created an association to act as situation in a country. Two measures commonly consid- an advocate on their behalf. ered are the rate of inflation and the rate of growth of These observations lead to the following conclusions: gross domestic product (GDP). These measures provide I Bank supervisors need to learn how to supervise the some indication of the economic stability of a country. special characteristics of MFIs. I There is a need for a more involved and active super- visory service provided perhaps by an apex institu- INFLATION. Some degree of inflation is usually present in tion or subcontracted to an external body. all economies and must be considered when designing and pricing loan and savings products. Both practitioners Source: Dadson 1997. and donors should be aware of the inflation rate in coun- tries where they are working, as inflation results in a real UNDERSTANDING THE COUNTRY CONTEXT 27 cost to MFIs (and their clients) that must be covered by additional time spent with borrowers, at least initially, the interest generated on loans. which in turn may increase costs. (For detailed informa- MFIs can and do operate in countries with high rates tion regarding the provision of microfinance in transition of inflation. One possible means of addressing hyperin- economies, see Carpenter 1996.) Both donors and practi- flation is to index the loan contract to a hard currency tioners should be aware of this. such as the U.S. dollar or to a commodity. Other means General instability, corruption, civil war, party or trib- include maintaining funding and assets in U.S. dollars, al rivalries--all have obvious direct economic effects such utilizing debt rather than equity (if possible), and provid- as internal migration, outmigration, capital flight, and ing short-term loans (ensuring that they match the currency instability. However, there are also less tangible clients cash flow needs), which allow for periodic adjust- social and psychological effects. How people see their ment of interest rates. future, where they want to invest, the degree to which If MFIs borrow funds in a currency other than their they are willing to set up shop on a permanent basis, the own (such as U.S. dollars), inflation can result in either very kinds of products they will trade in, and what people foreign exchange gains or losses. Both donors and practi- are willing to buy with whatever assets they have available tioners need to determine who will benefit from these will all be affected by the political atmosphere. potential gains or will cover these losses. During periods of severe political unrest, MFI activi- ties may need to be postponed until a more stable envi- GROWTH RATE. A country's annual GDP growth provides ronment ensues, particularly where concern for the safety an indication of that country's economic stability and of both staff and clients exists. Donors must understand future growth prospects. Examining the rate of growth the need to postpone activities or slow down growth and gives both donors and practitioners an indication of the not pressure the MFI to continue to meet growth targets potential opportunities available to microentrepreneurs during periods of civil unrest (see box 1.14). and hence can affect the forecasted growth and funding requirements of the MFI. Although positive GDP growth is always preferred, Box 1.13 Microfinance in Areas of Political Unrest stagnant economies can be potential markets for the provi- I N THE WESTERN PART OF K ENYA , MICROFINANCE sion of microfinance services if unemployment increases operations exist, but they seem to be reaching limits of and more people become self-employed. However, stag- both outreach and sustainability. Initial loan uptake is nant or falling GDP levels also result in lower incomes and high, but the market for repeat loans and especially for less demand for products and services, which could result scaling up enterprise productivity is limited. Part of the in the failure of microenterprises. Furthermore, volatility reason has to do with the general neglect of the area by in the economy affects the risk and hence choices of finan- the government, which in turn is a function of political cial institutions and microentrepreneurs. In general, the and tribal rivalry. As a result, public sector investment and stability of financial and other markets makes microenter- other development efforts go elsewhere, and in these con- ditions, more and more people enter the informal econo- prises and consequently microfinance services more viable. my, not with an ambition to become entrepreneurs, but simply as a means of survival. An MFI wishing to operate TRANSITION AND POLITICAL UNREST. Transition economies in this area needs to make strategic decisions that may and situations of conflict or political unrest, where exist- involve compromising the orthodoxies of sustainability. ing systems of social networks have broken down and For example, it would make sense to reduce objectives need to be reestablished, pose additional concerns for so that in the medium term, the objectives would be MFIs (see box 1.13). Transition economies, by definition, to use microfinance as a means to smooth consumption are beginning to develop private sector markets and busi- and alleviate poverty, with the hope that as conditions im- nesses. Often the population is not familiar with business prove, a more viable market for a true MFI might become transactions and is generally not entrepreneurial in nature. possible. MFIs need to develop trust among their clients and Source: Contributed by Thomas Dichter, Sustainable Banking ensure that borrowers understand the responsibilities asso- with the Poor Project, World Bank. ciated with financial transactions. This will likely require 28 MICROFINANCE HANDBOOK also help to clarify or establish an MFI's or donor's Box 1.14 The Albanian Development Fund objectives. IN LATE 1992, AS ALBANIA ENTERED ITS TRANSITION TO A Investment in Infrastructure and Human Resource market economy, the Albanian Development Fund's rural Development credit program began forming village associations for the purpose of lending. By the end of 1996 the fund was An important consideration when providing microfi- active in 170 villages in 10 districts and had close to nance services is the existence of adequate infrastructure 7,000 outstanding loans totaling more than US$2.4 mil- lion (243 million lek). In spite of a general lack of under- (roads, communications facilities, water and sewer sys- standing of market forces in Albania, the Albanian tems) and of affordable social services such as health, Development Fund has achieved commendable loan education, and nutrition. repayment, with only one village fund failing to repay. "Improving rural roads and electricity, or provid- This success stems from a good understanding of the vil- ing matching grants for village-determined invest- lage networks and from the initial design of the project, ments such as watershed management, to increase which required full loan repayment before disbursement the earnings capacity of poor households is often of further loans. more cost-effective than providing financial ser- In 1992 Albania recorded an inflation rate of 237 per- vices, particularly where infrastructure is underde- cent a year, and its GDP growth was negative ­7 percent. veloped. Similarly, providing or upgrading rural It was during this time that the Albanian Development primary education and health facilities can achieve Fund began lending. To address such high rates of infla- longer-term benefits for the rural poor, especially tion and their effect on the value of its capital, the fund set the interest rate at 5 percent per year, with repayment for the poorest of the poor, by increasing the pro- of principal pegged to the U.S. dollar. In 1995 inflation ductivity of labor, the main asset of the poor." had decreased to 6 percent, and the Albanian (Yaron, Benjamin, and Piprek 1997, 42) Development Fund began charging 10 percent, with loan A lack of roads, electricity, communication facilities, repayment no longer pegged to the U.S. dollar. and other infrastructure will affect the means by which Beginning in 1997 Albania experienced civil unrest an MFI and the microenterprises it supports operate and caused by the collapse of pyramid savings schemes. should be taken into consideration by both donors and Although all borrowing was halted, credit officers contin- practitioners. ued to maintain contact with borrowers. Repayment of Identifying the availability of and access to social ser- loans in January and February (when the pyramid schemes vices for an MFI's client base helps to determine whether were beginning to collapse) continued to be 100 percent in clients' needs for services are being met and whether the rural areas and remained high in urban areas. This occurred despite an estimated 80 percent participation rate objectives of the MFI are suitable (that is, whether the of farmers in the pyramid schemes and is testimony to the MFI can provide only financial services while other gov- strength of the Albanian Development Fund system. ernmental organizations and NGOs provide other need- ed services). For example, the availability of education Source: Ledgerwood 1997. and health services greatly influences the capacity of microentrepreneurs to increase their enterprise activities. Poverty Levels Any provision of financial services to the poor should consider the availability of other nonfinancial services When examining the country context, it is important and the environment in which the MFI and the clients for both practitioners and donors to understand the operate. depth of poverty in the country in which they are Some MFIs form relationships with other service orga- operating and to determine government and donor nizations to coordinate the provision of financial services, policies that work toward alleviating or reducing nonfinancial services, or social services, taking advantage of poverty. An understanding of the degree of poverty in the comparative strengths of each organization. However, a country helps in estimating the size and needs of the it should be noted that the provision of social services such potential market for microfinancial services and may as health, education, and family planning should not be UNDERSTANDING THE COUNTRY CONTEXT 29 directly combined with the provision of financial services Microenterprises and small businesses may be affected (a point discussed in more detail in chapter 3). by government policies, including excessive regulation, prohibitive levels of taxation, inadequate government pro- Government View of the Microenterprise Sector tection against cheap imported products, laxity about black markets (which results in unfair competition for the It is important to determine the government's position microbusiness sector), harassment by government officials regarding the informal sector and microenterprise devel- for operating businesses on the streets, and inadequate opment, as this affects policies that may influence the services and high user fees in public market structures. behavior of microentrepreneurs. Some governments rec- Many of these regulations work effectively to encourage ognize the positive contribution of microenterprises to microenterprises to remain outside the legal or formal the economy and may actively include informal sector mainstream (see the discussion of Argentina's tax laws in development in the national plan. However, in many box 1.15). countries informal sector issues and their relationship to When policies and practices negatively affect clients' government policy receive little attention. Most policy businesses, an MFI or donor may choose to undertake frameworks favor large manufacturing sectors and are environment-level interventions, such as policy and biased against the informal sector and small enterprises. advocacy work, in addition to providing or supporting Most governments want to encourage the develop- the provision of financial services. Advocacy can include ment of businesses in their countries. Some governments helping clients organize to protest unfair policies or supplement general policy goals that apply to business treatment. MFIs can influence policy by working alone with specific policies and programs aimed at micro and small enterprises. It is helpful if policies are in place that establish a favorable climate for the start-up of new busi- Box 1.15 Tax Laws in Argentina nesses and the growth of existing businesses. Examples I N A RGENTINA ALL FORMAL ENTERPRISES MUST BE are policies that minimize the costs of licensing and regis- registered with the Dirección General Impositiva, the tax tering a business, provide easy access to information bureau, and must comply with various tax requirements. about laws and regulations, and facilitate commercial The most important tax is the impuesto al valor añadido, a codes, which establish rules to minimize the cost of 21 percent value-added tax. Businesses must charge this doing business by defining the rights and responsibilities tax on their sales and pay it on all products they buy. The of all parties to a transaction (USAID 1995). net difference between the two is due to the tax bureau The decision of governments to become actively annually. involved in microenterprise development through cred- Small businesses, including small farmers, that are not formalized enterprises are strongly penalized by this sys- it programs or other enterprise development services tem. When purchasing supplies, they must pay the tax up can affect the environment for private microfinance front and are sometimes charged a surtax of 50 percent on providers, either by negatively distorting the market or purchases as a penalty for not being registered enterprises, by positively contributing to the supply of services. raising the effective value-added tax to 31.5 percent. They Alternatively, governments can choose to support the are then unable to deduct the amount of tax paid from informal sector through macropolicies, the allocation of their revenue because, as nonregistered entities, they can- resources that affect microproduction, or work with not charge a value-added tax on the sale of their products NGOs that provide services and training. or services. Buyers purchasing crops from these unregis- "Active collaboration in this sense involves the tered businesses (small farmers) offer low prices, arguing establishment of a favorable climate to enable that they are taking a risk in buying from them and may these institutions to continue and expand their be penalized for doing so. The number of buyers willing to work with the informal sector is hence restricted, offer- work with support but no interference from gov- ing a limited range of channels for commercialization. ernment entities. This can include national recog- nition of the microenterprise sector, support to Source: Contributed by Cecile Fruman, Sustainable Banking with discuss the issue, funding research, and scaling up the Poor Project, World Bank. pilot programs." (Stearns and Otero 1990, 27) 30 MICROFINANCE HANDBOOK or through coalitions of similar organizations to lobby regulated financial intermediary, hindering bank appropriate government or regulatory bodies on behalf regulators from gaining an accurate financial pro- of their clients (Waterfield and Duval 1996). file of the regulated entity. Successful arrangements between the social organization (the nongovern- mental organization) and the regulated MFI can be Appendix 1. Risks in the Microfinance attained if the structure adheres to certain basic Industry principles, including transparency, arm's-length transactions, honest transfer pricing, and opera- There are four main areas of risk that are specific to tional independence. The regulated MFI must MFIs: portfolio risk, ownership and governance, man- maintain independent management and oversight agement, and "new industry." Portfolio risk was dis- of its financial services. Regulators can require that cussed in the main text of chapter 1. The remaining risks the composition of the MFI's board of directors are discussed below. The following discussion is taken include professional individuals who are prepared from Berenbach and Churchill 1997. to define sound policies and to oversee manage- ment. All directors should be held legally liable for Ownership and Governance the performance of the MFI, as is the case for directors of private sector companies. Although effective external regulation and supervision by I Sufficient financial depth. MFIs may be capable of rais- regulatory bodies are important to the health of the finan- ing the initial capital requirements from their found- cial system, no amount of external oversight can replace ing shareholders. However, these owners may lack the accountability that stems from proper governance and financial depth or the motivation to respond to addi- supervision performed by the owners of financial institu- tional calls for capital as required. Development insti- tions. The following points highlight issues of ownership tutions may require a lengthy approval process to and governance relative to adequate supervision of MFIs: secure disbursement of funds. Regulators can intro- I Adequate oversight of management. Often, investors in duce measures to compensate for the owners' limited MFIs are motivated by social objectives. As a result, capacity to provide additional capital. This can be they may not hold this investment to the same stan- addressed by the following options: establishing addi- dards that they apply to commercial investments. tional reserve funds, limiting dividend distribution Regulators should encourage meaningful participation until capital benchmarks are reached, and requiring of private investors, particularly local business leaders. standby financing commitments by MFI owners. These private investors are an important source of local governance and a valuable resource if the MFI Management Risks encounters difficulties. Furthermore, MFIs benefit from the participation of several significant sharehold- The management risks that apply to MFIs are generated ers who bring diverse backgrounds and perspectives to by the specific methods of providing financial services. the governance process. In addition to private inves- I Decentralized operational systems. A decentralized tors, regulators should encourage investment by other organizational structure that permits the provision of private or public development-oriented institutions financial services directly at the borrower's or saver's with microfinance or related development finance location is central to microfinance. Consequently, experience. senior management must train and supervise mid- I Organizational and ownership structures. If a social level management, introduce appropriate reporting organization, which is funded with public resources systems, and maintain adequate communication sys- and does not have owners, oversees the manage- tems so that uniform policies and procedures are ment and determines the policies of the regulated adopted. Furthermore, decentralized operating meth- financial intermediary, social objectives may take ods create an environment that can easily be subject priority over financial objectives. This may make it to fraudulent practices. Regulators should require difficult to determine the true performance of the MFIs to maintain strong internal auditing capabili- UNDERSTANDING THE COUNTRY CONTEXT 31 ties and aggressive internal auditing procedures. implemented on a broad scale. It may be appropriate Guidelines should be established for the performance to limit the number of new products or services that of external audits for MFIs. Finally, adequate mea- are introduced at any one time. The challenge facing sures of internal communications and financial con- MFIs is to conduct a large volume of very small trans- trols are essential. actions and to do so sustainably. Given this challenge, I Management efficiency. MFIs offer a high-volume, it is most appropriate to limit MFIs to relatively sim- repetitive service that operates on very tight margins. ple products and services that can be easily mastered. If funds are not relent promptly, earnings will suffer. Regulators should ensure a high quality of manage- ment to ensure that brisk and timely services are Sources and Further Reading provided. I Management information. Decentralized operating Basle Committee on Banking Regulations and Supervisory methods, high volume of short-term loans, rapid Practices. 1988. "International Convergence of Capital portfolio turnover, and the requirement for effi- Measurement and Capital Standards." Basle, Switzerland. cient service delivery make accurate and current Berenbach, Shari, and Craig Churchill. 1997. Regulation and Supervision of Microfinance Institutions: Experience from portfolio information essential for effective MFI Latin America, Asia and Africa. Occasional Paper 1. management. In general, MFIs have not focused Washington, D.C.: MicroFinance Network. on providing adequate and appropriate financial Carpenter, Janney. 1996. Microcredit in Transitional Economies. information for making judgments about their Organisation for Economic Co-operation and financial viability. Government regulators, donors, Development, Centre for Co-operation with the Economies potential depositors, and other types of potential in Transition, Paris, France. creditors all need fairly standard information CGAP (Consultative Group to Assist the Poorest). 1996a. about financial viability in order to make informed "The Consultative Group to Assist the Poorest--A judgments. Reporting requirements for regulated Microfinance Program." CGAP Focus Note 1. World institutions make it necessary for MFIs to be able Bank, Washington, D.C. to produce accurate, useful, and timely manage- ------. 1996b. "Regulation & Supervision of Micro-Finance ment information. Institutions: Stabilizing a New Financial Market." CGAP Focus Note 4. World Bank, Washington, D.C. ------. 1997. "How CGAP Member Donors Fund Micro- New Industry finance Institutions." CGAP Focus Note 11. World Bank, Washington, D.C. A number of the risks that face MFIs stem from the fact Chaves, Rodrigo A., and Claudio Gonzalez-Vega. 1993. that microfinance is a relatively new field. Formal finan- "Should Principles of Regulation and Prudential cial services may also be new to the micro market. Supervision Be Different for Microenterprise Finance I Growth management. MFIs that expand into new Organizations?" GEMINI Working Paper 38. U.S. markets often face little competition. These institu- Agency for International Development, Washington, tions can experience dramatic growth in their initial D.C. years of operation. Regulators should closely monitor ------. 1994. "Principles of Regulation and Prudential Supervision and their Relevance for Microenterprise MFIs that dramatically surpass the growth projections Finance Organizations." In Maria Otero and Elisabeth presented in the license application. Rhyne, eds., New World of Microenterprise Finance. West I New products and services. Although this industry has Hartford, Conn.: Kumarian Press. made considerable advances in the design of appropri- ------. 1995. "Making the Leap into the Formal Financial ate microfinance products and services, the field System: Structuring Regulation and Prudential Supervision remains relatively young and untested. It is difficult to for Microenterprise Finance Organizations." GEMINI assess when a new product, or service is an ill- Microenterprise Development Brief 19. U.S. Agency for conceived deviation from an existing model or a International Development, Washington, D.C. breakthrough in new services for the market. New Chaves, Rodrigo, and Susana Sánchez. 1997. "Formal Sector products and services must be well tested before being Suppliers in Rural Mexico." World Bank, Latin America and 32 MICROFINANCE HANDBOOK the Caribbean Region, Finance, Private Sector, and ACCION International, November 27­28, Washington, Infrastructure Unit, Washington, D.C. D.C. Christen, Robert P. 1995. "Issues in the Regulation and Rhyne, Elisabeth. 1995. "Major Issues in Supervision and Supervision of Microfinance." Paper presented at the con- Regulation." Paper presented at the MicroFinance Network ference on Regulation and Supervision of Microfinance Conference, November, Manila. Institutions, sponsored by ACCION International, Robinson, Marguerite. 1994. "Savings Mobilization and November 27­28, Washington, D.C. Microenterprise Finance: The Indonesian Experience." In Maria Churchill, Craig, ed. 1997. Regulation and Supervision of Otero and Elisabeth Rhyne, eds., The New World of Micro- Microfinance Institutions Case Studies. Occasional Paper 2. enterprise Finance. West Hartford, Conn.: Kumarian Press. Washington, D.C.: MicroFinance Network. Rock, Rachel, and Maria Otero, eds. 1997. From Margin to Cuevas, Carlos E. 1996. "Enabling Environment and Mainstream: The Regulation and Supervision of Microfinance. Microfinance Institutions: Lessons from Latin Amer- Monograph Series 11. Washington, D.C.: ACCION ica." Journal of International Development 8 (2): 195­ International. 210. Rosenberg, Richard. 1994. "Beyond Self-Sufficiency: Licensed Dadson, Christine. 1997. "Citi Savings and Loans, Ghana." Leverage and Microfinance Strategy." U.S. Agency for Adapted from Craig Churchill, ed., "Establishing a International Development, Washington, D.C. Microfinance Industry." MicroFinance Network, Stearns, Katherine, and Maria Otero, eds. 1990. The Critical Washington, D.C. Connection: Governments, Private Institutions, and the Dichter, Thomas. 1997. "Alexandria Business Association" Informal Sector in Latin America. Monograph Series 5. Case Study for the Sustainable Banking with the Poor Washington, D.C.: ACCION International. Project. World Bank, South Asia Social Development Unit, USAID (U.S. Agency for International Development). 1995. Washington, D.C. "Policy Goals, Reform, and Microenterprise Development." Drake, Deborah, and Maria Otero. 1992. Alchemists for the Microenterprise Development Brief 25. GEMINI Project, Poor: NGOs as Financial Institutions. Monograph Series 6. Washington, D.C. Washington, D.C.: ACCION International. ------. 1998. Microenterprise Development Policy Paper. FAO (Food and Agriculture Organization). 1995. Washington, D.C. "Safeguarding Deposits: Learning from Experience." FAO Vogel, Robert C. 1994. "Other People's Money: Regulatory Agricultural Services Bulletin 116. Rome. Issues Facing Microenterprise Finance Programs." Gray, Tom, and Matt Gamser. 1994. "Building an International Management and Communications Institutional and Policy Framework to Support Small and Corporation, Washington, D.C. Medium Enterprises: Learning from Other Cultures." Waterfield, Charles, and Ann Duval. 1996. CARE Savings and Implementing Policy Change Project. Development Credit Sourcebook. Atlanta, Ga.: CARE. Alternatives, Inc., Bethesda, Md. Women's World Banking Global Policy Forum. 1995. Krahnen, Jan Pieter, and Reinhard H. Schmidt. 1994. "The Missing Links: Financial Systems That Work for Development Finance as Institution Building: A New Approach the Majority." Women's World Banking (April). to Poverty-Oriented Lending. Boulder, Colo.: Westview Press. Yanovitch, Lawrence, and Dennis Macray. 1996. "The Krutzfeldt, Herman. 1995. "The Experience of BancoSol." Multilateral Development Banks and the Microfinance Sector: Paper presented at the MicroFinance Network Conference, Opening Financial Markets to the World's Self-Employed November, Manila. Poor." Paper prepared for the Congressional Task Force on Ledgerwood, Joanna. 1997. "Albania Development Fund." the U.S. and Multilateral Development Banks and the Center Case Study for the Sustainable Banking with the Poor for Strategic and International Studies. Foundation for Project. World Bank, South Asia Social Development International Community Assistance, Washington, D.C. Unit, Washington, D.C. Yaron, Jacob, McDonald P. Benjamin, Jr., and Gerda L. Lobíere, Jacques Trigo. 1995. "Supervision and Regulation Piprek. 1997. Rural Finance; Issues, Design, and Best of Microfinance Institutions: The Bolivian Experience." Practices. Environmentally and Socially Sustainable Paper presented at the conference on Regulation and Development Studies and Monographs Series 14. Supervision of Microfinance Institutions, sponsored by Washington, D.C.: World Bank. CHAPTER TWO The Target Market and Impact Analysis A target market is a group of potential clients who oriented businesses, or specific economic sectors) it wish- share certain characteristics, tend to behave in es to support. Furthermore, the clients' cash flow and similar ways, and are likely to be attracted to a capacity for debt must match the services being offered. specific combination of products and services. A target Once a target market is selected, it is important to market represents a defined market segment that con- determine if that market is in fact being reached and tains identifiable clients who demand or represent a what impact the provision of financial services has on potential demand for microfinance services. In selecting that market. This is done by carrying out impact analy- a target market for microfinance services, MFIs need to sis, discussed in the second half of this chapter. determine their own objectives, understand what moti- Chapter 1 examined the supply of financial services; vates a group of clients, and assess whether the target this chapter focuses on demand and on how to determine market can be reached in a way that will eventually be the impact of meeting that demand on the target market. financially sustainable. Most important, the market must It summarizes what stakeholders need to know to deter- be chosen based on effective demand for financial ser- mine whom they want to provide services to (practition- vices and the capacity within that market to take on ers) or support the provision of services to (donors) and debt. whether they achieve the desired outcomes. Organizations that do not define their objectives, and This chapter covers two main topics: identifying a hence their target market, or that fail to design their target market, and analyzing the impact of services on products to meet the needs of this market often have dif- that market. ficulty managing their operations and staying focused. For example, an organization that simply wants to pro- vide financial services to the poor without defining who Objectives of the Microfinance Institution the poor are and which level of the poor it wants to reach often ends up operating with different credit mod- Selecting a target market depends on the objectives of els, trying to serve different groups and satisfy different the microfinance service provider and the perceived donors. Both donors and practitioners should be clear demand for financial services. In any country there are about appropriate target markets for particular MFIs and unserved or underserved enterprises and households, must continually focus on meeting those markets' needs. ranging from the ultra-poor, who may not be economi- Specifically, donors should not encourage MFIs to oper- cally active, to small growing enterprises that provide ate in areas or market segments where they have no expe- employment in their communities. This range or con- rience, expertise, or resources. tinuum constitutes the demand side for microfinance Target markets can be identified by the characteristics services. Often the supply side does not offer a corre- of the clients (poverty level, gender, ethnicity, caste, reli- sponding continuum of services. MFIs need to supply gion, and so forth) the MFI wants to serve and the type services that fill the gaps and integrate the unserved or level of business activity (existing businesses, growth- groups into the market. 33 34 MICROFINANCE HANDBOOK The goal of MFIs as development organizations is to Direct and Indirect Targeting service the financial needs of unserved or underserved markets as a means of meeting development objectives. Before discussing how MFIs identify their target markets, These development objectives generally include one or it is worthwhile to clarify the meanings of and differences more of the following: between direct and indirect targeting and to explain why I To reduce poverty indirect targeting (or "identifying" the market) is the pre- I To empower women or other disadvantaged popula- ferred means of reaching a target market with financial tion groups services. I To create employment Direct targeting generally refers to the allocation of a I To help existing businesses grow or diversify their specific amount of funds to provide credit to a particular activities sector of the economy or population. Direct targeting is I To encourage the development of new businesses. founded on the belief that because certain groups (the In a World Bank study of lending for small and poor, specific castes) or sectors (agriculture, fisheries) are microenterprise projects, three objectives were most fre- unable to access credit (or to access it at affordable quently cited (Webster, Riopelle, and Chidzero 1996): prices), credit must be made accessible through a govern- I To create employment and income opportunities ment or donor mandate. In some cases the government through the creation and expansion of microenterprises or donor also subsidizes the clients' cost of borrowing. I To increase the productivity and incomes of vulnera- In spite of the good intentions of such a strategy, one ble groups, especially women and the poor should be skeptical about its usefulness (see box 2.1). I To reduce rural families' dependence on drought- First, due to the fungibility (exchangeability) of money, prone crops through diversification of their income- it is almost always impossible to know what any given generating activities. loan "really" finances. Second, it is highly doubtful that Given the large number of conditional variables in anyone knows better than the recipients themselves what each country context, every organizational decision to is good for them. Third, the strategy rests on the idea enter or serve a target market will involve balancing the that "commerce" and "consumption" are less valuable conditions in that market. This decisionmaking process must keep in mind the two long-term goals of microfi- nance: outreach, serving those who have been consistently Box 2.1 Targeted Credit Activities underserved by financial institutions (such as women, the poor, and indigenous and rural populations), and sus- I N MOST DEVELOPING COUNTRIES THERE HAS BEEN tainability, generating enough revenue to cover the costs extensive government intervention in the allocation of of providing financial services. Depending on which tar- credit. Although a degree of intervention may have been get market is selected, there are consequences to the useful during the early stages of development, many coun- tries have come to recognize that this policy has had an MFI's financial position, because costs will be affected. adverse effect on industrial and financial development. In short, there are trade-offs involved in the decisions The evidence suggests that direct targeted credit programs about objectives and how to reach them. The central cal- have been an inefficient way of redistributing income and culus for an MFI concerns which objectives it can afford dealing with imperfections in the goods market. However, to set and for how long. some programs that were well designed and narrowly MFIs need to determine where there is unmet focused have been reasonably successful in dealing with demand for microfinance services and which target group specific imperfections in the financial markets (for exam- matches their objectives. For example, if an MFI's objec- ple, the lack of risk capital). In the future governments tive is to reach the very poor with financial and other ser- should attack the conditions that made direct credit vices, its target market will differ from an MFI that appear desirable--imperfections in markets or extreme wishes to serve the economically active poor with only inequalities in income--instead of using directed credit programs and interest rate subsidies. financial services. In addition, some MFIs may wish to focus on a particular economic sector or level of business Source: Pederson 1997. activity as a means of achieving their objectives. THE TARGET MARKET AND IMPACT ANALYSIS 35 socially than "production" and "income generation" and tors, but direct targeting imposes eligibility criteria, while should therefore not be financed. This naively transfers indirect targeting designs appropriate products and notions from industrialized to developing countries and services. from the world of the wealthy to the world of the poor. For example, if people are engaged in trade, this is likely to be a better line of business for them. However, con- The Importance of Adequate Cash Flow and the sumption by a poor household is very often a way of Capacity to Service Debt strengthening income-generation capabilities through improved education and nutrition. Debt capacity is an important consideration in determin- Direct targeting generally leads to credit diversion and ing the demand for financial services. When identifying low repayment. It also results in substantial costs of mon- their target market, MFIs must consider clients' and itoring eligibility and compliance. Furthermore, potential potential clients' cash flow as well as their ability to repay clients who have profitable but unfinanced or underfi- loans. Cash flow is the comparison of cash inflows and nanced businesses may be excluded because they do not outflows. Debt capacity is the amount of additional debt fit the profile. Alternatively, people who do match the a client can take on without running the risk of inade- qualifications and receive credit may not have entrepre- quate cash flow and consequent loan default.1 neurial skills or a profitable venture in need of financing. MFIs need to consider debt capacity as opposed to Indirect targeting means that products and services are basing credit decisions on a "credit need" approach that designed for and aimed at people who are beyond the risks future trouble for both lenders and borrowers. A normal frontiers of formal finance, instead of mandating credit-need assessment yields unreliable information specific funds to particular groups who fit a narrowly because self-reported credit need involves "wishing." By defined profile. Indirect targeting focuses on those who focusing on credit need rather than debt capacity the cannot take advantage of income-generating opportuni- lender risks not getting the money back, and the borrow- ties because of market imperfections or other barriers to er risks serious debt. This is because the need for credit financial services. With indirect targeting, self-selection and the ability to repay debt cannot be assumed to takes place by virtue of the design of microfinance match. services. Economists refer to this as "incentive com- No matter how the target market is identified, it is patibility"--the terms and conditions are such that imperative for MFIs to ensure that each client and target unwanted clients will not be interested, both because the group can generate enough cash to repay the loan on products are less attractive and because the set of require- time. This in turn determines the size of the potential ments imposed to access the services will seem too bur- target market (differentiating between "credit need," or densome. This happens because populations outside of what borrowers say they want, and "effective demand," the target group have other alternatives for financial ser- or what borrowers can and are willing to borrow and vices that the target group does not have. repay). For example, an MFI that seeks to provide the very "Lenders are able to recover loans on schedule only poor with credit should design its loan products so that when the repayment capacity of the borrower the relatively high interest rates and small size of loans are equals or exceeds debt service, which consists of attractive only to the very poor. The MFI may also principal and interest due for payment. Borrowers require group guarantees and weekly attendance at group are able to repay their loans on time without suf- meetings. More affluent clients usually see this as an fering hardship only when their repayment capaci- inconvenience, which makes the credit attractive only to ty equals or exceeds the debt service due according poorer clients. to the loan contract. These simple, self-evident The primary difference between direct and indirect relationships define the role that credit plays in targeting lies in the means that the MFI uses rather than development and influence the fate of efforts to in the target group. Both direct and indirect targeting expand the frontier of formal finance." (Von may reach the same population groups or economic sec- Pischke 1991, 277) 1. This discussion is drawn from Von Pischke (1991). 36 MICROFINANCE HANDBOOK To determine potential clients' debt capacity, the first equally important to invoke the underlying psychological consideration is their cash flow. It is then necessary to basis for a minimal equity contribution. People care assess the degree of risk associated with this cash flow and about an asset if they have worked for it or own it. This other claims that may come before repaying the MFI seems to be a universal trait of human nature. If the MFI loan. Adjusting the debt capacity of a borrower for risk lends 100 percent of the cost of projects or assets, bor- should reflect reasonable expectations about adverse con- rowers have little or no risk since their own money is not ditions that may affect the borrower's enterprise. at stake. By requiring a contribution by the borrower Adjustment for adversity should reflect the lender's will- (even if it is in fact a token from the lender's viewpoint), ingness to assume the risks of borrowers' inability to MFIs hope to increase responsible behavior among bor- repay. The greater the MFI's capacity to assume risk, the rowers and thereby reduce the chances of loan default. higher the credit limits the lender can offer. However, MFIs need to monitor where their clients Other claims that need be considered are debts to are accessing the capital for this minimal equity contribu- other lenders (claims by informal lenders generally rank tion. If clients need to borrow funds from another source ahead of those of formal credit institutions) and house- (at potentially higher interest rates) to contribute equity, hold expenses such as food and fuel, taxes, school fees, they may be increasing the risk of default. and expenditures for emergencies, important social oblig- ations, and ceremonies. Moral Hazard MFIs need to be conservative when estimating the debt capacity of a potential target market, because deter- Managing debt capacity and ensuring a minimal equity mining clients' debt capacity is an important part of requirement offset moral hazard. Moral hazard is identifying a target market and designing appropriate defined as "the incentive by someone (an agent) who products and services for this market. For the most part, holds an asset belonging to another person (the princi- borrowers do not go into debt to the full extent of their ple) to endanger the value of that asset because the agent debt capacity. Economists refer to this as "internal credit bears less than the full consequence of any loss" (Chaves rationing." If borrowers attempt to fully exhaust their and Gonzalez-Vega 1994). For example, a borrower's debt capacity, they are usually behaving opportunistically efforts to repay may be largely unobservable by an MFI; and are at risk of not being able to manage debt. it cannot readily determine what is attributable to the lack of effort by the client as opposed to bad luck or Minimal Equity Requirement external forces. An MFI that could observe the borrower could relate the terms of the loan contract to the effort In addition to assessing clients' debt capacity, MFIs put in by the borrower, but because it costs too much to should consider clients' ability to contribute a minimum do this, the best the MFI can do is to ensure that the amount of equity. In other words, loans should not cash flow and debt capacity of the borrower are suffi- finance the entire business activity. Some MFIs set a cer- cient to service the debt. MFIs must also set terms of the tain percentage of equity as one of the loan conditions. loan that are acceptable to the borrower and result in Even when the target group is start-up businesses owned behavior that the MFI prefers. This usually means giv- by the very poor, where equity contributions in the strict ing the borrower more risk (by being less willing to for- sense are not possible, some MFIs require the pledging of give loans when things go wrong) than the MFI would a household asset before granting the loan. Other forms be ready to assume if it had better information (Yaron, of equity can be compulsory savings or an amount con- Benjamin, and Piperk 1997). tributed by the borrower to the project in the form of a "membership fee" or "loan application fee." While these Market Size fees or contributions are financially insignificant from the lenders' point of view, they carry financial and psycho- MFIs should estimate the size of the market for microen- logical weight for the prospective borrowers. terprises that can benefit from financial services, lest self- In banking terms, minimal equity requirements reported credit need be confused with debt capacity and reduce the lender's risk. However, in microfinance it is effective demand. Obviously, since there are costs THE TARGET MARKET AND IMPACT ANALYSIS 37 involved in starting the delivery of financial services in a a narrowly defined group of clients to establish a market given area, it is important to know: niche and develop a thorough understanding of this I What kind of financial service will both benefit client base. households or enterprises and have a high likelihood The target market for MFIs generally takes into con- of good repayment to the lender? sideration a combination of two factors: I How much effective demand for the product(s) will I Characteristics of the population group, including the there be initially, and how expandable will that mar- level of poverty ket be? I The type of microenterprises being financed. These questions can be answered by undertaking lim- ited survey work, preferably in-depth interviews with Characteristics of the Population Group selected potential clients supplemented by some quanti- tative work, or, alternatively, by undertaking a full con- In many countries, people operating in the informal straints analysis and a large-scale survey to estimate the sector are illiterate. An MFI needs to understand the full range of possibilities. Whichever method is selected, level of literacy (including financial literacy) of its client it is important to estimate the market size to ensure that, base to design appropriate interventions. Some MFIs in the long term, enough demand exists to justify the require illiterate borrowers to use their thumbprint or MFI's continued existence. fingerprint as a means of formally agreeing to a finan- cial contract. Others have spent time teaching the bor- rowers how to sign their names and read numbers to be Identifying the Target Market able to verify the contracts they sign. Literacy and numeracy are not necessary prerequisites for accessing Profit-oriented companies either invest in identifying financial services from many MFIs; however, they must specific segments of the market for a product they want be taken into account when designing credit and sav- to sell or design products specifically with a market seg- ings transactions. ment in mind. For donors or development organizations Based on their objectives, microfinance providers may that want to achieve development goals (rather than want to select a target market to address a specific client profit), market identification serves a somewhat different population. Characteristics of the population group take purpose. In this case, the target market is identified into account various socioeconomic characteristics, because it is underserved and disadvantaged in some way including gender, poverty level, geographic focus, and that slows development. The goal is not simply profit, ethnicity, caste, and religion (figure 2.1). but rather a more equitable distribution of financial ser- vices to groups that can make productive use of them FOCUSING ON FEMALE CLIENTS. The objective of many (Bennett 1997). MFIs is to empower women by increasing their econom- Nevertheless, even though the goals may differ, devel- ic position in society. The provision of financial services opment organizations still need to use the approach of directly to women aids in this process. Women entrepre- profit-oriented companies to identify what the chosen neurs have attracted special interest from MFIs because clients want and what they can afford to pay, so that they almost always make up the poorest segments of soci- appropriate products and services can be offered. The ety; they are generally responsible for child-rearing question is not whether to identify a target market, but (including education, health, and nutrition); and they how to identify it. often have fewer economic opportunities than men. Understanding the characteristics of the target market Women face cultural barriers that often restrict them helps MFIs design products and services that attract dif- to the home (for example, Islamic purdah), making it ferent groups (as opposed to targeting them directly). difficult for them to access financial services. Women This becomes an iterative process as the MFI learns more also have more traditional roles in the economy and may about the target market and its needs. be less able to operate a business outside of their homes. While it is usual to broaden the client base over Furthermore, women often have disproportionally large time, most successful microfinance projects begin with household obligations. 38 MICROFINANCE HANDBOOK monitoring repayment data reported a higher rate of repay- Figure 2.1 Client Characteristics ment of loans in projects focused on women than in com- panion nontargeted projects (Rhyne and Holt 1994). The characteristics of women's businesses differ from those of men's in important ways. In general, women Men Women Poor Ultra poor tend to weigh household maintenance and risk reduction more heavily in their business strategies. Women also tend to give less emphasis to enterprise growth, prefer- ring to invest profits in their families rather than in expanding the enterprise. Other characteristics include: I A concentration in trade, services, and light manufac- turing (particularly in subsectors using traditional technologies) Religion Caste I A tendency to start smaller and stay smaller through- Urban Rural out their lifetimes, although women's businesses last Ethnicity as long (if not longer) than those of men I The frequent use of family labor and the location of their businesses in the household. Source: Contributed by Mike Goldberg, CGAP. In both urban and rural settings, women tend to engage in activities that offer easy entry and exit and do In some instances, commercial banks are unwilling to lend to women or mobilize deposits from them. This is Box 2.2 U.S. Agency for International Development based on their perception that women are unable to con- Findings on Female Borrowers trol household income. Moreover, because women's access to property is limited and their legal standing can IN A U.S. AGENCY FOR INTERNATIONAL DEVELOPMENT be precarious, women also have fewer sources of collater- (USAID) study of 11 successful MFIS, findings indicate al. Finally, in many countries women have lower literacy that the organizations studied do in fact reach large num- rates, which makes it more difficult for them to deal with bers of women, either because of direct policy decisions financial systems that depend on written contracts. (Grameen Bank, Association Dominicana para el Desarrollo de la Mujer, or ADOPEM) or because of a Experience has shown that women generally have a high commonly held belief that women demonstrate stronger sense of responsibility and are affected by social pressure repayment performance and are more willing to form (although, like men, they often fail to repay subsidized groups (the Kenya Rural Enterprise Programme). loans from government or other programs that they per- Among programs concentrating on women, motiva- ceive as charity rather than business). It has been argued tions generally include the belief or experience that that an increase in women's income benefits the household women are good credit risks and are more likely to have and the community to a greater extent than a commensu- poor access to resources and services. Female participation rate increase in men's income. Women have also demon- rates in programs without gender preference are deter- strated higher repayment and savings rates than male clients mined by the prevalence of women in client groups served (box 2.2). Anecdotal evidence suggests that arrears rates are and by features not studied that may impede or facilitate slightly higher among men; however, since many generally women's access. Some correlation exists between pro- accepted ideas about women and microfinance are based grams offering smaller loans and programs serving more women, but the correlation is far from perfect. For exam- on the Grameen Bank's experience, which lacks a control ple, BancoSol has a relatively high average outstanding group of men, it is difficult to generalize conclusively about loan size, yet 71 percent of its clients are women. the repayment behavior of males and females. However, a review of World Bank projects supporting enterprise devel- Source: Christian and others, 1995. opment for women found that the majority of projects THE TARGET MARKET AND IMPACT ANALYSIS 39 not require large amounts of working capital, fixed assets, est of the poor." While there are now many examples of or special skills (beyond those already acquired in the programs and institutions serving the working poor with household). Such activities offer the flexibility regarding financial services in a self-sustainable manner, there is less time commitments that enables women to balance their experience of successfully serving the very poor, the desti- work and family obligations. The activities are often sea- tute, and the disabled. Many donors, practitioners, and sonal, geographically portable, and fit household condi- academics believe they ought to be beneficiaries of trans- tions and space limitations. The market is usually limited fer programs that do not entail an additional liability for to local consumers. These characteristics imply that the recipient (Hulme and Mosley 1996). women's household duties limit their choice of business One of the most important studies on whether micro- activity, which is an important consideration when pro- finance is appropriate for the poorest of the poor is that viding financial services specifically for women. by Hulme and Mosley in their book Finance against MFIs need to be proactive in identifying female clients. Poverty. Using data from MFIs in seven countries, they They need to look beyond areas with high concentrations compare increases in borrowers' income with those in a of manufacturing enterprises to promoting services through control group. Their findings suggest that successful insti- existing women's networks and by word of mouth. The tutions contributing to poverty reduction are particularly gender of loan officers may also affect the level of female effective in improving the status of the middle and upper participation, depending on the social context. In such segments of the poor. However, clients below the poverty cases, if female credit officers are not available (because they line were worse off after borrowing than before. lack educational opportunities or are unable to walk on the Furthermore, the impact on the client's income seems to streets alone or at night), it may be more acceptable for be directly related to the level of income, which would male credit officers to work with women in groups. reinforce the tendency of those MFIs wishing to preserve A study by the World Bank's Sustainable Banking their viability to concentrate on the less-poor. A related with the Poor project titled "Worldwide Inventory of consequence is that the development and refinements of Microfinance Institutions" (Paxton 1996) found that microfinance products remain focused on the middle and predominately female programs are typically group- upper segments of the poor, leaving the poorest behind. based, with small loan sizes and short loan terms. "The Hulme and Mosley suggest that recognizing the het- median loan size for female-based programs is US$162, erogeneity of the poor should lead to more innovation compared to US$768 for the predominately male pro- and experimentation, which deepens the downward grams" (Paxton 1996, 13). However, this finding does reach of financial services. However, there is still a large not definitively support smaller loan sizes for women, unsatisfied demand for financial services among the because it is based on the identification of MFIs simply bankable poor (those with the capacity for debt), which as having more than 50 percent women (rather than an could be met by mainstreaming the known successful actual percentage of women clients). approaches. This debate will likely continue for some time. The "depth" of outreach achieved by an MFI will THE LEVEL OF POVERTY. Because the objective of many depend to a great extent on its objectives and its ability MFIs is poverty reduction, they often wish to focus on to design products and services suitable to the level of the poorest segments of the population. In most coun- poverty it is targeting. tries many people do not have access to financial services, In light of the need for most MFIs to reach financial from the poorest of the poor, who may not be economi- sustainability, consideration must be given to the trade- cally active, to small business operators, who may not off between minimizing costs and focusing on the poor- qualify for formal financial sector services. MFIs com- est clients. While serving the ultra-poor may indeed be monly measure their outreach of services in terms of scale, possible in a financially sustainable way, it is likely that or the number of clients they reach, and depth, or the the time frame to reach financial self-sufficiency will be level of poverty of their clients. An MFI's products and shorter for MFIs serving the economically active poor. If services will vary with the extent of outreach. the target market identified is the poorest of the poor, There is much debate in the field of microfinance as donors and practitioners alike need to be committed to to whether access to financial services benefit the "poor- supporting the institution over a longer period. 40 MICROFINANCE HANDBOOK GEOGRAPHIC FOCUS. One of the most important consid- mobilization, or credit tied to specific activities or erations for an MFI is whether it will serve urban or rural purchases). clients. This decision greatly affects the development of products and services, and it should be based on both the activities characteristic of different geographical settings Box 2.3 The Kenya Rural Enterprise Programme and the varying levels of infrastructure development in urban and rural areas. THE KENYA RURAL ENTERPRISE PROGRAMME (K-REP) WAS Choosing a target market that is based in urban areas established in 1984 as a development NGO that provides has both advantages and disadvantages. It is highly credit for on-lending and technical assistance to other NGOs. To promote growth and generate employment in dependent on the objectives of the MFI. The advantages the microenterprise sector, K-REP lends indirectly through of focusing on an urban market may include: NGOs and directly to groups that would otherwise find it I Lower transaction costs (shorter distances) for clients extremely difficult to access credit from commercial banks I Greater chance that clients will be literate and other formal financial intermediaries. I Potential higher chance of repayment, since interac- K-REP offers credit directly to groups through its tions with clients can be more frequent Juhudi and Chikola loan products. Juhudi, initiated in I Possible leveraging through relationships with formal 1989, provides loans based on a modification of the financial institutions, since urban clients may be group lending methodology used by the Grameen Bank physically closer to formal sector banks and more in Bangladesh. K-REP facilitates the formation of five to comfortable with visiting banks seven member groups called wantanos. Up to six wan- I More developed local infrastructure and more varied tanos confederate into a kiwa, which is registered by the Kenyan Ministry of Culture and Social Services as a self- markets. help group. The Chikola program, initiated in 1991, However, urban clients may be more transitory, provides credit to individual entrepreneurs through exist- resulting in a higher risk of potential default. Character- ing rotating savings and credit associations. Under the based lending may be more difficult. In addition, covari- Chikola program, K-REP provides a single loan to an ance risk can exist if most clients are active in the same established group, which then retails the loan to its indi- economic sector; in other words, if they are all traders in vidual members. To be eligible the group must be regis- the same area or manufacturers of the same products. tered as a self-help group, be in existence for at least one When the loan portfolio of any financial institution is year, and have an average membership of 20. The heavily concentrated in a few activities, risks can increase Chikola must be operating or must intend to operate a substantially. The same thing happens in all markets if revolving loan fund and must have a group savings the region has one principal economic sector that enters account. The qualifying level of savings is not less than into decline. Financing larger clients in these environ- 10 percent of the requested loan amount. A K-REP credit officer appraises the loan application ments does not effectively diversify risks. and makes the disbursement by bank check to the Some MFIs provide financial services in rural areas group. Each group must meet at least once a month to only, based on a general lack of supply of services outside conduct group activities, including collection of savings urban centers and the fact that in some countries poverty and loan repayments. Scheduled group loan repayments is largely a rural phenomenon. Providing services to rural are made (due to regulation) by transfers from the group clients can therefore be an effective means of reaching a savings account to K-REP's bank account. The loan large number of poor households. Furthermore, there are limit is 25,000 Kenyan shillings (about US$450) per often local informal organizations that can be used to member for the first loan, with subsequent loans varying deliver financial services. For example, the Kenyan chikola in size according to the needs of the group and its mem- system, a form of village rotating savings and credit sys- bers. The savings of the group serve as collateral for the tem, has become the mainstay of the Kenya Rural loan, and each member agrees to forfeit his or her sav- ings in the event of default by a group member. Enterprise Programme (K-REP; box 2.3). However, rural markets may also have disadvantages: Source: Contributed by Stephanie Charitonenko Church, I There may be a long history of poorly designed rural Sustainable Banking with the Poor Project, World Bank. credit programs (with subsidized credit, no savings THE TARGET MARKET AND IMPACT ANALYSIS 41 I There may be a less diversified economic base. viding microfinance services more difficult. (Caste, while I Covariance risk can be significant. For example, if the not solely an ethnic factor, can pose cultural barriers simi- MFI is engaged in agricultural lending, many farmers lar to ethnic differences.) Societies differ in their stock of may grow the same crop or raise the same type of ani- "social capital--those features of social organization such mals, resulting in higher risk in case of drought or as networks, norms, and trust that facilitate coordination other climate disorder. and cooperation for mutual benefit" (Putnam 1993, 36). I There may be no branches of formal financial institu- These structures depend on traditions of collaboration tions in the area. This can create problems when and a certain level of trust between members of society. In clients need access to a branch network for savings societies with high social capital, where systems and struc- deposits or loan repayment. tures have been developed to build trust and foster social I It may be more difficult to reach the minimal scale and economic transactions beyond the family and required to break even. kingroup, it will be easier and less costly to build sustain- I There is likely to be a poorly developed infrastructure able systems for financial intermediation (Bennett 1997). and a more dispersed population. When identifying a target market, MFIs must ensure Whether microentrepreneurs are based in a rural or that they are able to communicate with their clients clear- urban area, they need access to markets and supplies. ly. If serious language barriers exist, it may make financial Rural areas are often isolated from markets. Furthermore, services to certain clients too costly. In addition, legal or an inability to produce and deliver goods due to a lack of financial jargon, even when in the client's language, may infrastructure may limit microentrepreneurs' success and create a communication barrier. MFIs must ensure that growth possibilities, thus limiting the demand for finan- communication with clients is appropriate to their level of cial services. A good transportation system reduces trans- understanding, particularly if clients are anxious to access action costs for both clients and MFIs. Some MFIs, such financial services and do not fully understand the implica- as the Grameen Bank, operate branches in the same geo- tions and responsibilities of doing so (box 2.4). graphic areas as their clients, reducing the barrier to accessible financial services due to poor roads and inade- quate transportation services. Box 2.4 The Influence of Ethnicity and Language in MFIs depend on attaining a certain scale of operations Microfinance to achieve sustainability. Providing services to populations that are widely dispersed incurs greater transaction costs. D OGON VILLAGES IN M ALI HAVE REMAINED VERY However, while less densely populated areas are harder to secluded over the years because of their physical isolation service, new methods, such as self-managed village banks, and the poor condition of communication networks. As a are being developed to overcome this problem by decen- result, the Dogon culture is relatively homogeneous and traditions have remained very strong. In addition, villages tralizing most of the financial transactions at the village have maintained their own linguistic identity. Dialects level, thus limiting transportation costs. differ from one village to another, even though they are only a few miles away, making communication difficult. ETHNICITY, CASTE, AND RELIGION. In most countries there The total Dogon population amounts to only 600,000 are ethnically or traditionally defined groups that are people speaking at least 30 different dialects. unserved or underserved by existing formal financial For a local MFI, the multitude of dialects has three institutions. There are cases in which a certain group important effects. First, all staff have to be hired locally, within a community cannot or will not take part in a taking into consideration their ability to master several financial services project due to a religious, ethnic, or dialects. Second, training activities and documents for vil- other social influence. It is important to understand these lagers on bank management have to be adapted into sev- restrictions when identifying a target market, so that eral dialects. Third, networking among village banks products and services can be developed that take into involves many translations. account the limitations on some groups. Source: Contributed by Cecile Fruman, Sustainable Banking with Building and maintaining a level of trust when differ- the Poor Project, World Bank. ent ethnic or religious groups are involved can make pro- 42 MICROFINANCE HANDBOOK Some religious groups do not allow their members tant to consider the types of activities in which the target to enter into financial transactions. Both donors and market is active and the level of development of the enter- practitioners should be aware of religious practices that prise being financed. This will further define the types of might affect the delivery of financial services and conse- products and services suitable for the MFI's market. quently the objectives and goals of the MFI. Islamic Enterprises vary by whether they are existing or start-up banking provides an example of financial markets businesses; unstable, stable, or growing; and involved in directly affected by religious practices (box 2.5). production, commercial, or service activities (figure 2.2). Types of Microenterprises EXISTING OR START-UP MICROENTERPRISES. When identi- fying a target market, an MFI needs to consider whether In addition to determining the characteristics of the pop- it will focus on entrepreneurs already operating a ulation group to be served by the MFI, it is also impor- microenterprise or on entrepreneurs (or potential entre- preneurs) who need financial services to start a business and possibly some form of business training. Working capital is the most common constraint identi- Box 2.5 Islamic Banking fied by entrepreneurs of existing microenterprises. To access F INANCIAL PROVIDERS IN I SLAMIC COUNTRIES ARE working capital, microentrepreneurs often borrow from subject to specific banking laws. The primary difference informal financial sources, such as family or friends, sup- between Islamic and conventional banking products is pliers, or a local moneylender. Usually moneylenders that in Islamic banking, money cannot be treated as a com- charge relatively high interest rates and may not offer loan modity. Money must be used productively (like labor or products or terms suited to the borrower. The ability to land), with the return based solely on actual profit or loss. both borrow and save with an MFI may increase microen- The bank and the entrepreneur share the risk of a project. trepreneurs' profits (through lower interest rates and access Islamic banks can neither charge nor pay interest (riba), to appropriately designed loan products) and improve nor can they require specific provisions for inflation. Islamic banks tend to make their profits through fees and their ability to manage working capital needs (through non-interest-bearing investments (that is, through project borrowing and saving at different times as required). profits). It should be noted, however, that due to the competitive nature of the banking field in general, returns Figure 2.2 Types of Microenterprises on Islamic banking products are close to those of similar products developed by conventional banks. Islamic banks are governed by sharia (Islamic law and councils of advisers), which dictate the products and ser- Start-up Unstable Growing vices that are Islamically correct. Sharia and their interpre- business tations of the religious commentary vary from country to Existing country, making it difficult to develop standard products. business This complicates efforts to establish an adequate regulato- Stable ry environment. In some ways Islamic banking is well suited to microfi- nance. In many cases collateral is not required. Instead, emphasis is placed on the profitability of the project. However, because the nature of Islamic banking often Agriculture Production requires the bank to share project risks with the client, a sector more detailed study of the project is required than a con- ventional bank would carry out. Consequently, some Stable Islamic banks maintain staff with technical backgrounds in addition to more conventional financial staff. Source: Khayat 1996. Source: Contributed by Mike Goldberg, CGAP. THE TARGET MARKET AND IMPACT ANALYSIS 43 Profits can also be increased through the acquisition of capital assets, such as sewing machines or rickshaws. Box 2.6 The Association for the Development of Access to continued financial services, including loans Microenterprises' Work with Existing Microenterprises for capital purchases and savings services to build up THE ASSOCIATION FOR THE DEVELOPMENT OF reserves, allows microentrepreneurs to increase their asset Microenterprises (ADEMI) has developed an effective base and improve their ability to generate revenue. lending methodology to meet the needs of low-income There are many advantages to working with existing urban clients in the Dominican Republic. As of June microentrepreneurs. Active businesses have a history of 1997 it had a network of 25 branches serving over success, which greatly reduces the risk to the MFI. 16,300 clients. Its assets totaled RD606 million (US$43 Furthermore, existing microentrepreneurs have the million) and its equity was RD242 million (US$17 mil- potential to grow and create employment opportunities. lion). ADEMI has served more than 40,000 small and However, some may have other debts (to moneylenders, micro enterprises during the past 15 years and had a suppliers, family, or other MFIs), which they may pay off direct impact on the jobs of more than 200,000 owners with the proceeds of new loans, thereby increasing the and employees of micro and small-scale enterprises. risk of default to the new lender (box 2.6). Objectives. Job creation and delivery of sustainable MFIs that target potential entrepreneurs often have financial services. Target clients. Established microenterprises with at poverty alleviation as an objective. The belief is that by least one year of experience and the potential to create aiding potential entrepreneurs to start up their own busi- new jobs. Clients must be over 18 years old, show initia- nesses, they will increase their incomes and consequently tive, and have a minimum level of business aptitude to reduce their level of poverty. ensure their success and growth. However, potential entrepreneurs often need more Services. Loans for working capital or the acquisition than financial services. Many need skills training or of equipment and machinery (fixed assets); and ADEMI other inputs to make their enterprises a success. When Mastercard; home improvement loans; and minimal sav- there are significant barriers to entry into certain fields ings services. ADEMI does not make consumption (due to minimum investment requirements, technolo- loans. gy levels, and market contacts), an integrated approach Method. Individual loans; no formal collateral, can prepare potential entrepreneurs prior to taking on although ADEMI does encourage the use of guarantors and will repossess equipment or seize other assets if a debt. However, the impact of training courses and client defaults. All borrowers must sign a legal contract technical assistance is not clearly linked to increased and promissory note, stating their obligation to repay production, profitability, job creation, and reinvest- funds at specified terms. Loan terms are on average ment. If services are subsidized, it can be difficult to between 10 and 34 months; repayment is monthly; aver- remove these subsidies and put the enterprise on an age loan sizes vary from US$1,000 to US$10,000 for equal footing with local competitors. Also, training microenterprises; interest rates range from 2.2 percent to programs often assume that anyone can become an 3 percent a month. entrepreneur, which is not the case, because not every- Source: Benjamin and Ledgerwood 1998. one is willing to take the risks inherent in owning and operating a business (although social intermediation services can help--see chapter 3). Furthermore, train- up businesses, on the one hand, and sustainability, on ing courses linked to credit access sometimes assume the other. that entrepreneurs cannot contribute their own equity and that credit should be arranged for 100 percent of LEVEL OF BUSINESS DEVELOPMENT. The level of business the investment. development is another consideration when identifying the Most MFIs prefer to focus on existing businesses, types of microenterprise to which an MFI wishes to pro- with perhaps a small portion of their portfolio invested vide financial services. This is closely linked with the level in start-up businesses, thereby reducing their risk. This of poverty existing in a potential target market. There are is again dependent on their objectives and the trade-off typically three levels of business development of microen- between increased costs (and lower loan sizes) for start- terprises that benefit from access to financial services: 44 MICROFINANCE HANDBOOK I Unstable survivors, with operators who have not found Generally, profits are low, leading to low reinvestment, other employment and tend to have very unstable low output, and a high level of vulnerability. Profits enterprises for a limited time remain low due to: I Stable survivors, with operators for whom the I The unspecialized nature of the product microenterprise provides a modest but decent living I The lack of timely and complete market information while rarely growing (beyond the local market) I Growth enterprises, or businesses that have the poten- I Underdeveloped infrastructure facilities tial to grow and become genuinely dynamic small I The lack of value-added services (such as packaging) enterprises. I The number of producers with similar products. Unstable survivors comprise the group most diffi- Experiences with this target group have demonstrat- cult to provide financial services to in a sustainable ed both advantages and disadvantages. Advantages may fashion, because loan sizes tend to remain small and include: the risk of business failure is high. Focusing on unsta- I The high poverty impact of a financial services pro- ble survivors as a target market can result in a great ject, since these enterprises are run by poor house- deal of time spent with the clients just to ensure that holds their businesses will survive and that they will contin- I High repayment, due to limited access to alternative ue to be able to make loan payments. Some technical sources of credit and the economic, social, and assistance may also be required, resulting in further financial costs of those alternatives time and cost increases. Also, unstable survivors often I Effective savings services, since there are rarely need credit for consumption-smoothing rather than secure, liquid alternative forms of savings that offer a income-generating activities. Depending on the objec- return for the operators of these enterprises (they tives of the MFI, these stopgap loans may or may not also help to smooth consumption for poor house- be appropriate. holds) Generally, the debt capacity of unstable survivors does I A general willingness to work with new credit tech- not increase. Accordingly, the MFI is limited in its nologies (such as groups) as an alternative to tangible attempts to reduce costs or increase revenue, because collateral. loan sizes remain small. While not all MFIs have the Disadvantages may include: immediate goal of reaching financial self-sufficiency, over I Little or no new job creation resulting from support the long term the choice to focus on unstable survivors to these enterprises will likely be a time-bound strategy, because access to I Limited growth potential or high covariance risk, donor funding may be limited. because many entrepreneurs are active in the same Stable survivors comprise the group that many MFIs businesses (financing them may create excess compe- focus on and for which access to a permanent credit sup- tition, meaning that the loan portfolio has to grow ply is vital. This is the group that benefits from access to by increasing the number of clients rather than financial services to meet both production and consump- increasing loan amounts to good clients) tion needs, while not necessarily requiring other inputs I Difficulty in mobilizing long-term savings, since from the MFI. households are accustomed to seasonal savings build- Stable survivors are targeted by microfinance providers up and liquidation cycles. with poverty reduction objectives. For these businesses, Growth enterprises are often the focus of MFIs returns on labor are relatively low, and market imperfec- whose objective is job creation and whose desire is to tions and near monopsony conditions may result in move microentrepreneurs from the informal sector to a uneven bargaining positions. Stable survivors are often progressively more formal environment. These MFIs women who simultaneously maintain family-related often establish linkages with the formal sector and pro- activities (providing food, water, cooking, medicine, and vide additional products and services. Growth enter- child care) while engaging in income-generating activities. prises represent the upper end of the poverty scale: Seasonal changes and household life cycles often force they usually pose the least risk to the MFI. While gen- such people to consume rather than invest in the business. erally a heterogeneous collection of enterprises, they THE TARGET MARKET AND IMPACT ANALYSIS 45 tend to share some characteristics and face similar tifying a target market, the economic sector of activities problems. Most have both production and risk-taking is also important. Enterprises can generally be divided experience, keep minimal accounting records, and usu- among three primary sectors: production, services, and ally do not pay taxes. In addition, they often have little agriculture. Each sector has its own specific risks and or no formal management experience. Other similari- financing needs, which directly influence the choices ties include: made by the MFI and the products and services provid- I Product line and labor. Firms that produce a single prod- ed. Table 2.1 identifies the loan purpose, the loan term, uct or line of products serving a narrow range of market and the collateral available for each of the three primary outlets and clients tend to use labor-intensive produc- sectors. tion techniques and rely on family and apprentice labor. There are several advantages to focusing on one eco- I Working capital and fixed asset management. These nomic sector: firms build their asset base slowly, in an ad hoc man- I Credit officers can focus their learning on one sector, ner. They depend largely on family credit for initial thereby developing an understanding of the charac- investment capital and on informal sector loans for teristics and issues that their borrowers face. working capital. Cash flow is a constant concern, and Consequently, they may be able to provide technical they are very sensitive to output and raw material price assistance more easily if it is required. changes. They often use second-hand equipment. I One loan product may be sufficient for all, thereby Growth-oriented microenterprises may be an attrac- streamlining operations and reducing transaction tive target group, because they offer potential for job costs. creation and vocational training within the community. However, MFIs are subject to covariance risk when They can resemble formal sector enterprises in terms of selecting a target market active in only one economic fixed assets, permanence, and planning, which offers sector. the potential for physical collateral and more thorough Not all MFIs target a single economic sector. Many business analysis. All these offset risk for the MFI. provide financial services for a combination of sectors, However, selecting growth-oriented microenterprises designing loan or savings products or both for each. can require a more involved approach on the part of the However, it is generally recommended that MFIs focus MFI. Growth-oriented businesses may need some or all on one sector until they have developed a sustainable of the following services: approach before developing new products (bearing in I Assistance in choosing new product lines and value- mind the higher covariance risk). added services Depending on the MFI's objectives, selecting a target I Working capital and sometimes longer-term invest- market based on the business sector allows it to clearly ment credit differentiate its products and influence the type of activi- I Accounting systems to track costs ties in a given area. This is evidenced by the Foundation I Marketing advice to help find new markets. for Enterprise Finance and Development in Albania (box 2.7). TYPE OF BUSINESS ACTIVITY. While the level of business Once a target market has been identified, the MFI development is an important consideration when iden- needs to design its products and services to meet the Table 2.1 Enterprise Sector Credit Characteristics Criterion Agriculture Production Services Use of loan proceeds Working capital, Working capital, fixed Working capital, some fixed assets assets, infrastructure some fixed assets Loan term Growing season 6 months­5 years 4 months­2 years Amount of collateral available Minimal Good Minimal Source: Waterfield and Duval 1996. 46 MICROFINANCE HANDBOOK the desired outcome. If the intervention is, for example, Box 2.7 Foundation for Enterprise Finance and an immunization program and the desired outcome is to Development's Approach to Sector Development prevent polio, the impact analysis would focus on polio rates. If it could be shown that polio rates went down as T HE F OUNDATION FOR E NTERPRISE F INANCE AND a result of the immunizations, the program's impact Development (FEFAD) began operations in Tirana, could be deemed successful. In other words, the impact Albania, in January 1996. FEFAD provides loans to indi- analyzed should correspond to the desired outcome. viduals in urban areas for small and medium-size busi- Microfinance impact analysis is the process by which nesses. At the end of 1996 it had over 300 outstanding one determines the effects of microfinance as an inter- loans for a total outstanding balance of 165 million lek vention. The effects examined depend on the outcomes (US$1.5 million). FEFAD focuses on three business sectors: production that were sought (the objectives of the MFI). Generally, (40 percent), services (20 percent), and trade (40 percent). the narrower the goals of the intervention, the less prob- Construction projects and agricultural activities are not lematic the impact analysis. financed. Annual interest rates in February 1997 were 26 Decisions about the degree, frequency, and depth of percent for production loans, 30 percent for services impact analysis involve consideration of the following loans, and 32 percent for loans for trade. The differential factors: in interest rates is based on the desire of FEFAD to I The time and cost encourage production activities over those in the other I The disruption to the institution and its clients two sectors. I The way the results will be used (the fear of bad news). Source: Ledgerwood 1997. All interventions can have unintended consequences. It is optional to seek and investigate unintended impacts, but the analysis is more complete to the extent that these needs of that market (see chapter 3). However, before unintended consequences can be illuminated. doing this, it is important to consider the type of impact Some likely users of microfinance impact analysis are: that the MFI hopes to achieve. This will help develop I Microfinance practitioners products and services and should be considered part of I Donors the design process. This is particularly relevant if the I Policymakers MFI wishes to maintain a database on the target market I Academics. to see the impact of financial services over time. Both practitioners and donors are usually concerned about improving their institutions or those that they sup- port, as well as learning if their interventions are having Impact Analysis the desired impact. Thus they may also be interested in impact analysis as a form of market research through Analyzing the impact of microfinance interventions is which they can learn more about their clients' needs and especially important if the interventions are ultimately how to improve their services. Impact assessment can aimed at poverty reduction (as most are).2 If microfi- also contribute to budget allocation decisions. nance practitioners do not make efforts to determine Policymakers and academics are solely concerned who is being reached by microfinance services and how with attributing impact effects to microfinance inter- these services are affecting their lives, it becomes difficult ventions. They can use impact assessment information to justify microfinance as a tool for poverty reduction. to influence policy changes and budget allocation deci- In the most generic sense, impact analysis is any sions and to address questions suited to academic process that seeks to determine if an intervention has had research. 2. This section was written by Thomas Dichter. It assumes a basic knowledge of statistics, sampling techniques, and the broad spec- trum of field methods of assessment. If there are terms with which readers are unfamiliar, they should read some of the sources listed at the end of this chapter to understand better the skills and resources needed to conduct microfinance impact analysis. THE TARGET MARKET AND IMPACT ANALYSIS 47 Impacts to be analyzed should correlate with impacts that are intended. Most MFIs see microfinance as a cost- Box 2.8 The Impact of Finance on the Rural Economy effective means of poverty reduction or poverty alleviation, of India but the detailed intentions and expectations of microfi- A RECENT STUDY OF RURAL FINANCE IN INDIA ANALYZES nance programs can differ considerably. A good starting data for 1972 to 1981 for 85 rural districts. The study point for impact analysts is the MFI's mission or goal. found that the Indian government had pursued a poli- cy of rapid, forced expansion of commercial banks into Kinds of Impacts rural areas. This expansion resulted in a rapid increase of nonagricultural rural employment and a modest Broadly, impacts of microfinance activities fall into three increase in the rural wage rate. By inference the study categories: concludes that the policy must have significantly I Economic increased the growth of the nonagricultural rural sec- I Sociopolitical or cultural tor. These findings suggest that the expansion of com- I Personal or psychological. mercial banks into rural areas eliminated severe Within each of these categories there are different constraints in rural financial markets and led to signifi- cant deepening of rural finance. levels of effect and different targets. The Indian government also pursued a policy of directing commercial banks and the cooperative credit ECONOMIC IMPACTS. Economic impacts can be at the level institutions to lend specifically for agriculture. The study of the economy itself. A large MFI reaching hundreds of found that the expansion of rural and agricultural credit thousands of clients may expect or aim at impact in terms volumes had a small positive impact on aggregate crop of changes in economic growth in a region or sector. output. This small increase is accounted for by a large I One MFI may seek outcomes at the level of the enter- increase in fertilizer use and by increased investments in prise. If so, it will look for business expansion or trans- animals and irrigation pumpsets. At the same time, agri- formation of the enterprise as the primary impact. cultural employment declined. This implies that the poli- I Another may seek net gains in the income within a cy of forced lending to agriculture had more impact on subsector of the informal economy (for example, the the substitution of capital for labor than on agricultural output, a clearly counterproductive outcome when the operators and owners of tricycle rickshaws). abundance of labor in India is considered. I Another may seek impact in terms of aggregate accu- The study compares the modest increase in agricul- mulation of wealth at the level of the community or tural output with the costs to the government of the household. directed and modestly subsidized credit to agriculture. I Another may seek positive impacts in terms of income The government costs include rough estimates of the or economic resource "protection" (reducing the vul- transaction costs, interest subsidy, and loan losses in nerability of poor people through what has come to the system. Assuming a default rate of 10 percent, the be called consumption smoothing). findings suggest that the value of the extra agricultural output triggered by the targeted credit almost covered SOCIOPOLITICAL OR CULTURAL IMPACTS. An MFI may seek a the government's cost of providing it. Of course, farm- shift in the political-economic status of a particular subsector. ers have additional costs in making their investments I A project aimed at credit for tricycle rickshaw drivers that are not included in the government's costs. These findings indicate that deepening the system of may hope that the drivers' increased business will rural financial intermediation in India had high payoffs in enable them to move collectively to formal status, rural growth, employment, and welfare, but that specifical- either by forming an association or by being able to ly targeting credit to agriculture was of doubtful benefit. change policy in their favor. I An MFI in a remote rural area may expect to help shift Source: Binswanger and Khandker 1995. rural people from a barter to a monetarized economy. I Another may hope for changes in power (and status) relationships. For example, an MFI that targets a changing the balance of power between that group minority ethnic group may seek impact in terms of and the local majority group. 48 MICROFINANCE HANDBOOK I Another may seek, as a primary impact, the redistrib- begun to change. New empirical evidence makes it pos- ution of assets (and power or decisionmaking) at the sible to redirect some of the impact questions and with household level (for example, shifting part of econom- them microfinance itself. ic decisionmaking from men to women). Johnson and Rogaly (1997), for example, provide a I Another may seek changes in children's nutrition or useful distinction between three sources of poverty: education as the result of a microfinance activity I Lack of income aimed at their mothers. I Vulnerability to income fluctuations I Powerlessness (few choices and little control). PERSONAL OR PSYCHOLOGICAL IMPACTS. Microfinance can Interventions can therefore promote income, pro- have impacts on the borrower's sense of self. tect income, or empower people. They point out that These impacts are the other half of empowerment most MFIs, especially those run by NGOs, measure effects. The first half is in a sense political--people (or more often estimate) only the first of these. Thus achieve more power in the household or community as they look only at income changes as the result of cred- the result of financial services. The second half is internal it (Johnson and Rogaly 1997) This is gradually and has to do with the person's changed view of self. changing. Such a change, if positive, can prepare the way for other changes. For example, a person who feels more confident A T THE HOUSEHOLD LEVEL . While many MFIs still may be willing to take new kinds of risks, such as starting direct credit to preexisting informal sector microenter- or expanding a business. prises, many others now direct credit specifically to Of course, all three general impact categories (as well women who are engaged in what could be called an as the different levels and targets) can shift in terms of income-generating activity rather than a business. which are primary and which are secondary effects. These MFIs are less concerned about whether the cred- Finally, an impact from one of these categories can in it is used for "business purposes." A few MFIs also itself cause an impact in one or more of the others. explicitly lend for consumption and seek impacts at the household level. What Kinds of Impacts Have We Seen with Making households and the household economy the Microfinance? targets of microfinance is increasing as research shows how important it is to reduce the economic insecurity of In the short history of microfinance impact analysis, we poor people, not by raising their income, but by "pro- have seen relatively little impact on a macro level. tecting" what little they do have and reducing their vul- However, few MFIs aim for impacts at the level of the nerability (Corbett 1988). economy as a whole. A recent review of research on impact discusses the With regard to impacts at the level of microenterpris- concept of the "household economic portfolio" as the es in the informal economy, the body of evidence sug- sum total of human physical and financial resources, gests that microenterprise credit does not result in which is in a dynamic relationship to the sum total of significant net gains in employment, but it can and does household consumption, production, and investment lead to increased use of family labor. We can state that activities (Dunn 1996). This very realistic approach to there is as yet no solid evidence of business growth and the way in which credit is used by many poor borrowers transformation as the result of microenterprise credit, but helps us to understand that the categories of "borrowing there is evidence of credit enabling enterprises to survive for production" and "borrowing for consumption" are (remain in business) in crises. not always useful in practice. The microfinance field has not yet amassed a large body of well-researched impact analyses. In recent years AT THE INDIVIDUAL LEVEL. Similarly, as the household-- there have been reviews of existing impact studies, and seen as a tiny economy in itself--became a guiding con- these as well as other research into the nature of poverty cept in impact analysis, formal economic models of at the grassroots have begun to interact with the disci- household decisionmaking have entered into economic pline of impact analysis. As a result, the questions have thinking. For example, "the role of individual preferences, THE TARGET MARKET AND IMPACT ANALYSIS 49 resources and bargaining power in intra-household deci- High repayment rates and low arrears can be taken as sion making" is now recognized as important (Chen and prima facie evidence of willingness to pay. (This can also Dunn 1996). Such concepts are coming into use as ways be applied to savings services; if people continue to uti- to map changes at both the household level (as a social and lize the savings services on an MFI, it can be assumed economic unit) and among the persons living in the that the service is valued.) household. Thus, while the evidence of significant impacts But there are some caveats to this test. For example, on health and nutrition are negligible and the results on when a client uses a service only once (and drops out) education are mixed as a result of microfinance interven- or if the service is priced below its market value (for tions, we are now able to see "empowerment" outcomes example, through subsidized interest rates), it is unclear for women that are significant. (It should be emphasized whether this service would still be valued if the cost that we are referring to changes attributable to credit only. were raised to a sustainable amount. The willingness- There are studies indicating that when credit services are to-pay test may also provide unclear results if a single combined with health education or nutrition education service is linked to other services (such as nutrition edu- services to groups, they do result in improvements to cation), if coercion is involved (as when a woman is health, nutrition, and education.) forced by her husband to take out a loan), or if the intrahousehold effects are mixed (for example, a decline Impact Proxies in girls' school attendance due to increased labor demands at home). Doing impact analysis well (and therefore credibly) can be Thus while willingness to pay is a low-cost, simple difficult and expensive. Addressing this dilemma, there is a proxy for impact, the weaknesses are substantial and school of thinking that advocates certain "proxies" for include the following considerations: impact. Otero and Rhyne have summarized recent micro- I The magnitude of impact is hard to determine. finance history by saying that there has been an important I Intrahousehold effects are not captured. shift from focusing on the individual firm or client of I Long-term development impact (poverty reduction) is financial services to focusing on the institutions providing unclear. services. This financial systems approach "necessarily relax- Perhaps the biggest argument against market-deter- es its attention to `impact' in terms of measurable enter- mined proxies such as willingness to pay is that they pre- prise growth and focuses instead on measures of increased sume that microfinance is a product for the marketplace access to financial services" (Otero and Rhyne 1994). like any other. If microfinance were merely a product Willingness to pay is sometimes viewed as a proxy for there would be little concern to look beyond willingness impact. Essentially, this view derives from a market eco- to pay, just as the sellers of cigarettes construe their prof- nomics paradigm--the clients are customers; if the prod- its as sufficient evidence of the success of their product. uct or service is made available to them and they buy it, Microfinance, however, is intended as a tool for poverty then there is value. Therefore, microfinance impact reduction, which is why many experts argue that analyz- analysis should concentrate on evidence of good outreach ing its impact on poverty is an unavoidable task (see and evidence of willingness to pay. Hulme 1997, who states clearly that monitoring MFI The rational for the willingness-to-pay test of impact performance in not enough). is that financial services usually require clients to pay the cost of acquiring the service in the form of interest pay- Client-Oriented Impact Analysis ments and fees as well as the transaction and opportunity cost of the time required to come to group meetings or The other school of thought about impact analysis consists to deal with other aspects of the loan process. If clients of those who believe that attempts must be made to assess, use the services (credit) repeatedly and, therefore, pay for analyze, and measure direct impacts. Unfortunately, the them repeatedly (and on time), it is evident that they dilemma of cost and the inherent difficulty of conducting value the services more than the cost. One can surmise, such an analysis well have been persistent problems, which therefore, that the client believes that the benefits have led to a general avoidance of the task. Few MFIs (impact) of the service on their lives exceeds the cost. invest much in impact analysis, and the literature on 50 MICROFINANCE HANDBOOK microfinance and microenterprise development has been the result of a direct intervention, the more the analysis remarkably short on discussions of the subject. (for exam- will encounter the complexity of human dynamics. For ple, the Small Enterprise Development Journal has had very microfinance aimed at poverty reduction, the heart of few articles on the subject in the past seven years). One the impact measurement problem is that the subjects are exception is PRODEM in Bolivia (box 2.9). human beings. As such they are unique and also embed- We stated earlier that what one measures is related ded in a culture and a society. And as if it were not to an MFI's mission and goals. If, for example, the enough to deal with these three variables (the culture, MFI's mission is poverty reduction, a further question the society, and the person), all three can and do change to be asked is: How does the MFI intervene on behalf at different rates of speed. of the poor? If the provision of credit is the main If the mission of a development project were to build instrument, then regardless of whether the focus is on a dam to irrigate thousands of acres of previously non- increasing income or household economic portfolio arable land, the impact of such goals could be assessed effects or empowerment outcomes, the ultimate subject with some confidence. The impact of the dam on the for any impact research will of necessity be human land could be accurately measured. There could be little beings. This is where most of the methodological doubt of findings that pointed to a specific number of dilemmas begin. hectares of improved soil and a resulting number of tons of food produced where none had grown before. There THE DILEMMA OF THE HUMAN SUBJECT AS A "DYNAMIC would, of course, be impacts on people's lives, and these TARGET." If an MFI wants to effect positive change in would be more difficult to measure, but then they were the lives of poor people, the more the expected impact is not the stated objectives of the project. Box 2.9 PRODEM's Impact and Market Anaylsis Project PRODEM IS A BOLIVIAN NONPROFIT ORGANIZATION THAT I Evaluate indicators of socioeconomic change for clients and provides support to the microenterprise sector with the aim their households over a three-year period, which can be of improving the quality of life for informal sector entrepre- linked to their use of PRODEM credit services. neurs. In 1995 Calmeadow, a Canadian nonprofit organiza- I Assess changes in PRODEM client businesses and pro- tion working in the field of microfinance, developed an ductive activities resulting from access to PRODEM impact study for PRODEM. Unlike traditional approaches credit services. to impact analysis, which aim to link changes in borrowers' I Evaluate the impact of PRODEM credit services in quality of life to their use of financial services, this impact selected economic subsectors to provide insight into the study was designed to provide direct feedback on not only operating context of PRODEM clients' businesses. I Gain enhanced understanding of rural clients' markets to the changes that may occur in the lives of PRODEM's assess credit needs and market opportunities. clients during the period that they are associated with PRO- I Use the data collection and analytical processes as a staff DEM, but also on their changing credit needs and the evolv- development opportunity for both field officers and head ing nature of their productive activities. It was anticipated office staff through exposure to new methodological that the information generated would help PRODEM to approaches. further develop its lending products to enhance the services I Contribute to the field of impact assessment generally it offers as a rural microlender. by developing an innovative study model that incorpo- The study was designed to analyze the systems of rela- rates analysis of the impact of new credit on local mar- tionships that make up the productive lives of PRODEM's kets and ties impacts at the household and business clients (suppliers, traders, transporters, and buyer markets) levels with changes occurring in the markets on which and to give insight to PRODEM's current and potential role they depend. as part of the financial strategies (for both production and The study used data collected through surveys and in- household consumption) of these clients. depth interviews in two areas in which PRODEM operates The study was based on six objectives: and in one control community. Source: Burnett 1995. THE TARGET MARKET AND IMPACT ANALYSIS 51 In the case of microfinance we see a similar dilemma, In addition, there are many other reasons not to but with a difference. If an MFI aims at providing finan- report accurately, including: cial services to the poor to reduce their poverty, then the I Embarrassment institution cannot take refuge in simply measuring how I Fear of taxation well it made services available (which is analogous to how I Not wanting others in the community to know well the dam was constructed). Because the stated end I Wanting to impress the person asking the question result--reducing poverty--is fully measurable only in I Wanting to please the person asking the question direct relationship to the lives of human beings, the num- I A different way of calculating things than the ques- ber of possible effects, distortions, and permutations of tionnaire--that is, the respondent genuinely may not those effects is, to put it mildly, far greater than in the know the answers to the questions. case when a project's direct impact is intended to be an increase in arable land as the result of building a dam. THE VESTED INTEREST OF THE ASSESSMENT TEAM. Those The dynamic interaction of individual humans, their conducting the impact assessment may have an interest in local society, and their culture can produce surprising the outcome of the study and thus consciously or uncon- results and can spoil a positive impact. Yaqub (1995) sciously may distort some responses. Anyone who is paid describes a group-based microcredit scheme operated by to do something may have a concern to please the agency BRAC in Bangladesh. The scheme was deemed success- paying them, especially when they are being paid to con- ful--as measured by increases in aggregate wealth of the duct an impact assessment for the agency whose project is members. But it is precisely because of aggregate wealth being evaluated. Such an interest can be expressed uncon- that the nature of the relationships in the group and its sciously, even in the very design of the survey instrument culture began to change. Because neither all persons in or the sample selection. At the level of data collection, it is the group nor their "life chances" are the same, some possible that enumerators, interviewers, or translators may gradually became better off than others. As Yaqub subtly avoid, ignore, or not notice data that may be offen- showed, some members with new wealth began to feel sive to the agency sponsoring the evaluation. that they had less to lose by ignoring group pressure to repay and thus became defaulters. This was a function of T HE ATTRIBUTION DILEMMA . Attributing an income the new power of their wealth rather than their inability change to the credit taken by someone requires knowing to repay. If the defaults caused by the successful use of in detail all the sources and uses of funds by the client credit resulted in the dissolution of the group and the (which in turn is related to the fungibility issue). subsequent ineligibility of other members for future Finding out the client's sources and uses of funds is loans, even their gains might be in jeopardy. Simply put, made difficult by the challenge of respondent honesty if the impact analysis story stopped at aggregate wealth and transparency as well as the fundamental fact that accumulation, it would miss a very important sequel. the livelihood strategies of most people, especially poor However, even before one gets to these unintended con- people, are complex, involving daily judgments about sequences and feedback loop effects, there are other practi- (and juggling of ) new opportunities and obstacles as cal problems involved in dealing with human subjects. well as new costs and old obligations. Attribution further requires knowing all the other CLIENT TRANSPARENCY AND HONESTY. It is commonly events and influences that occurred while the credit understood that money is fungible--that is, someone intervention was undertaken and separating them from may take a loan for the stated purpose of starting or the specific impacts of the credit and savings program. financing a business and then put the loan proceeds into There are other dynamic interactions that are func- the business, or they may use it to pay for their child's tions of microfinance itself. The use of one financial ser- education or wedding and then use other household vice can affect or influence the demand and usefulness of income to put into the business. If clients feel that they others. These, too, are difficult to trace. will not receive another loan if they have not actually used the preceding loan for the stated purpose, they may HYPOTHETICAL IMPACTS IN THE ABSENCE OF SERVICES. A not accurately report the use of the loan. different facet of the attribution dilemma is the difficulty 52 MICROFINANCE HANDBOOK in determining how outcomes might have been different because the loan comes at a time of particular need, because if no intervention had been made. To determine this someone else has asked them to take the loan on his or her requires, at the least, the establishment of baselines and behalf, because they fear loss of face in their group, because control groups. These are tasks that are not only costly they are hedging their bets in the belief that they may get a but in some sense impossible to do perfectly, since it is larger loan in the future, or even because they are counting extremely unlikely that one can find a sample of people on being able to default later on (when loan follow-up who are similar in every way to another sample of people, eventually slacked off), based on experience with past credit other than they did not receive a loan and others did. interventions. In the case research for the Sustainable Moreover, estimating those similarities and conducting Banking with the Poor project, dropout rates have been baseline surveys is subject to the same kinds of reporting increasing, particularly among target groups in which the errors as we have already noted. Still, one can get a sense prospects for sustained income growth are not promising. of what might have happened in the absence of the ser- vices by looking at how many other financing options When Should Impact Be Assessed? exist, how accessible they are to the same clientele, and how attractively they are priced (including pricing the The question of when to conduct an impact assessment social transaction costs). is not just a practical matter of cost and methodological Similar to the question of what would have happened advantage. As with the other aspects of impact, this ques- if the loan had not been made, one must also consider tion also relates to each MFI's mission and purpose. what other options for interventions were available and what resources could have been used for interventions, IMPACT ASSESSMENT BEFORE INTERVENTION. If the mission other than the provision of financial services. contains an absolute commitment to reach directly the poorest segment of the population (as is the case for Credit TEMPORARY IMPACTS. An additional dilemma is the ques- and Savings for the Hard-core Poor Network (CASH- tion of the ultimate value of the finding. Assume that POR) in the Philippines) then conducting a baseline one finds that an intervention was or appeared to be suc- impact assessment prior to the initiation of financial ser- cessful in raising people's incomes or reducing their pow- vices is essential to finding out who is going to be the main erlessness. Have they moved across the poverty line? beneficiary of the loans in terms of relative wealth. This Have the local structures in which their lives have been does not have to be expensive. In many villages in Asia, for embedded in the past altered in significant ways? In example, the material used in people's roofs can be a reli- short, how lasting are the changes? It is empirically evi- able indicator of relative poverty. CASHPOR therefore dent that small changes can be undone very quickly. advocates a house index based simply on comparative Small incremental moves upward in income or nutrition measurements of the three dimensions of the house and levels or changes in power relationships within the family notation of the roof material (Gibbons 1997b). can be reduced or wiped out by other reactions and forces, some of which are environmental, macroeconom- IMPACT ASSESSMENT AFTER INTERVENTION. The conven- ic (even related to world trade), or political. tional approach to impact assessment and analysis is to The usefulness of a loan obviously depends on circum- conduct the research after the MFI has been operating for stances, which can change. If, for example, poor women a set period of time, such as three years (a period often take loans repeatedly through several cycles, this simply says imposed by donors). In this approach there are often that they have been able to find ways to ensure repayment. three phases: first, a baseline study establishing some con- Especially in cases where women do not have regular and trols at the beginning of the MFI's operations, then an direct access to income, field studies have shown that they interim or midterm impact assessment, followed 18 borrow money from others in the family, take from remit- months or 2 years later by the final assessment. tance income, or even take from their own stocks of food Increasingly, this approach is being viewed as inadequate. to repay the loan. There are highly variable reasons why they may make an effort that can appear to be a sign of IMPACT ASSESSMENT DURING INTERVENTION. Those who how much they value the service. They may borrow are most concerned with human and social dynamics and THE TARGET MARKET AND IMPACT ANALYSIS 53 the methodological problems they present recommend I Scant notice to questionnaire concerns, such as survey that impact assessment be done on a continuous basis fatigue and the need for back translation. and that someone from the MFI remain involved with a I Infrequent inclusion of issues, such as politics, representative cross section of clients over time. This is favoritism, corruption, accountability, and leakages, as coming to be called "impact monitoring." Obviously, part of the design. this approach will likely involve keeping someone on the Measuring asset accumulation at the household level is MFI staff full-time for this purpose. an approach that is also increasingly recommended as a focus of impact study, especially as a way to get around Methods of Impact Assessment the inherent difficulties in measuring income changes accurately. In assessing the reviews of microfinance impact studies con- Barnes (1996) identifies six approaches to measuring ducted in recent years (for example, Gaile and Foster 1996) assets in the microenterprise literature: and recalling the dilemmas outlined above, it is tempting to I Attaching a current monetary value to assets and lia- conclude that the number of problems unearthed is so bilities great that the prospects of overcoming them are slim. I Specifying whether or not a specific asset is held, However, the opposite seems to be the case. The consensus which may be used to discuss the structure of the seems to be that multiple methods (as opposed to a single holding or other qualitative dimensions method) must be used and that combining qualitative and I Computing the flow value from productive assets quantitative approaches may be the only way to overcome I Ranking assets based on their assumed monetary these dilemmas (Hulme 1995; Carvalho and White 1997). value of other qualities The growing literature on microfinance impact assess- I Constructing an index that is a composite of measures ment is replete with details on the pitfalls of various I Determining the meaning of the assets to the owners approaches. For example, Gaile and Foster reviewed 11 and the social effects of the assets. studies and concluded that "some form of quasi experi- Barnes also notes that the biggest quantitative prob- mental design is appropriate" along with multivariate sta- lem with respect to asset accumulation involves depreci- tistical analysis. Their recommendations include an ating and valuing physical assets. overall sample size of about 500, "which should allow for It is because these pitfalls are unavoidable that there is effective use of control variables and for dealing with a convergence of opinion that a mix of methods needs to problems associated with longitudinal analysis; longitudi- be used. The basics of both quantitative and qualitative nal studies should have an interval of 18 to 24 months methods are reviewed briefly below. between data collection rounds." At the same time, their review concluded that "none of the reviewed studies suc- Fundamental Characteristics of Qualitative Approaches cessfully controlled for the fungibility of resources between household and enterprise." The most commonly prescribed alternatives to quantita- They recommend that control methods include: tive instruments and measures of impact fall within the I "Statistically equated control methods," which they fields of sociology and anthropology. The most recent feel are sufficient to address most control issues approach is the rubric of participant observation, which I Gender, which is a critical control variable is the classic mode of anthropological fieldwork. I Continued efforts to control for fungibility Participant observation is not so much a single method I Control methods that are a function of the data avail- as a combination of methods and techniques used to deal able. with the problems posed by the complexity of social Among other lacunae in the studies they reviewed organizations. (As noted above, lone humans are com- were: plex enough, and in reality humans are almost never I Minimal consideration as to the location of the MFI, alone. They are always part of a social unit, with levels which is a major determinant of success. of complexity all the greater.) As one textbook on the I Too little attention to alternative methodologies, such subject of participant observation puts it, "this blend of as qualitative methods and counterfactual analysis techniques ...involves some amount of genuine social 54 MICROFINANCE HANDBOOK interaction in the field with the subjects of the study, ings by following their "tracks" as it were, to see what some direct observation of relevant events, some formal significance they may have. and a great deal of informal interviewing, some systemat- Translated into the technique of "sampling" (the ic counting... and open-endedness in the directions the process of selecting those with whom to interact), open- study takes." (McCall and Simmons 1969). endedness means using what is called "nonprobability" The important general characteristics of such sampling. This begins with common sense--ensuring approaches usually include the following: diversity by choosing a range of people in a range of situ- I Intensivity (a continuous or regular presence among ations. But it also means being in position to take advan- clients over time). This enables the researcher to deepen tage of an accidental meeting with someone or to follow his or her understanding of what is going on and in a lead. turn enables trust to develop on the part of the respon- Rapid rural appraisal methods are derivative of classic dent. Naturally, one cannot be present among people sociological and anthropological approaches in that they over time without social interaction--hence the word also involve the use of semistructured interviews with key "participant" in the term "participant observation." informants, participant observation, and the method- I Structured observation. Observation is not as passive a ological principles of triangulation and open-endedness. function as it might seem. The researcher studying But rapid rural appraisal is more interactive and aims at a microfinance impacts has in mind some hypotheses shorter duration. It also employs such new techniques as about indicators and will naturally pay close attention (Robert Chambers work in Carvalho and White 1997): to them. They are thus part of a structured observa- I Focus groups with 5 to 12 participants from similar tion plan. The actual activity in the field may include backgrounds or situations (these may be observants or formal interviewing, using a questionnaire or inter- the subjects under analysis) view guide. The observation part comes into play as a I Social mapping and modeling drawn by participants form of research control. As a simple example, in ask- (often on the ground) to indicate which institutions ing about household assets a respondent may mention and structures of their community are important in having recently purchased a radio or mirror. If the their lives interview takes place in the respondent's home, the I Seasonality maps or calendars on which communities trained (observant) interviewer will look around to see show how various phenomena in their lives vary over if the radio or mirror is in sight. If it is not, the inter- the course of a year viewer may then ask where it is. I Daily time-use analysis I Triangulation. Triangulation involves checking data I Participatory linkage diagramming showing chains of with respondents and cross-checking with others to causality get different points of view of the same phenomenon. I Venn diagrams showing the relative importance of dif- Careful use of triangulation helps to deal with the ferent institutions or individuals in the community problems of the inherently untrustworthy respondent I Wealth ranking. and causality. I Open-endedness. This is the characteristic that causes Fundamental Characteristics of Quantitative some methodological distress, because it implies that Approaches there may be no closure to the effort or that it has no methodological rigor. In fact, the notion of open-end- By definition quantitative approaches (the word "quantita- edness is a response (again) to human dynamism and tive" derives from the Latin "quantus," meaning "of what complexity. It is another way of saying that the size") involve quanta (units) or things that can be counted, impact assessment process must be prepared to go including income, household expenditures, and wages. where the data lead. Often findings or indications of Because conclusions must be drawn from these numbers, findings are a surprise and may not fit a prior hypoth- the rules of statistics must be rigorously applied. esis. In many cases these findings, especially if not Probability is one of the most important of these rules, and numerically significant, are dismissed. In participant thus sampling (choosing which persons, households, or observation the objective is to make use of such find- enterprises to survey) is extremely important and has to be THE TARGET MARKET AND IMPACT ANALYSIS 55 done carefully to ensure that randomness is achieved (that Quasi-experimental methods attempt to mimic the is, that every unit in the population to be studied has an analysis of controlled experiments, with treatment and equal chance of being selected). Probability theory control groups created from different people. The differ- demands that the numbers of units to be studied be rela- ences between the treatment and control groups may be tively large and widespread to have statistical validity. observable or unobservable due to their nonrandom Quantitative approaches to microfinance impacts rely selection. on predesigned and pretested questionnaires. These are Nonexperimental methods use nonexperimental (non- formal rather than unstructured, because the responses random) survey data to look at the differences in behav- must be comparable and easy to count. Because survey ior between different people and relate the degree of questionnaires are structured and relatively fixed, little exposure to the treatment to variations in outcomes. The interactivity is likely between surveyor and respondent. absence of a control group separates such methods from As a result, survey questionnaires rely almost totally on quasi-experimental methods. respondent recall. This may be unreliable if the events Econometrics is an analytical technique that applies being recalled are far back in time. statistical methods to economic problems. Econometric Because of the widespread coverage necessary, the cost techniques have been used extensively to analyze data of designing, testing, and administering quantitative gathered using quasi- and nonexperimental methods. approaches is usually greater than that of qualitative Economists and econometricians have been studying sta- approaches. tistical methods for program evaluation for at least 25 The reason quantitative approaches have traditionally years (see Moffitt 1991, 291). During this time many been attractive to policymakers, however, derives from econometric techniques have been developed to mitigate the notion that there is strength in numbers. Numbers the various problems associated with nonrandom data offer confidence and reliability precisely because they are selection. Some of the techniques available for perform- presumed to represent true measures. If the methods ing econometric analysis are described below, along with used have been applied with rigor, the precision of the their advantages and disadvantages. results can be verified. Perhaps the most fundamental choice in an economet- Appendix 1 provides a matrix comparing some of the ric analysis is to select a model. Every analysis of an eco- major quantitative methods used in microfinance impact nomic, social, or physical system is based on some analysis. underlying logical structure, known as a model, which In general, the available quantitative methods fall into describes the behavior of the agents in the system and is three categories: the basic framework of analysis. In economics, as in the I Experimental methods physical sciences, this model is expressed in equations I Quasi-experimental methods describing the behavior of economic and related variables. I Nonexperimental methods. The model that an investigator formulates might consist of Experimental methods involve a natural or devised exper- a single equation or a system involving several equations. iment in which some randomly chosen group is given the In single-equation specification (regression analysis), "treatment" or, in development terms, the "intervention" the analyst selects a single (dependent) variable (denoted as (here, microfinance services). The results are then com- Y ) whose behavior the researcher is interested in explain- pared to a randomly selected group of controls from whom ing. The investigator then identifies one or a number of the treatment is withheld. The randomization guarantees (independent) variables (denoted by X ) that have causal that there are no differences--neither observable nor unob- effects on the dependent variable. In some econometric servable--between the two groups. Consequently, in theo- studies the investigator may be interested in more than ry, if there is a significant difference in outcomes, it can one dependent variable and hence will formulate several only be the effect of the treatment. Although the role of equations at the same time. These are known as simultane- experimental methods has been expanded in recent years, ous equation models (Ramanathan 1997, 6) there are many cases in which such experiments are difficult To estimate the econometric model that a researcher or impossible because of the cost or the moral or political specifies, sample data on the dependent and independent implications (Deaton 1997, 92). variables are needed (Hulme 1997, 1). Data are usually 56 MICROFINANCE HANDBOOK drawn in one of two settings. A cross section is a sample of interviewers, participant observers, translators, or focus a number of observational units all drawn at the same group facilitators. A large potential weakness in both quan- point in time. A time series is a set of observations drawn titative and qualitative methods is that the quality of these on the same observational unit at a number of (usually personnel can vary greatly. Consequently, if impact analy- evenly spaced) points in time. Many recent studies have sis is to be done, those paying for it should invest up front been based on time-series cross section, which generally in high-quality field personnel. This implies careful consist of the same cross section observed at several points recruiting, training, and supervision, as well as a sensible in time. Since the typical data set of this sort consists of a remuneration policy. large number of cross-sectional units observed at a few points in time, the common term "panel data set" is more Integrating Methodologies fitting for this type of study (Greene 1997, 102). "The key is to tap the breadth of the quantitative Comparisons of Quantitative and Qualitative Approaches approach and the depth of the qualitative approach. The exact combination of qualitative and quantita- The major underlying difference between quantitative and tive work will depend on the purpose of the study qualitative approaches is their philosophical position on and the available time, skills, and resources. In gen- "reality." The qualitative approach essentially says there are eral, integrating methodologies can result in better multiple forms of reality and which "reality" is in play measurement; confirming, refuting, enriching, and depends on whose point of view one takes. Advocates of explaining can result in better analysis; and merging the qualitative approach say, for example, that poverty the quantitative and qualitative findings into one set cannot be defined only by the donor or project personnel of policy recommendations can lead to better but must also be defined by the poor themselves, because action." (Carvalho and White 1997) poverty is more than simply lack of income. The quantita- Carvalho and White (1997) recommend specific types tive approach essentially assumes that there is only one of integration such as: reality that can be counted. In the case of poverty, the I Using the quantitative survey data to select qualitative quantitative approach would think it appropriate to have samples external surveyors determine adequate poverty indicators I Using the quantitative survey to design the interview such as amount of income, assets, access to basic services, guide for the qualitative work health, nutrition, potable water, and other measurable I Using qualitative work to pretest the quantitative ques- variables, either by yes or no answers or actual counting. tionnaire These basic differences have been summed up in the I Using qualitative work to explain unanticipated results commonly held notion that quantitative impact analysis is from quantitative data. more of a "science," while qualitative approaches are more of an "art" (Hulme 1997, 1). The Choice of Unit of Analysis Whereas for data in quantitative methods the vari- ables (for example, food expenditure) must be count- Regardless of the unit of analysis (the individual client, the able, in qualitative methods only the responses to enterprise, the household, or the community), it is impor- questions or the number of times a phenomenon was tant to investigate, get to know, and interview nonusers as observed can be counted, since the variables are atti- well as users of the financial service. Nonusers can be of tudes, preferences, and priorities. As a result, while two types: those who were and ceased to be users quantitative approaches are not prone to errors involv- (dropouts) and those who were never users. ing the representativeness of the sample, they are highly prone to what is called "nonsampling error." The oppo- THE CLIENT AS A CLIENT. This unit of analysis focuses on site is the case with qualitative approaches. the individual client as a client of the microfinance ser- The processes by which both quantitative and qualitative vices. The main purpose of such an analysis is for mar- methods are implemented necessarily depends on field ket research--to gather the clients' perceptions of the workers, whether employed as enumerators, surveyors, services so as to improve them. This has validity if con- THE TARGET MARKET AND IMPACT ANALYSIS 57 cerns about outreach have come up or if there are prise or the household (especially when this distinction dropouts. is not clear to the entrepreneur). THE CLIENT AS AN INDIVIDUAL. Both quantitative and THE SUBSECTOR. This unit and level of analysis is rarely if qualitative methods can focus on the client as an indi- ever used in microfinance impact analysis. It is brought to vidual (also called intrahousehold impact analysis when the reader's attention to recall that the level of analysis it extends to members of the client's household). The depends on the nature of the intervention and the intend- advantage of this level of investigation is that individual ed outcomes. It is possible that a microfinance program is clients are easily defined and identified. They can be sur- designed to focus on one subsector of the economy. Thus veyed or interacted with in focus groups through inter- besides using the enterprises in that subsector (for exam- views or participant observation. ple, all basket weavers) as units of analysis, the entire sub- Among the impacts to be sought at this level are sector would also be looked at in the aggregate. Indicators changes in income, allocation of income (to pay for could include jobs created; aggregate output; changes in girls' schooling, for example), changes in behavior (such subcontracts with other subsectors; alleviating constraints as willingness to take risks), changes in status or sense of affecting the whole subsector, such as input supply; and self, and so on. price changes brought about by changes in the subsector. The disadvantages of focusing on the client as an Needless to say, this level of analysis is highly research- individual is that estimations of the magnitude of oriented and requires significant time and resources. impact can be highly subjective and are likely to go beyond the person or persons looked at. Thus it THE HOUSEHOLD ECONOMIC PORTFOLIO. This approach looks becomes difficult to disaggregate the impacts on the pri- at the household, the individual, his or her economic mary subjects and those that extend to others. activity, and the local society in which he or she is embed- ded and traces interactions among them. Such an analysis THE ENTERPRISE. Using the enterprise as the unit of is difficult and costly and usually requires both a high analysis assumes that the goal of the MFI is largely degree of technical specialization and a good knowledge of economic: a change in the overall prospects of the local contexts. It can involve living on site for a period of enterprise (growth, greater profits, increase in produc- time. The payoff, however, is that it provides a more com- tion, higher sales, acquisition of a productive asset, prehensive picture of the impacts and the linkages between general increase in competitiveness, or some evidence different parts of the whole. of business transformation, such as moving to a fixed In summary, impact analysis is not easy or without real location). costs. At a time in microfinance when institutional sustain- The main methodological advantage of this unit of ability is an almost universal goal, finding ways to pay for analysis is the availability of well-understood analytical impact analysis poses serious challenges. However, the tools, such as profitability and return on investment future of microfinance may be at stake as a deliberate tool (Hulme 1997, 6). One disadvantage lies in defining the for poverty reduction. It is probably best if provisions for enterprise clearly enough for comparability with other impact analysis are made part of MFI activities from the enterprises. Many microenterprises are survivalist activi- beginning and the subject is given the same priority as ties rather than businesses in the conventional sense. financial sustainability. Impact analysis, for all its difficul- Thus the information needed for analysis will often be ties--and they are daunting--can be done well. The cur- lacking. The main disadvantage, however, is the prob- rent consensus seems to be that the first step is to take the lem of fungibility: the fact that there is no sure way to task seriously so that appropriate care and talent are devot- tell whether loan money has gone to benefit the enter- ed to carrying it out. 58 MICROFINANCE HANDBOOK Appendix 1. Quantitative Impact Assessment Most Common Impacts Analyzed with Quantitative Methods Enterprise level Household level Individual level Output Income Women's empowerment Asset accumulation Specialization Control over finances and other resources Risk management Diversification Contraceptive use Technology Asset accumulation Children Employment Savings Survival rates Management Consumption Health and nutrition Market Food Education Income Nonfood Exploitation Most Common Quantitative Methods Used to Analyze Impacts Quantitative method Advantages Disadvantages Experimental I Easiest to identify the effect of the I Often difficult to find or create situation program since treatment and control where a true experiment exists. groups are randomized Quasi-experimental I Can solve some problems plaguing I Hard to create a control group without bias nonexperimental data by simulating a I Some assumptions regarding the data true control group are required Nonexperimental I Assumptions made are rarely I Most complicated method for separating the case-specific and can lead, therefore, effects of program participation from to generalizable results other factors I Many assumptions are required of the data, which may not be met Econometric Options Quantitative method Advantages Disadvantages Choice of model Simple linear regression I Simplest of models I Assumptions required are often unrealistic (two-variable regression) I Can be used to handle nonlinear for most applications (estimates are relationships provided the model is unbiased, consistent, and efficient) linear in parameters Mulitvariate regression I The assumptions of the simple linear I Inclusion of too many variables can make regression model can be relaxed the relative precision of the individual somewhat, allowing for a more coefficients worse generalized application I Resulting loss of degrees of freedom would I Can accommodate nonlinearities reduce the power of the test performed on I Allows for the control of a subset of the coefficients independent (explanatory) variables to examine the effect of selected independent variables Simultaneous equation I Allows for the determination of several I Must consider additional problems of dependent variables simultaneously simultaneity and identification THE TARGET MARKET AND IMPACT ANALYSIS 59 Choice of data set Single cross section I Provides a relatively quick, I Cannot gauge the dynamics (magnitude inexpensive snapshot of a or rate) of change at the individual level representative group at a given point in time Repeated cross section I Can track outcomes or behaviors for I Cannot gauge the dynamics (magnitude or groups of individuals with less rate) of change at the individual level expense and time Longitudinal or panel I Allows analysis of dynamic changes I Attrition of and change in original panel at the individual level over time I Can be used to measure dynamic I Measurement periods are usually long with changes of the representative group high associated costs over time Source: Contributed by Stephanie Charitonenko-Church, Sustainable Banking with the Poor Project, World Bank. Sources and Further Reading Blair, Edmund. 1995. "Banking: MEED Special Report." Middle East Business Weekly 39 (June): 23­38. Almeyda, Gloria. 1996. Money Matters: Reaching Women Burnett, Jill. 1995. "Summary Overview of PRODEM Impact Microentrepreneurs with Financial Services. Washington, and Market Analysis Project." Calmeadow, Toronto, Canada. D.C.: United Nations Development Fund for Women and Berger, Marguerite, Mayra Buvinic, and Cecilia Jaramillo. 1989. Inter-American Development Bank. "Impact of a Credit Project for Women and Men Microentre- Barnes, Carolyn. 1996. "Assets and the Impact of Micro- preneurs in Quito, Ecuador." In Marguerite Berger and Mayra enterprise Finance Programs." AIMS (Assessing the Impact Buvinic, eds., Women's Ventures: Assistance to the Informal of Microenterprise Services) Brief 6. U.S. Agency for Sector in Latin America. Hartford, Conn.: Kumarian Press. International Development, Washington, D.C. Bolinick, Bruce R., and Eric R. Nelson. 1990. "Evaluating the Benjamin, McDonald, and Joanna Ledgerwood. 1998. "The Economic Impact of A Special Credit Programme Association for the Development of Microenterprises KIK/KMKP in Indonesia." Journal of Development Studies; (ADEMI): `Democratising Credit' in the Dominican 26 (2): 299­312. Republic." Case Study for the Sustainable Banking with the Carvalho, Soniya, and Howard White. 1997. Combining the Poor Project, World Bank, Washington, D.C. Quantitative and Qualitative Approaches to Poverty Bennett, Lynn. 1993. "Developing Sustainable Financial Systems Measurement and Analysis: The Practice and Potential. for the Poor: Where Subsidies Can Help and Where They World Bank Technical Paper 366. Washington, D.C. Can Hurt." Talking notes for the World Bank AGRAP CGAP (Consultative Group to Assist the Poorest). 1996. Seminar on Rural Finance, May 26, Washington, D.C. "Financial Sustainability, Targeting the Poorest and Income ------. 1997. "A Systems Approach to Social and Financial Impact: Are There Trade-offs for Micro-finance Intermediation with the Poor." Paper presented at the Institutions?" CGAP Focus Note 5. World Bank, Banking with the Poor Network/World Bank Asia Washington, D.C. Regional Conference on Sustainable Banking with the Chambers, Robert. 1992. "Rural Appraisal: Rapid, Relaxed and Poor, November 3­7, Bangkok. Participatory." Discussion Paper 311. University of Sussex, Bennett, Lynn, and Mike Goldberg. 1994. Enterprise Institute of Development Studies, United Kingdom. Development and Financial Services for Women: A Decade Chaves, Rodrigo A., and Claudio Gonzalez-Vega. 1994. of Bank Experience. World Bank, Asia Technical De- "Principles of Regulation and Prudential Supervision and partment, Washington, D.C. Their Relevance for Microenterprise Finance Binswanger, Hans, and Shahidur Khandker. 1995. "The Organizations." In Maria Otero and Elizabeth Rhyne, eds., Impact of Formal Finance on the Rural Economy of The New World of Microenterprise Finance. W. Hartford, India." Journal of Development Studies 32 (2): 234­62. Conn.: Kumarian Press. 60 MICROFINANCE HANDBOOK Chen, Martha, and Elizabeth Dunn. 1996. "AIMS Brief 3." ability." Paper presented at the Finance against Poverty U.S. Agency for International Development, Washington, Seminar, March 27, Reading University, United D.C. Kingdom. Christen, Robert Peck, Elisabeth Rhyne, Robert C. Vogel, and ------. 1997. "Impact Assessment Methodologies for Cressida McKean. 1995. Maximizing the Outreach of Microfinance: A Review." Paper prepared for the CGAP Microenterprise Finance: An Analysis of Successful Micro- Working Group on Impact Assessment Methodologies, finance Programs. Program and Operations Assessment Washington, D.C. Report 10. Washington, D.C.: U.S. Agency for Inter- Hulme, David, and Paul Mosley. 1996. Finance Against national Development. Poverty. London: Routledge. Corbett, Jane. 1988. "Famine and Household Coping Johnson, Susan, and Ben Rogaly. 1997. Microfinance and Strategies." World Development 16 (9): 1099­112. Poverty Reduction. London: Oxfam and ActionAid. Deaton, A. 1997. The Analysis of Household Surveys: A Khayatt, Djenan. 1996. Private Sector Development. Microeconomic Approach to Development Policy. Baltimore, Washington, D.C.: World Bank. Md.: The Johns Hopkins University Press. Lapar, Ma. Lucila A., Douglas H. Graham, Richard L. Meyer, Downing, Jeanne, and Lisa Daniels. 1992. "Growth and and David S. Kraybill. 1995. "Selectivity Bias in Dynamics of Women Entrepreneurs in Southern Africa." Estimating the Effect of Credit on Output: The Case of GEMINI Technical Paper 47. U.S. Agency for Rural Non-farm Enterprises in the Philippines." Ohio International Development, Washington, D.C. State University, Department of Agricultural Economics Dunn, Elizabeth. 1996. "Households, Microenterprises, and and Rural Sociology, Columbus, Ohio. Debt." AIMS Brief 5. U.S. Agency for International Lawai, Hussain. 1994. "Key Features of Islamic Banking." Development, Washington, D.C. Journal of Islamic Banking 11 (4): 7­13. Gaile, Gary L., and Jennifer Foster. 1996. "Review of Ledgerwood, Joanna. 1997. "Albanian Development Fund." Methodological Approaches to the Study of the Impact of Case Study for the Sustainable Banking with the Poor Microenterprise Credit Programs." AIMS (Assessing the Project. World Bank, Washington, D.C. Impact of Microenterprise Services project) paper, Liedhom, Carl, and Donald C. Mead. 1987. "Small Scale Management Systems International, Washington, D.C. Industries in Developing Countries: Empirical Evidence Garson, José. 1996. "Microfinance and Anti-Poverty and Policy Implications." International Development Strategies. A Donor Perspective." United Nations Paper 9. Michigan State University, East Lansing, Mich. Capital Development Fund Working Paper. UNDCF, Mahajan, Vijay, and Bharti Gupta Ramola. 1996. "Financial New York. Services for the Rural Poor and Women in India: Access Gibbons, David. 1997a. "Targeting the Poorest and Covering and Sustainability." Journal of International Development 8 Costs." Paper presented at the Microcredit Summit, (2): 211­24. February 3, Washington, D.C. Malhotra, Mohini. 1992. "Poverty Lending and Micro- ------. 1997b. Remarks made at the Microcredit Summit, enterprise Development: A Clarification of the Issues." February, Washington, D.C. GEMINI Working Paper 30. U.S. Agency for Inter- Goetz, Anne Marie, and Rina Sen Gupta. 1993. "Who Takes national Development, Washington, D.C. Credit? Gender, Power, and Control over Loan Use in McCall, George J., and J.L. Simmons. 1969. Issues in Rural Credit Programmes in Bangladesh." Institute of Participant Observation. Reading, Mass.: Addison-Wesley. Development Studies, University of Sussex, and Bangladesh McCracken, Jennifer A., Jules N. Pretty, and Gordon R. Institute of Development Studies, Dhaka. Conway. 1988. An Introduction to Rapid Rural Appraisal for Goldberg, Mike. 1992. Enterprise Development Services for Agricultural Development. London, U.K.: International Women Clients. Population and Human Resources, Institute for Environment and Development, Sustainable Women in Development. Washington, D.C.: World Bank. Agriculture Programme. Greene, W.H. 1997. Econometric Analysis. Upper Saddle River, Moffit, R. 1991. "Program Evaluation with Nonexperimental N.J.: Prentice Hall. Data." Evaluation Review 15 (3): 291­314. Hulme, David. 1995. "Finance for the Poor, Poorer, or Naponen, Helzi. 1990. "Loans to the Working Poor: A Poorest? Financial Innovation, Poverty, and Vulner- Longitudinal Study of Credit, Gender, and the Household THE TARGET MARKET AND IMPACT ANALYSIS 61 Economy." PRIE Center for Urban Policy Research, Schuler, Sidney Ruth, and Syed M. Hashemi. 1994. "Credit Rutgers University, New Brunswick, N.J. Programs, Women's Empowerment and Contraceptive Use O'Sullivan, Edmund. 1994. "Islamic Banking: MEED Special in Rural Bangladesh." Studies in Family Planning 25 (2): Report." Middle East Business Weekly 38 (August): 7­13. 65­76. Otero, Maria, and Rhyne, Elizabeth, eds. 1994. The New Sebstad, Jennifer, Catherine Neill, Carolyn Barnes, with Gregory Chen. 1995. "Assessing the Impacts of World of Microenterprise Finance. West Hartford, Conn.: Microenterprise Interventions: A Framework for Analysis." Kumarian Press. Managing for Results Working Paper 7. U.S. Agency for Parker, Joan, and C. Aleke Dondo. 1995. "Kenya: Kibera's International Development, Office of Microenterprise Small Enterprise Sector--Baseline Survey Report." GEMI- Development, Washington, D.C. NI Working Paper 17. U.S. Agency for International Sharif bin Fazil, Muhammad. 1993. "Financial Instruments Development, Washington, D.C. Conforming to Shariah." Journal of Islamic Banking 10 Paxton, Julia. 1996. "Worldwide Inventory of Microfinance (4): 7­20. Institutions." Sustainable Banking with the Poor Project, Von Pischke, J.D. 1991. Finance at the Frontier. World Bank, World Bank, Washington, D.C. Economic Development Institute. Washington, D.C. Pederson, Glen D. 1997. "The State of the Art in Waterfield, Charles, and Ann Duval. 1996. CARE Savings and Microfinance. A Summary of Theory and Practice." Credit Sourcebook. Atlanta, Ga: CARE. Sustainable Banking with the Poor Project, World Bank, Webster, Leila, Randall Riopelle, and Anne-Marie Chidzero. Washington, D.C. 1996. World Bank Lending for Small Enterprises, Pitt, Mark, and Shahidur R. Khandker. 1996. Household and 1989­1993. World Bank Technical Paper 311. Intrahousehold Impact of the Grameen Bank and Similar Washington, D.C. Targeted Credit Programs in Bangladesh. World Bank Weidemann, Jean. 1993. "Financial Services for Women." Discussion Paper 320. Washington, D.C. GEMINI Technical Note 3. U.S. Agency for International Putnam, Robert D., with Robert Leonardi, and Raffaella Y. Development, Washington, D.C. Nanetti. 1993. Making Democracy Work: Civic Traditions Yaqub, S. 1995. "Empowered to Default? Evidence from in Modern Italy. Princeton, N.J.: Princeton University BRAC's Microcredit Programmes." Journal of Small Press. Enterprise Development 6 (4) December. Ramanathan, R. 1997. Introductory Econometrics with Yaron, Jacob, McDonald P. Benjamin, Jr., and Gerda L. Applications. Baltimore, Md.: Academic Press. Piprek 1997. "Rural Finance: Issues, Design, and Best Rhyne, Elisabeth, and Sharon Holt. 1994. "Women in Finance Practices." Environmentally and Socially Sustainable and Enterprise Development." ESP Discussion Paper 40. Development Studies and Monographs Series 14. World World Bank, Washington, D.C. Bank, Rural Development Department. Washington, D.C. CHAPTER THREE Products and Services I n chapter 1 we examined various contextual must be developed. In many cases clients also need factors that affect the supply of financial services specific production and business management skills as to low-income women and men. In chapter 2 we well as better access to markets if they are to make looked at the objectives of MFIs and ways to identify a profitable use of the financial services they receive target market and assess impacts. In this chapter we (Bennett 1994). look at the range of products and services that an MFI Providing effective financial services to low-income might offer, taking into account the supply and women and men therefore often requires social interme- demand analysis outlined in the previous two chapters. diation--"the process of creating social capital as a sup- MFIs can offer their clients a variety of products port to sustainable financial intermediation with poor and services. First and foremost are financial services. and disadvantaged groups or individuals" (Bennett However, due to the nature of an MFI's target cli- 1997). Most MFIs provide some form of social interme- ents--poor women and men without tangible assets, diation, particularly if they work with groups. In some who often live in remote areas and may be illiterate-- cases social intermediation is carried out by other orga- MFIs cannot operate like most formal financial insti- nizations working with MFIs. tutions. Formal financial institutions do not generally In addition, some MFIs provide enterprise develop- regard the tiny informal businesses run by the poor as ment services such as skills training and basic business attractive investments. The loan amounts the business- training (including bookkeeping, marketing, and pro- es require are too small, and it is too difficult to obtain duction) or social services such as health care, education, information from the clients (who may communicate and literacy training. These services can improve the in local dialects that the lender does not know). The ability of low-income men and women to operate clients are also often too far away, and it takes too long microenterprises either directly or indirectly. to visit their farms or businesses, especially if they are Deciding which services to offer depends on the located in the urban slum. All of this means that the MFI's objectives, the demands of the target market, the costs per dollar lent will be very high. And on top of existence of other service providers, and an accurate cal- that, there is no tangible security for the loan (Bennett culation of the costs and feasibility of delivering addi- 1994). tional services. This means that low-income men and women face This chapter will be of interest to donors who are formidable barriers in gaining access to mainstream considering supporting a microfinance activity or exist- financial service institutions. Ordinary financial inter- ing MFIs that are considering adding new services or mediation is not usually enough to help them partici- changing the way in which they operate. In particular, pate, and therefore MFIs have to create mechanisms to the overview of the products and services an MFI might bridge the gaps created by poverty, illiteracy, gender, offer its clients enables the reader to determine the level and remoteness. Local institutions must be built and at which it meets the needs and demands of the target nurtured, and the skills and confidence of new clients market. 63 64 MICROFINANCE HANDBOOK The Systems Framework ing subsidy for this work. The systems approach allows us to deal with the fact that microfinance involves a mix Providing microfinance services to marginal clients is a of "business" and "development." Instead of trying to complex process that requires many different kinds of force formal banks to become NGOs worrying about skills and functions. This may require more than one social intermediation, empowerment, and participation, institution. Even MFIs taking an integrated approach do and instead of urging NGOs to transform themselves not often provide all of the services demanded by the into banks that must make a profit, we can encourage target group (see below). Thus understanding how the institutions to form partnerships in which each does what process of financial and social intermediation takes place it does best, pursuing its own corporate mission while requires a systems analysis rather than a simple institu- combining different skills to provide a lasting system that tional analysis.1 provides the poor with access to financial services. The systems perspective is important not only (However, it is important to note that a new breed because there may be a number of different institutions of NGOs, called "business NGOs," have begun to involved, but also because these institutions are likely to emerge. In the discourse currently surrounding microfi- have very different institutional goals or "corporate mis- nance, there seems to be an expectation that all NGOs sions." Thus if a commercial bank is involved, its goal is involved in microfinance will basically transform them- to build its equity and deliver a profit to its owners. In selves into financial institutions to achieve the goal of contrast, both government agencies and nongovern- financial self-sustainability. Because not all NGOs are mental organizations (NGOs), despite their many the same, considerable caution is required before draw- differences, are service organizations rather than profit- ing this conclusion. Before encouraging an NGO to making institutions. Credit and savings clients them- become a formal financial institution, it is important to selves, if formed as a membership organization, have be very clear about the organization's institutional their own corporate mission: to serve its members who goals, management capacity, and experience. Not all are both clients and owners. NGOs can--or should--become banks. Indeed, there The systems approach is useful for both donors and are many valid roles for NGOs to play within a sustain- practitioners, because it makes the issue of subsidies much able system of financial intermediation--even if they easier to understand and deal with honestly. It allows us themselves are not able to become financially self-sus- to see each institution involved in the intermediation tainable in the way that a well performing bank or process as a separate locus of sustainability when we are cooperative can.) assessing the commercial viability of the whole system. Within the systems framework there are four broad When it comes to covering the costs of providing services categories of services that may be provided to microfi- within the systems framework, each institution may have nance clients: a different perspective. For formal financial institutions I Financial intermediation, or the provision of financial and membership organizations, financial sustainability is products and services such as savings, credit, insur- an essential goal. NGOs although expected to operate ance, credit cards, and payment systems. Financial efficiently and to cover as much of their costs as possible, intermediation should not require ongoing subsidies. are not expected to generate a profit. Typically they work I Social intermediation, or the process of building the to deliver financial services to a target group for whom human and social capital required by sustainable "the market" has failed. For many NGOs financial self- financial intermediation for the poor. Social interme- sustainability as an institution is not a goal consistent diation may require subsidies for a longer period than with their corporate missions. financial intermediation, but eventually subsides If we consider the system as a whole, we can be much should be eliminated. clearer about where the subsidies are. If a given institu- I Enterprise development services, or nonfinancial services tion is solely involved in providing social intermediation, that assist microentrepreneurs. They include business then we can make an informed decision to provide ongo- training, marketing and technology services, skills 1. This discussion is taken from Bennett (1997). PRODUCTS AND SERVICES 65 development, and subsector analysis. Enterprise devel- Microfinance Institutions--Minimalist or opment services may or may not require subsidies, Integrated? depending on the willingness and ability of clients to pay for these services. MFIs by definition provide financial services. However, I Social services, or nonfinancial services that focus on an MFI may also offer other services as a means of improving the well-being of microentrepreneurs. improving the ability of its clients to utilize financial ser- They include health, nutrition, education, and litera- vices. There is much debate in the field of microfinance cy training. Social services are likely to require ongo- as to whether MFIs should be minimalist--that is, offer- ing subsidies, which are often provided by the state or ing only financial intermediation--or integrated--offer- through donors supporting NGOs. ing both financial intermediation and other services The degree to which an MFI provides each of these (figure 3.1). Most MFIs offer social intermediation to services depends on whether it takes a "minimalist" or some extent. The decision to offer nonfinancial services "integrated" approach. determines whether an MFI is minimalist or integrated. Figure 3.1 Minimalist and Integrated Approaches to Microfinance MINIMALIST APPROACH INTEGRATED APPROACH One "missing piece"--credit Financial and nonfinancial services Source: Author`s schematic. 66 MICROFINANCE HANDBOOK MFIs using the minimalist approach normally offer "Finance in the form of savings and credit arises to per- only financial intermediation, but they may occasional- mit coordination. Savings and credit are made more effi- ly offer limited social intermediation services. cient when intermediaries begin to transfer funds from Minimalists base their approach on the premise that firms and individuals that have accumulated funds and there is a single "missing piece" for enterprise growth, are willing to shed liquidity, to those that desire to usually considered to be the lack of affordable, accessi- acquire liquidity" (Von Pischke 1991, 27). ble short-term credit, which the MFI can offer. While While virtually all MFIs provide credit services, other "missing pieces" may exist, the MFI recognizes its some also provide other financial products, including comparative advantage in providing only financial savings, insurance, and payment services. The choice of intermediation. Other organizations are assumed to which financial services to provide and the method of provide other services demanded by the target clients. providing these services depends on the objectives of This approach offers cost advantages for the MFI and the MFI, the demands of its target market, and its insti- allows it to maintain a clear focus, since it develops and tutional structure. provides only one service to clients. Two key imperatives that must be considered when The integrated approach takes a more holistic view of providing financial services are: the client. It provides a combination or range of finan- I To respond effectively to the demand and prefer- cial and social intermediation, enterprise development, ences of clients and social services. While it may not provide all four, I To design products that are simple and can be easily the MFI takes advantage of its proximity to clients and, understood by the clients and easily managed by the based on its objectives, provides those services that it MFI. feels are most needed or that it has a comparative The range of products commonly provided includes: advantage in providing. I Credit An MFI chooses a minimalist or more integrated I Savings approach depending on its objectives and the circum- I Insurance stances (demand and supply) in which it is operating. If I Credit cards an MFI chooses to take an integrated approach, it I Payment services. should be aware of the following potential issues: The following section provides a brief description of the I Providing financial and nonfinancial services are two products MFIs offer their clients. This overview is provided distinct activities, which may at times lead an insti- to familiarize the reader with the unique ways that financial tution to pursue conflicting objectives. services are provided to microentrepreneurs. Specific infor- I It is often difficult for clients to differentiate "social mation on how to design credit and savings products can services," which are usually free, from "financial ser- be found in chapters 5 and 6. Appendix 1 summarizes the vices," which must be paid for, when they are receiv- main approaches to financial intermediation. ing both from the same organization. I MFIs offering many services may have difficulties Credit identifying and controlling the costs per service. I Nonfinancial services are rarely financially sustainable. Credit is borrowed funds with specified terms for repay- ment. When there are insufficient accumulated savings to finance a business and when the return on borrowed Financial Intermediation funds exceeds the interest rate charged on the loan, it makes sense to borrow rather than postpone the business The primary role of MFIs is to provide financial interme- activity until sufficient savings can be accumulated, diation. This involves the transfer of capital or liquidity assuming the capacity to service the debt exists from those who have excess at a particular time to those (Waterfield and Duval 1996). who are short at that same time. Since production and Loans are generally made for productive purposes-- consumption do not take place simultaneously, some- that is, to generate revenue within a business. Some thing is needed to coordinate these different rhythms. MFIs also make loans for consumption, housing, or spe- PRODUCTS AND SERVICES 67 cial occasions. While many MFIs insist that only produc- cover costs, and that clients are motivated to repay loans. tive loans be made, any loan that increases the liquidity MFIs can be sustainable providing they have enough funds of the household frees up enterprise revenue, which can to continue operating in the long term. These funds can be put back into the business. be obtained solely through operational revenue or through Most MFIs strive to reach sustainability (which may or a combination of grants and operating revenue. As micro- may not include financial sustainability--see chapter 9) by finance develops, clear principles are being established that ensuring that the services offered meet the demands of lead to financially viable lending (box 3.1). clients, that operations are as efficient as possible and costs Methods of credit delivery can generally be divided into are minimized, that interest rates and fees are sufficient to the two broad categories of individual and group Box 3.1 Principles of Financially Viable Lending to Poor Entrepreneurs Principle 1. Offer services that fit the preferences of poor entre- rooms in local government buildings, paying little or no preneurs. overhead while reaching deep into rural areas. Select staff These services might include: from local communities, including people with lower levels I Short loan terms, compatible with enterprise outlay and of education (and hence lower salary expectations) than income patterns. ACCION International programs typi- staff in formal banking institutions. cally lend for three-month terms, and Grameen Bank for one year. Principle 3. Motivate clients to repay loans. I Repeat loans. Full repayment of one loan brings access to Substitute for preloan project analysis and formal collateral another. Repeat lending allows credit to support financial by assuming that clients will be able to repay. Concentrate management as a process rather than as an isolated event. on providing motivation to repay. These motivations might I Relatively unrestricted uses. While most programs select include: customers with active enterprises (and thus the cash I Joint liability groups. An arrangement whereby a hand- flow for repayment), they recognize that clients may ful of borrowers guarantees each other's loans is by far need to use funds for a mixture of household or enter- the most frequently used repayment motivation. This prise purposes. technique is employed by Grameen and in a slightly I Very small loans, appropriate for meeting the day-to-day different form by ACCION affiliates. It has proved financial requirements of businesses. Average loan sizes at effective in many different countries and settings Badan Kredit Kecamatan in Indonesia and Grameen are worldwide. Individual character lending can be effec- well under $100, while most ACCION and Bank tive when the social structure is cohesive, as has been Rakyat Indonesia activities feature average loans in the demonstrated throughout Indonesia's array of credit $200 to $800 range. programs. I A customer-friendly approach. Locate outlets close to I Incentives. Incentives such as guaranteeing access to loans entrepreneurs, use extremely simple applications (often motivate repayment, as do increases in loan sizes and one page), and limit the time between application and preferential pricing in exchange for prompt repayment. disbursement to a few days. Develop a public image of Institutions that successfully motivate repayments devel- being approachable by poor people. op staff competence and a public image that signals that they are serious about loan collection. Principle 2. Streamline operations to reduce unit costs. Develop highly streamlined operations, minimizing staff Principle 4. Charge full-cost interest rates and fees. time per loan. The small loan sizes necessary to serve the poor may result Standardize the lending process. Make applications very in costs per loan requiring interest rates that are significant- simple and approve on the basis of easily verifiable criteria, ly higher than commercial bank rates (though significantly such as the existence of a going enterprise. Decentralize lower than informal sector rates). Poor entrepreneurs have loan approval. Maintain inexpensive offices--Badan Kredit shown a willingness and an ability to pay such rates for ser- Kecamatan operates its village posts once a week from vices with attributes that fit their needs. Source: Rhyne and Holt 1994. 68 MICROFINANCE HANDBOOK approaches, based on how the MFI delivers and guaran- I The screening of potential clients by credit checks and tees its loans (adapted from Waterfield and Duval 1996). character references. I Individual loans are delivered to individuals based on I The tailoring of the loan size and term to business their ability to provide the MFI with assurances of needs. repayment and some level of security. I The frequent increase over time of the loan size and I Group-based approaches make loans to groups--that term. is, either to individuals who are members of a group I Efforts by the staff to develop close relationships with and guarantee each other's loans or to groups that clients so that each client represents a significant then subloan to their members. investment of staff time and energy. Individual lending requires frequent and close contact INDIVIDUAL LENDING. MFIs have successfully developed with individual clients. It is most often successful in urban effective models to lend to individuals, which combine areas where client access is possible. Individual lending can formal lending, as is traditional in financial institutions, also be successful in rural areas, particularly through sav- with informal lending, as carried out by moneylenders. ings and credit cooperatives or credit unions. In both Formal financial institutions base lending decisions on urban and rural areas individual lending is often focused business and client characteristics, including cash flow, on financing production-oriented businesses, whose opera- debt capacity, historical financial results, collateral, and tors are generally better off than the very poor. character. Formal sector lenders have also proven the The Fédération des Caisses d'Epargne et de Crédit usefulness of personal guarantors to motivate clients to Agricole Mutuel in Benin provides an example of effec- repay loans. They have demonstrated the value of a busi- tive individual lending in rural areas (box 3.2). nessike approach and the importance of achieving cost Because credit officers need to spend a relatively long recovery in their lending operations. Finally, they have period of time with individual clients, they usually serve established the importance of external regulation to safe- between 60 and 140 clients. Loans to individuals are guard client savings and the institution itself. However, usually larger than loans to members in a group. formal sector lending practices are often not suitable for Accordingly, given an equal number of loans, these microfinance institutions, as many microbusinesses or loans to individuals provide a larger revenue base to cover business owners do not have many assets or adequate the costs of delivering and maintaining the loans than financial reporting systems. group loans. The revenue base is the outstanding Informal sector lenders approve loans based on a person- amount of a loan portfolio that is earning interest rev- al knowledge of the borrowers rather than on a sophisti- enue. Because interest revenue is based on a percentage cated feasibility analysis, and they use informal collateral of the amount lent, the larger the amount lent the sources. They also demonstrate the importance of greater the revenue and hence the more funds available responding quickly to borrowers' needs with a minimum to cover costs, even though costs may not be that much of bureaucratic procedures. Perhaps most important, higher than for loans of smaller amounts (see chapter 9). moneylenders demonstrate that the poor do repay loans Furthermore, individual lending models may be less and are able to pay relatively high interest rates. However, costly and less labor-intensive to establish than group- loans from moneylenders are often not taken for produc- based models. tive purposes but rather for emergency or consumption The Association for the Development of Microenter- smoothing. They are usually for a relatively short period prises in the Dominican Republic provides an excellent of time. example of a successful MFI that provides individual loans Characteristics of individual lending models include (box 3.3). (Waterfield and Duval 1996, 84): I The guarantee of loans by some form of collateral GROUP-BASED LENDING. Group-based lending involves the (defined less stringently than by formal lenders) or a formation of groups of people who have a common wish cosigner (a person who agrees to be legally responsible to access financial services. Group-lending approaches for the loan but who usually has not received a loan of frequently build on or imitate existing informal lending her or his own from the MFI). and savings groups. These groups exist in virtually every PRODUCTS AND SERVICES 69 Box 3.2 Individual Loans at Fédération des Caisses d'Epargne et de Crédit Agricole Mutuel, Benin THE FÉDÉRATION DES CAISSES D'EPARGNE ET DE CRÉDIT farmers group participates in the credit analysis and acts as a Agricole Mutuel (FECECAM) is a large network of savings guarantor. and credit cooperatives with more than 200,000 clients. Most Individual lending has been very much geared toward of the loans are provided to individuals. Loan analysis is con- men because they have the most assets (and hence collat- ducted by a credit committee composed of representatives eral) and as heads of households receive support from the from the villages to which the local savings and credit cooper- farmers group. For women this approach has been rather ative provides financial services. Loans are therefore character- ineffective. In 1993 the federation introduced group based, although some form of collateral is also required. lending for female clients. The goal is to help women Borrowers must also have a savings account with the coopera- become familiar with the savings and credit cooperatives tive, and the loan amount they receive is related to their vol- and build up enough savings to become individual bor- ume of savings. Savings cannot be drawn on until the loan is rowers after taking two or three loans with the backing of paid back in full. In the case of agricultural loans, the village their group. Source: Fruman 1997. country and are called by various names, the most rather than having to wait for their turn. More well- common being rotating savings and credit associations known group-lending models include the Grameen (box 3.4). Bank in Bangladesh and ACCION International's soli- Group-lending approaches have adapted the model darity group lending, both of which facilitate the forma- of rotating savings and credit associations to provide tion of relatively small groups (of 5 to 10 people) and additional flexibility in loan sizes and terms and general- make individual loans to group members. Other models, ly to allow borrowers to access funds when needed such as the Foundation for International Community Box 3.3 The Association for the Development of Microenterprises T HE A SSOCIATION FOR THE D EVELOPMENT OF M ICRO - ADEMI uses a combination of collateral and guarantors to enterprises (ADEMI) provides credit and technical assistance secure its loans. All borrowers are required to sign a legal con- to microenterprises and small enterprises. The credit service tract stating their obligation to repay funds at specified terms. has the following characteristics: Most clients also have a guarantor or cosigner who assumes I Emphasis on providing credit to microenterprises, fol- full responsibility for the terms of the contract in case the lowed by appropriate amounts of noncompulsory, com- client is either unable or unwilling to repay the loan. When plementary, direct technical assistance. feasible, borrowers are also required to sign deeds for property I Initial, small, short-term loans provided for working cap- or machinery as loan guarantees. In some case guarantors may ital and later increased in successively larger amounts and be asked to pledge security on behalf of the borrower. longer terms to purchase fixed assets. However, ADEMI strongly recommends that if appli- I Use of commercial banks for loan disbursements and col- cants are not able to provide collateral they should be limited lection. in their ability to access the association's credit. When first- I Positive real rates of interest charged on the loans. time borrowers are unable to provide security, loan officers I Caution in disbursements, so that the timing of the loan rely heavily on information as a collateral substitute. In these and the amount are appropriate. situations the loan officer will investigate the reputation of The size and terms of each loan provided by the associa- the applicant in much the same way that an informal tion are set individually by the loan officer based on the moneylender might. Visits are made to neighboring shops client's needs and capacity to repay. Loans are approved by and bars. Applicants may be asked to produce financial state- management. Most loans to microenterprises are for up to one ments if possible or receipts for utility payments. Each appli- year, although the term may be longer for fixed asset loans. cation is considered on a case-by-case basis. Source: Ledgerwood and Burnett 1995. 70 MICROFINANCE HANDBOOK Box 3.4 Rotating Savings and Credit Associations ROTATING SAVINGS AND CREDIT ASSOCIATIONS (ROSCAS) switch turns if someone is in need, and some associations are informal groups developed by their members at the grass- build up security funds for mutual assistance. Variations roots level. They bear different names in different countries. exist in the size of groups, the amounts "saved," the frequen- For example, in West Africa they are tontines, paris, or susus; cy of meetings, and the order of priority in receiving loans. in South Africa they are stokvels; in Egypt they are gam'iyas; In Latin America commercial associations (tandas) are used in Guatemala they are cuchubales; and in Mexico they are to buy expensive consumer durables such as cars or machin- tandas. ery. Some people belong to several associations. The rotating Individuals form a self-selected group and all members savings and credit associations can be rural or urban and agree to contribute a regular fixed amount every week or include men and women, usually in separate groups, some- month. Members then take turns receiving the full amount times jointly. Members can be at any income level; many collected in the period until all members have had a turn to civil servants or office employees participate. receive funds. The order in which members receive funds is However, rotating savings and credit associations are lim- determined by lottery, mutual agreement, or need or person- ited due to their lack of flexibility. Members are not always al emergencies of the group members. Often the person who able to receive a loan when needed or in the amount initiated the formation of the association receives the funds required. Also, membership in some associations involves first. high social and transaction costs, such as regular meetings Interest is not directly charged but is implicit. Those who and providing refreshments for group members. Rotating benefit from the loan early are privileged over those who savings and credit associations also imply risk in the form of receive it in the late stages. Rotating savings and credit asso- the chance that some members will drop out before everyone ciations also play an important social role. Members might has had a turn at receiving the contributions. Source: Krahnen and Schmidt 1994. Assistance (FINCA) village banking model, utilize larger member and offer to help until the problem is resolved. groups of between 30 and 100 members and lend to the In still other cases, the mandatory savings of group mem- group itself rather than to individuals. bers may be used to pay off the loan of a defaulter. Several advantages of group-based lending are cited fre- Another advantage of group lending is that it may quently in microfinance literature. One important feature reduce certain institutional transaction costs. By shifting of group-based lending is the use of peer pressure as a screening and monitoring costs to the group, an MFI can substitute for collateral. Many group-based lending pro- reach a large number of clients even in the face of asym- grams target the very poor, who cannot meet the tradi- metric information through the self-selection of group tional collateral requirements of most financial members. One of the reasons that self-selection is so institutions. Instead, group guarantees are established as important is that members of the same community gener- collateral substitutes. While many people presume that ally have excellent knowledge about who is a reliable credit group guarantees involve the strict joint liability of group risk and who is not. People are very careful about whom members, members are seldom held responsible. Instead, they admit into their group, given the threat of losing their the default of one member generally means that further own access to credit (or having their own savings used to lending to other members of the group is stopped until repay another loan). Hence group formation is a critical the loan is repaid. The financial and social grouping elicits component of successful group lending. In addition to several types of group dynamics that may increase repay- cost reduction through internal group monitoring and ment rates. For example, peer pressure from other group screening of clients, MFIs using group-based lending also members can act as a repayment incentive, since members can save by using a hierarchical structure. Typically, loan do not want to let down the other members of their officers do not deal with individual group members but group or suffer any social sanctions imposed by the group rather collect repayment from a group or village leader, as a result of defaulting. In other cases, the group may rec- thereby reducing transaction costs. Puhazhendhi (1995) ognize a legitimate reason for the arrears of a certain found that microfinance banks that use intermediaries PRODUCTS AND SERVICES 71 such as self-help groups reduce transaction costs more suc- Savings cessfully than banks that work directly with clients. While numerous potential advantages of group lend- Savings mobilization has long been a controversial issue in ing have been reported, several disadvantages exist as microfinance. In recent years there has been increasing well. Bratton (1986) demonstrates how group lending awareness among policymakers and practitioners that institutions have better repayment rates than individual there is a vast number of informal savings schemes and lending programs in good years but worse repayment MFIs around the world (in particular, credit union orga- rates in years with some type of crisis. If several mem- nizations) have been very successful in mobilizing savings. bers of a group encounter repayment difficulties the These developments attest to the fact that low-income entire group often collapses, leading to a domino effect. clients can and do save. The World Bank's "Worldwide This effect was a strong and significant determinant of Inventory of Microfinance Institutions" found that many group loan repayment in a study in Burkina Faso of the largest, most sustainable institutions in microfi- (Paxton 1996b) and signals the inherent instability of nance rely heavily on savings mobilization. "In 1995, over group lending in risky environments (box 3.5). $19 billion are held in the surveyed microfinance institu- Some critics question the assumption that transaction tions in more than 45 million savings accounts compared costs are indeed lower with group lending (for example, to nearly $7 billion in 14 million active loan accounts. Huppi and Feder 1990). Group training costs tend to be Often neglected in microfinance, deposits provide a high- quite high, and no individual borrower-bank relationship ly valued service to the world's poor who seldom have is established over time. Not only are institutional trans- reliable places to store their money or the possibility to action costs high, but client transaction costs are quite earn a return on savings" (Paxton 1996a, 8). high as well as more responsibility is shifted from the The survey also found that the ability to effectively MFI to the clients themselves. mobilize deposits depends greatly on the macroeconomic Finally, while it appears that some people work well and legal environment. "Statistical analysis of the surveyed in groups, there is the concern that many people prefer institutions reveals a positive correlation between the to have individual loans rather than being financially amount of deposits mobilized and the average growth in per punished for the irresponsible repayment of other group capita GNP of the country from 1980 to 1993. Likewise, members. Given the presence of opposing advantages higher deposit ratios are negatively correlated with high lev- and disadvantages associated with group lending, it is els of inflation. Finally, the amounts of deposits are positive- not surprising to see group lending work well in some ly correlated with high levels of population density" (Paxton contexts and collapse in others (see box 3.6). 1996a, 8). Furthermore, institutions operating with donor Box 3.5 Repayment Instability in Burkina Faso A 1996 STUDY OF GROUP LENDING EXAMINED THE INFLU- Thus an incentive existed for correctly paying groups to ences of group dynamics on loan repayment in 140 lending default if any default existed within the village. In addition, groups of Sahel Action. Several variables were shown to have groups tended to default after several loan cycles. A possi- a positive influence on successful loan repayment, including ble explanation for this phenomenon is that the terms and group solidarity (helping a group member in need), location conditions of the group loan no longer fulfilled the (urban outperformed rural), access to informal credit, and demand of each group member in consecutive loan cycles, good leadership and training. Nevertheless, other variables led leading one or more members to default. to repayment instability. Positive and negative externalities exist within any The variable that had the most destabilizing effect on group-based lending program. However, given the impor- loan repayment was the domino effect. Joint liability exist- tance and delicacy of maintaining strong repayment rates, a ed not only at the group level but also at the village level, thorough understanding of the factors leading to repay- in the sense that if any group in the village defaulted on its ment instability is critical to reduce the causes of wide- loan then no other loans would be issued in the village. spread default. Source: Paxton 1996b. 72 MICROFINANCE HANDBOOK Box 3.6 The Role of Groups in Financial Intermediation Guidelines for effective use of groups: I Improved loan collection through screening and selecting, I Groups are more effective if they are small and homoge- peer pressure, and joint liability, especially if group penal- neous. ties and incentives are incorporated in the loan terms. I Imposing group penalties and incentives (such as no I Improved savings mobilization, especially if incentives access to further loans while an individual is in default) are incorporated in a group scheme. improves loan performance. I Lower administrative costs (selection, screening, and loan I Loan sizes that increase sequentially appear to allow collection) once the initial investment is made in estab- groups to screen out bad risks. lishing and educating the groups. I Staggered disbursements to group members can be based on the repayment performance of other members. Risks associated with using groups: I Poor records and lack of contract enforcement. Possible advantages of using groups: I Potential for corruption and control by a powerful leader I Economies of scale (a larger clientele with minimal within the group. increases in operating costs). I Covariance risk due to similar production activities. I Economies of scope (an increased capacity to deliver I Generalized repayment problems (domino effect). multiple services through the same group mechanism). I Limited participation by women members in mixed- I Mitigation of information asymmetry related to potential gender groups. borrowers and savers through the group's knowledge of I High up-front costs (especially time) in forming viable individual members. groups. I Reduction of moral hazard risks due to group monitor- I Potential weakening of group if group leader departs. ing and peer pressure. I Increased transaction costs to borrowers (time for meet- I Substitution of joint liability for individual collateral. ings and some voluntary management functions). Source: Yaron, Benjamin, and Piprek 1997. funds generally were found to have a high rate of loan port- a nominal amount. For the most part, compulsory sav- folio growth, while deposit-based programs grew more slow- ings can be considered part of a loan product rather than ly (probably due to a lack of funds). Also, institutions that an actual savings product, since they are so closely tied to mobilized deposits were found to have higher average loan receiving and repaying loans. (Of course, for the borrow- sizes and were more likely to work in urban areas than insti- er compulsory savings represent an asset while the loan tutions providing only credit. This latter finding relates to represents a liability; thus the borrower may not view the fact that most institutions that collect savings must be compulsory savings as part of the loan product.) regulated to do so. Larger urban institutions tend to be reg- Compulsory savings are useful to: ulated more often than smaller MFIs, and they often reach a I Demonstrate the value of savings practices to borrowers clientele at a higher income level. I Serve as an additional guarantee mechanism to ensure Bank Rakyat Indonesia and the Bank for Agriculture the repayment of loans and Agricultural Cooperatives in Thailand are well-known I Demonstrate the ability of clients to manage cash flow examples of established MFIs with fast-growing savings and make periodic contributions (important for loan mobilization. The Caisses Villageoises in Pays Dogon, repayment) Mali, provides another less well-known example (box 3.7). I Help to build up the asset base of clients. However, compulsory savings are often perceived COMPULSORY SAVINGS. Compulsory savings differ sub- (rightly) by clients as a "fee" they must pay to participate stantially from voluntary savings. Compulsory savings (or and gain access to credit. Generally, compulsory savings compensating balances) represent funds that must be cannot be withdrawn by members while they have a contributed by borrowers as a condition of receiving a loan outstanding. In this way, savings act as a form of loan, sometimes as a percentage of the loan, sometimes as collateral. Clients are thus not able to use their savings PRODUCTS AND SERVICES 73 Box 3.7 Caisses Villageoises, Pays Dogon, Mali T HE C AISSES V ILLAGEOISES WERE ESTABLISHED IN P AYS Traditionally, farmers prefer to save in livestock or invest Dogon, Mali, in 1996. There are now 55 such self-managed their savings in their small businesses. village banks. The caisses mobilize savings from their mem- Even with high interest rates, relatively few members save bers, who are men and women from the village or neighboring with a caisse. Of the 21,500 members of the Caisses villages. Deposits are used for on-lending to members accord- Villageoises, fewer than 20 percent used a deposit account ing to the decisions made by the village credit committee. during the course of 1996. As a result, the volume of the By December 31, 1996, the Caisses Villageoises had loan applications exceeds the amount of mobilized savings. mobilized $320,000. The average outstanding deposit was The caisses have been borrowing from the National $94, equivalent to 38 percent of GDP per capita. Each vil- Agriculture Bank since 1989 to satisfy their clients' needs. lage bank set its own interest rates based on its experience However, the amount an individual caisse can borrow is a with traditional village groups providing loans to their mem- function of the amount of savings it has mobilized (one and bers or informal sources of financial services. In 1996 the a half to two times this amount can be borrowed from the nominal interest rate on deposits was on average 21 percent;, National Agriculture Bank). In this way, the incentive to with an inflation rate averaging 7 percent, the real rate was mobilize savings remains high for the caisse and members 14 percent. Such a high level of interest was required because feel a significant sense of ownership, since their deposits are the caisses operate in an environment in which money is what determines its growth. As of December 1996 the vol- very scarce, due to minimal production of cash crops in the ume of deposits to the volume of loans outstanding was 41 area and irregular amounts of excess grain for trade. percent. Source: Contributed by Cecile Fruman, Sustainable Banking with the Poor Project, World Bank. until their loan is repaid. In some cases compulsory sav- ent philosophies (CGAP 1997). The former assumes that ings cannot be withdrawn until the borrower actually the poor must be taught to save and that they need to withdraws his or her membership from the MFI. This learn financial discipline. The latter assumes that the sometimes results in the borrowing by clients of loan working poor already save and that what is required are amounts that are in fact less than their accumulated sav- institutions and services appropriate to their needs. ings. However, many MFIs are now beginning to realize Microfinance clients may not feel comfortable putting the unfairness of this practice and are allowing their voluntary savings in compulsory savings accounts or even clients and members to withdraw their compulsory sav- in other accounts with the same MFI. Because they often ings if they do not have a loan outstanding or if a certain cannot withdraw the compulsory savings until their loan amount of savings is still held by the MFI. (Compulsory is repaid (or until after a number of years), they fear that savings are discussed further in chapter 5.) they may also not have easy access to their voluntary sav- ings. Consequently, MFIs should always clearly separate VOLUNTARY SAVINGS. As the name suggests, voluntary sav- compulsory and voluntary savings services. ings are not an obligatory part of accessing credit services. There are three conditions that must exist for an MFI Voluntary savings services are provided to both borrowers to consider mobilizing voluntary savings (CGAP 1997): and nonborrowers who can deposit or withdraw according I An enabling environment, including appropriate legal to their needs. (Although sometimes savers must be mem- and regulatory frameworks, a reasonable level of polit- bers of the MFI, at other times savings are available to the ical stability, and suitable demographic conditions general public.) Interest rates paid range from relatively I Adequate and effective supervisory capabilities to pro- low to slightly higher than those offered by formal finan- tect depositors cial institutions. The provision of savings services offers I Consistently good management of the MFI's funds. advantages such as consumption smoothing for the clients The MFI should be financially solvent with a high and a stable source of funds for the MFI. rate of loan recovery. The requirement of compulsory savings and the Requirements for effective voluntary savings mobi- mobilization of voluntary savings reflect two very differ- lization include (Yaron, Benjamin, and Piprek 1997): 74 MICROFINANCE HANDBOOK I A high level of client confidence in the institution (a cards, and payment services. Many group lending sense of safety) programs offer an insurance or guarantee scheme. A I A positive real deposit interest rate (which will require typical example is Grameen Bank. Each member is positive real on-lending interest rates) required to contribute 1 percent of the loan amount I Flexibility and diversity of savings instruments to an insurance fund. In case of the death of a client I Security this fund is used to repay the loan and provide the I Easy access to deposits for clients deceased client's family with the means to cover bur- I Easy access to the MFI (either through its branch net- ial costs. work or through mobile banking officers) Insurance is a product that will likely be offered I MFI staff incentives linked to savings mobilization. more extensively in the future by MFIs, because there The provision of savings services by an MFI can con- is growing demand among their clients for health or tribute to improved financial intermediation by (Yaron, loan insurance in case of death or loss of assets. The Benjamin, and Piprek 1997): Self-Employed Women's' Association in Gujarat, I Providing clients with a secure place to keep their India, provides a good example of an MFI that is savings, smooth consumption patterns, hedge against meeting the insurance service needs of its clients (box risk, and accumulate assets while earning higher real 3.8). returns than they might by saving in the household or saving in kind Credit Cards and Smart Cards I Enhancing clients' perception of "ownership" of a MFI and thus potentially their commitment to repay- Among the other services that some MFIs are beginning ing loans to the MFI to offer are credit cards and smart cards. I Encouraging the MFI to intensify efforts to collect loans due to market pressures from depositors (espe- CREDIT CARDS. These cards allow borrowers to access a cially if they lose confidence in the MFI) line of credit if and when they need it. Credit cards are I Providing a source of funds for the MFI, which can used when a purchase is made (assuming the supplier of contribute to improved loan outreach, increased the goods accepts the credit card) or when access to cash autonomy from governments and donors, and is desired (the card is sometimes called a "debit card" if reduced dependence on subsidies. the client is accessing his or her own savings). Savings mobilization is not always feasible or desir- The use of credit cards is still very new in the field able for MFIs: the administrative complexities and costs of microfinance; only a few examples are known. Cre- associated with mobilizing savings--especially small dit cards can only be used when adequate infrastructure amounts--may be prohibitive. Institutions may also is in place within the formal financial sector. For exam- find it difficult to comply with prudential regulations ple, some credit cards provide access to cash through that apply to deposit-taking institutions. automated teller machines. If there is not a wide net- Furthermore, the volatility of microfinance loan portfo- work of such machines and if retailers are not willing lios may put deposits at unusually high risk if the MFI uses to accept credit cards, then the cards' usefulness is lim- savings to fund unsafe lending operations. In addition, MFIs ited. that offer voluntary saving services must have greater liquidi- Credit cards do offer considerable advantages to both ty to meet unexpected increases in savings withdrawals. The clients and MFIs. Credit cards can: provision of savings services creates a fundamentally different I Minimize administrative and operating costs institution than one that provides only credit. Caution must I Streamline operations be taken when deciding to introduce savings. I Provide an ongoing line of credit to borrowers, enabling them to supplement their cash flow accord- Insurance ing to their needs. A recent example of an MFI that provides its low- MFIs are beginning to experiment with other fin- income clients with credit cards is the Association for ancial products and services such as insurance, credit the Development of Microenterprises (box 3.9). PRODUCTS AND SERVICES 75 Box 3.8 Self-Employed Women's Association Insurance THE SELF-EMPLOYED WOMEN'S ASSOCIATION (SEWA) IS THE members and helping women to apply for and settle insurance mother institution of a network of interrelated organizations. claims. It began in 1972 as a trade union of self-employed women Every member who wishes to purchase insurance is given in Ahmedabad in the state of Gujarat, India. The association two options: pay a 60 rupee premium each year and receive all is active in three areas: organizing women into unions; estab- benefits or make a one-time fixed deposit of 500 rupees in the lishing alternative economic organizations in the form of association's bank. The interest on the fixed deposit is used to cooperatives; and providing services such as banking, hous- pay the annual premium. For members who opt for the fixed ing, and child care. deposit scheme, the association will also provide them with a The association's bank was established in 1974 to provide 300 rupee maternity benefit paid through its bank. banking and insurance services. It acts as manager of the insur- The bank does not charge its own members or the associa- ance scheme run by the Self-Employed Women's Association tion's members for this service and does not receive any com- for its members. The insurance services are actually offered by mission from the insurance companies. However, in exchange the Life Insurance Corporation of India and the United India for managing these services on behalf of the association, it Insurance Company, supplemented by a special type of insur- receives an annual sum of 100,000 rupees. In addition, the ance for women workers offered by the association. The bank association also pays the salaries of staff responsible for admin- acts as an intermediary, packaging insurance services for its istering the insurance schemes. Source: Biswas and Mahajan 1997. SMART CARDS. The Swazi Business Growth Trust in the Payment Services Kingdom of Swaziland provides its clients with smart cards, which are similar to credit cards but are generally In traditional banks payment services include check not available for use in retail outlets (box 3.10). Smart cashing and check writing privileges for customers who cards contain a memory chip that contains information maintain deposits (Caskey 1994). In this sense the about a client's available line of credit with a lending banks' payment services are bundled with their savings institution. services. MFIs may offer similar payment services either with their savings services (if applicable) or separately for a fee. If payment services are bundled with savings ser- Box 3.9 The Association for the Development of vices, the MFI can pay an artificially low interest rate on Microenterprises' MasterCard customer deposit accounts to cover the cost of those ser- IN MAY 1996 THE ASSOCIATION FOR THE DEVELOPMENT of vices. Otherwise, a fee is charged to cover these costs, Microenterprises (ADEMI) and the Banco Popular which include personnel, infrastructure, and insurance Dominicano, the largest private commercial bank in the costs. Fees can be based on a percentage of the amount Dominican Republic, launched the ADEMI MasterCard. of the check or they can be a flat minimum fee with In addition to the usual uses of the card at purchasing out- additional charges for first-time clients. Moreover, lets, clients can withdraw cash at 45 Banco Popular branches because the MFI advances funds on checks that must and 60 automated teller machines. To qualify for the card subsequently be cleared through the banking system, it clients must have operated their businesses for at least one incurs interest expenses on the funds advanced and runs year, have monthly incomes of at least 3,000 pesos (about US$230), have borrowed at least 3,000 pesos from the risk that some cashed checks will be uncollectible ADEMI, and have a good repayment record. The associa- due to insufficient funds or fraud. MFIs, therefore, must tion aims to reach 5,000 cardholders in the first year. It calls have a relationship with at least one bank to clear the the MasterCard initiative a further step in building trust checks being cashed. with owners of microenterprises and enhancing their status. In addition to check cashing and check writing privi- Source: Contributed by McDonald Benjamin, East Asia and Pacific leges, payment services include the transfer and remit- Region Financial Sector and Development Unit, World Bank. tance of funds from one area to another. Microfinance clients often need transfer services, however, the amounts 76 MICROFINANCE HANDBOOK Box 3.10 Swazi Business Growth Trust Smart Cards SMART CARDS ARE A NEW FINANCIAL TECHNOLOGY PIONEERED provided with the equipment to read the cards, and Swazi by the Growth Trust Corporation, a for-profit affiliate of Business Growth Trust clients can draw funds and make Swazi Business Growth Trust, with the assistance of repayments at various commercial bank branches in Development Alternatives, Inc. and the U.S. Agency for Swaziland. A smart card transaction takes a fraction of the International Development. time of a regular teller transaction. The Growth Trust Commercial banks in Swaziland are not interested in Corporation holds a line of credit with participating banks providing financial services to trust customers due to the and collects transaction information from the banks at the high transaction costs of servicing these small businesses. end of the day. The Growth Trust Corporation was thus established to pro- Smart cards enable the trust to monitor disbursements vide financial services in the form of housing loans and and repayments easily. The trust plans to move the smart small business loans. To minimize administrative and oper- card network on line in commercial banks once the Swazi ating costs, the corporation introduced the smart card. Each telephone system allows. Commercial banks are willing to customer is issued a smart card with a memory chip on its provide this service free of charge because the trust is a surface containing information about the amount of the nonprofit organization and targets a different, low-income line of credit the client has obtained. Commercial banks are clientele. Source: McLean and Gamser 1995. that formal financial institutions require to make a trans- Social Intermediation fer may be beyond the limits of the client. Without transfer services clients may be forced to carry (relatively) For individuals whose social and economic disadvantages large amounts of money with them, thus incurring place them "beyond the frontier" of formal finance (Von unnecessary risks. To offer payment services MFIs must Pischke 1991), successful financial intermediation is have an extensive branch network or relationships with often accompanied by social intermediation. Social inter- one or more banks (box 3.11). This is often difficult to mediation prepares marginalized groups or individuals to achieve. enter into solid business relationships with MFIs.2 Few MFIs are currently offering these services, but Evidence has shown that it is easier to establish sus- it is anticipated that in the future more and more will tainable financial intermediation systems with the poor recognize a demand for payment services and will in societies that encourage cooperative efforts through boost their ability to cover the costs associated with local clubs, temple associations, or work groups--in such services. other words, societies with high levels of social capital. Box 3.11 Demand for Payment Services at the Fédération des Caisses d'Epargne et de Crédit Agricole Mutuel, Benin THE FÉDÉRATION DES CAISSES D'EPARGNE ET DE CRÉDIT and donors who work closely with farmers or poor communi- Agricole Mutuel (FECECAM) has the largest outreach of any ties are also interested, because they would be able to transfer financial institution in Benin, with 64 local branches, many in subsidy payments. remote areas. For this reason there is widespread pressure on For now the federation does not have the management the bank to offer payment services, since its extensive network capability to offer payment services. In addition, Article 24 offers a unique opportunity to transfer funds from the capital of the regulatory framework for credit unions (known as the city, Cotonou, to some very distant villages. The demand for PARMEC or Programme d'appui à la réglementation des payment services comes from a variety of actors. Member Mutuelles d'épargne et de Crédit law) does not allow these clients are interested because this service would help them re- institutions to transfer funds or offer checks which strongly ceive remittances or transfer money to relatives. Government limits the possibility of offering payment services. Source: Fruman 1997. PRODUCTS AND SERVICES 77 Perhaps more than any other economic transaction, financial intermediation without first receiving some financial intermediation depends on social capital, capacity-building assistance. because it depends on trust between the borrower and There is a range of capacity-building that can take place the lender. Where neither traditional systems nor mod- with social intermediation. Some MFIs simply focus on ern institutions provide a basis for trust, financial inter- developing group cohesiveness to ensure that members mediation systems are difficult to establish. continue to guarantee each other's loans and benefit as Social intermediation can thus be understood as the much as possible from networking and the empowerment process of building the human and social capital required that comes from being part of a group. Other MFIs seek for sustainable financial intermediation with the poor. to develop a parallel financial system whereby the group MFIs that provide social intermediation services most actually takes over the financial intermediation process often do so through groups, but some also work with (box 3.12). This often happens with the group first man- individuals. The following discussion focuses primarily on aging its own internal savings funds and then later taking group-based social intermediation, including approaches on the lending risk and banking duties associated with on- that lend directly to groups and those that lend to indi- lending external funds. This aspect of social interme- vidual members of groups. Loosely defined, "groups" diation mostly involves training group members in include small solidarity groups (such as those found in participatory management, accounting, and basic financial ACCION and Grameen lending models) as well as large management skills and helping groups to establish a good credit-unions or caisses villageois. Where relevant, indi- record-keeping system. This often requires initial subsidies vidual-based social intermediation services are also similar to the time-bound subsidies required for financial discussed. intermediation (figure 3.2). Group social intermediation is defined as the effort to Ultimately, it is the groups' cohesiveness and self-man- build the institutional capacity of groups and invest in the agement capacity that enables them to lower the costs of human resources of their members, so that they can begin financial intermediation by reducing default through peer to function more on their own with less help from out- pressure--and consequently to lower the transaction costs side. This aspect of social intermediation responds to the that MFIs incur in dealing with many small borrowers and growing awareness that some of the poor--especially savers (Bennett, Hunte, and Goldberg 1995). those in remote, sparsely populated areas or where levels However, not all social intermediation results in the of social capital are low--are not ready for sustainable formation of groups, nor is it always provided by the Box 3.12 Village Banks: An Example of a Parallel System STARTED IN 1985 AS AN EXPERIMENT IN COMMUNITY DEVEL- The model usually establishes a parallel system managed opment with a program for women's savings and credit in by elected leaders, although there are an increasing number Costa Rica, the village banking model has become an increas- of cases of village banks initiating relationships with formal ingly popular model for providing poor women with access sector financial institutions or federating into apex organiza- to financial services. FINCA, CARE, Catholic Relief Services, tions. The village bank provides a joint liability system for its Freedom From Hunger, and a host of local nongovernmental 25 to 50 members and also acts as a secure savings facility. organizations (NGOs) have begun to implement village Individual members accumulate savings, which eventually banks throughout Latin America, Africa, and Asia. The Small become their own source of investment capital. Enterprise Education and Promotion Network (SEEP) The "rules of the game" for membership--meeting atten- recently reported that there are "over 3,000 banks associated dance, savings, credit, and nonfinancial services--are devel- with SEEP members and their partners serving close to oped by the women themselves, usually with guidance from 34,000 members, 92 percent of whom are women. Average an NGO field worker. Enforcement of credit contracts is the loan sizes are $80 and the collective portfolio stands at responsibility of the leaders, with member support. (For more approximately $8.5 million." information on village banking, see appendix 1.) Source: Bennett 1997. 2. This section is based on writings by Lynn Bennett between 1994 and 1997. 78 MICROFINANCE HANDBOOK Figure 3.2 Group Social Intermediation Source: Bennett 1997. same institution that provides the financial intermedia- Enterprise Development Services tion. In West Africa NGOs or other local organizations (public or private) visit remote villages where access to MFIs adopting an integrated approach often provide financial services is limited to develop awareness among some type of enterprise development services (some MFIs individual village members of available services. They refer to these services as business development services, help individuals open accounts and manage their funds, nonfinancial services, or assistance).3 If these services are and generally increase their level of self-confidence when not offered directly by the MFI, there may be a number dealing with financial institutions. Social intermediation of other public and private institutions that provide enter- services for individuals can be provided by the MFI itself prise development services within the systems framework or by other organizations attempting to link low-income to which the MFI's clients have access. clients with the formal financial sector. Enterprise development services include a wide range Agence de Financement des Initiatives de Base, a of nonfinancial interventions, including: Benin Social Fund, provides a good example (box 3.13). I Marketing and technology services Box 3.13 Social Intermediation in a Social Fund in Benin AGENCE DE FINANCEMENT DES INITIATIVES DE BASE IS A that, it has relied on NGOs that work with small entrepre- social fund in Benin supported by the World Bank and neurs in the most remote villages or poor communities, Deutsche Gesellschaft für Technische Zusammenarbeit. Its helping them to understand the principles according to goal is to help alleviate poverty through infrastructure which the existing MFIs function and then linking them to development, capacity building of communities and the MFI of their choice. The NGOs also convey to the NGOs, and social intermediation. Given the dozens of MFIs the messages they hear in the village, so that the microfinance institutions or providers in Benin, many of MFIs can better adapt their services and products to the which are performing well, the social fund has chosen not specific demand of this target clientele. This role of social to engage in financial intermediation. Instead, it has intermediation played by the NGOs should lead to focused on helping the existing institutions expand their increased financial intermediation for the underserved pop- outreach to the poorer segments of the population. To do ulation in Benin. Source: Contributed by Cecile Fruman, Sustainable Banking with the Poor Project, World Bank 3. This section and appendix 2 were written by Thomas Dichter, Sustainable Banking with the Poor Project, World Bank. PRODUCTS AND SERVICES 79 enterprise development service programs could be divid- Box 3.14 Business Skills Training ed into: I Enterprise formation programs, offering training in FUNDACIÓN CARVAJAL IN COLOMBIA PROVIDES SHORT practical courses in a range of accounting, marketing, and sector-specific skills such as weaving as well as training management topics adapted to microenterprise needs. for persons who might start up such businesses Empretec in Ghana provides a two-tiered manage- I Enterprise transformation programs, providing technical ment training course for microenterprise operators who assistance, training, and technology to help existing exhibit specific entrepreneurial characteristics and are will- microenterprises make a quantitative and qualitative ing to adopt new management attitudes and tools. leap in terms of scale of production and marketing. Source: Contributed by Thomas Dichter, Sustainable Banking An additional distinction is between direct and indi- with the Poor Project, World Bank. rect services, which has to do with the interaction between the provider of services and the enterprises (box 3.15). Direct services are those that bring the I Business training (box 3.14) client in contact with the provider. Indirect services are I Production training those that benefit the client without such direct con- I Subsector analysis and interventions. tact, as in policy-level interventions. The most common providers of enterprise develop- All enterprise development services have as their ment services to date have been: goal the improved performance of the business at I NGOs running microfinance projects with integrated hand, which in turn improves the financial condition approaches of the owner or operator. In a sense, enterprise devel- I Training institutes opment services are a residual or catch-all category-- I Networks they can be almost everything except financial services. I Universities As a result, they are a far broader and more complex I Private firms (training companies, bilateral aid agency arena in which to work than credit and savings (box subcontractors) 3.16). I Producer groups (or chambers of commerce) Some experts suggest that enterprise development I Government agencies (small business assistance, services be provided by a different institution than the export promotion bureaus) one providing financial services. However, if an MFI I Informal networks providing production training chooses to provide such services, they should be clearly through apprenticeships. separated from other activities and accounted for sepa- Studies sponsored by the U.S. Agency for rately. Furthermore, enterprise development services International Development in the late 1980s found that Box 3.16 Cattle Dealers in Mali Box 3.15 Direct and Indirect Business Advisory Services IN PAYS DOGON, MALI, AN MFI PROVIDING ENTERPRISE development services helped to organize a gathering with THE BUSINESS WOMEN'S ASSOCIATION OF SWAZILAND cattle dealers, where they identified their main need as a visits clients' workshops on a regular basis to provide basic reliable means of transportation to move cattle from the vil- management advice for a fee (direct services). lages to the towns. They decided to call a second meeting Alternatively, the carpenters association in Oua- and invite some of the local truck owners to discuss the gadougou, Burkina Faso, developed a promotional problems and identify solutions. As a result, the cattle deal- brochure to increase the number and size of contracts with ers started sharing truck loads to reduce costs. They also small and medium-size firms and government agencies began to pressure the truck owners to provide timely and (indirect services). high-quality transportation services. Source: Contributed by Thomas Dichter, Sustainable Banking Source: Contributed by Cecile Fruman, Sustainable Banking with with the Poor Project, World Bank. the Poor Project, World Bank. 80 MICROFINANCE HANDBOOK should never be a requirement for obtaining financial with little impact on the businesses; or they have pro- services. vided closely tailored assistance to a few enterprises, They should also be fee-based. While full cost recov- at high cost per beneficiary." ery may not be possible, the goal should be to generate I Outreach. A common complaint about enterprise as much revenue as possible to cover the costs of deliver- development services is that women's access to ing the services. This is why some traditional providers these services is limited. A study by the United (such as government agencies) have been in this arena-- Nations Development Fund for Women in 1993 they know that enterprise development services must be found that only 11 percent of government-spon- subsidized. The services will almost always involve some sored business training slots went to women, due type of subsidy, which has to be external to the enter- to time conflicts between their work and house- prises unless it is provided by groups of enterprises that hold responsibilities, the location (often distant devote a portion of their profits to keeping the services from the village or slum), and bias in some formal going. Even private firms (or bilateral aid agency sub- training institutions against working with women contractors) and universities usually provide enterprise clients. development services under grants from foundations The main response to these doubts is that enterprise and governments. development services approaches are highly varied, and Direct services to enterprises (such as entrepreneur- to date only a few of many possibilities have been tried. ship and business skills training), the organization of It is premature to conclude that these services have little microenterprises into clusters, and so on require sensitivi- impact or have inherent limits in terms of outreach. Nor ty on the part of the provider to the circumstances of are there enough cases in which enterprise development poor people, a capacity to stick with and work for long services are provided exclusively rather than as part of an periods in poor communities, and dedication to helping integrated package. However, they are likely most valu- the poor. These characteristics are usually found among able for larger microenterprises, since smaller ones are people who work in NGOs. often engaged in traditional activities that people know In general, enterprise development services are some- how to operate already. There is a high cost involved in what problematic, because even though the payoff can be providing enterprise development services directly to great, it is difficult to judge either the impact or the per- individual enterprises, but if the assisted business grows formance of the service provider itself. Enterprise devel- and creates significant employment, the long-term pay- opment services projects do not generally involve money offs can outweigh the initial cost. The dissenting views as a commodity, but rather the transfer of knowledge in mentioned above refer only to the direct provision of the form of skills, information, research, and analysis. enterprise development services and do not take into The returns to the client as a result of this new knowl- account the potential high returns accruing to indirect edge are difficult to assess. Likewise, because the client provision. cannot measure the value of "knowledge" at the "point of There has been little documentation of the results purchase," paying its full cost to the provider is rarely and outcomes of enterprise development services possible. because these services often have been part of inte- Certain studies have questioned the value of enter- grated approaches. And unlike financial services, the prise development services in the following areas: wide range of approaches and strategies possible I Impact. Some studies have shown that in many cases under the rubric of enterprise development services there is little difference in profits and efficiency means that there are fewer agreed techniques to be between those microenterprises that received credit learned and adapted by MFIs wishing to take this alone and those that received an integrated enterprise approach. development services and credit package. Whether an MFI provides enterprise development I Cost and impact. According to Rhyne and Holt services depends on its goals, vision, and ability to (1994), "these [enterprise development service] pro- attract donor funds to subsidize the costs. (See appendix grams tend to fall into one of two patterns. Either 2 for more information about enterprise development they have provided generic services to large numbers, services.) PRODUCTS AND SERVICES 81 Social Services tional NGO), or a community organization, some MFIs have chosen to provide social services in addition to While social services such as health, nutrition, education, financial intermediation (box 3.17). In this way, they are and literacy training are often provided by the state, a able to take advantage of contact with clients during loan local NGO (with or without assistance from an interna- disbursement and repayment. Box 3.17 Who Offers Social Services? A RECENT WORLD BANK SURVEY EXAMINED THE ARRAY of Ratio of social to financial staff financial and nonfinancial services offered by more than 1.4 200 microfinance institutions around the world. Analysis of the types of institutions that offer social 1.2 services showed that nongovernmental organizations 1.0 (NGOs) were more active in providing services and train- ing relating to health, nutrition, education, and group 0.8 formation. This trend was particularly pronounced in 0.6 Asia. Credit unions tended to have more of a financial focus than NGOs did but were also quite involved in lit- 0.4 eracy and vocational training. Banks and savings banks were more narrowly focused on financial services, as 0.2 shown in the ratio of social staff to financial staff in the 0.0 figure. Savings Credit Banks banks unions NGOs Source: Paxton 1996a. Box 3.18 Freedom From Hunger--Providing Nutrition Training with Microfinance F REEDOM F ROM H UNGER IS AN INTERNATIONAL NON - for educational as well as financial purposes. The organi- governmental organization (NGO) based in Davis, zation believes that group-based lending models have California, with operations in Latin America, Africa, and social as well as economic potential and that this dual Asia. It believes that microfinance services are unlikely to potential must be realized to have an impact on poverty. relieve most of the burdens of poverty without important By helping the poor to manage their own self-help groups economic, social, medical, and nutritional changes in par- successfully and use credit to increase their incomes and ticipants' behavior. begin saving, these programs engage in vital activities that Their basic premise is that very poor people--especial- improve confidence, self-esteem, and control over one's ly women in rural areas--face tremendous obstacles to environment. becoming good financial services clients. The reasons In credit with education programs, a part of each reg- include social isolation, lack of self-confidence, limited ular meeting is set aside for a learning session. Initially, a entrepreneurial experience, and major health and nutri- member of the program staff (usually a moderately edu- tional problems. When the goal is to provide credit for cated person from the local area) facilitates each learning the alleviation of an entire family's poverty, these clients session. Village bank members identify the poverty-relat- need more than credit alone. They need educational ser- ed problems they confront in their daily lives and become vices that specifically address these obstacles and that pro- motivated to develop and use solutions that are appropri- vide a framework for solutions. ate to the location. The topics addressed in learning ses- Freedom From Hunger utilizes the "credit with educa- sions range from village bank management to the basic tion" village banking model--group-based lending to the economics of microenterprise to the improvement of the poor that cost effectively uses the regular group meetings health and nutrition of women and children. (Box continues on next page.) 82 MICROFINANCE HANDBOOK Box 3.18 Freedom From Hunger--Providing Nutrition Training with Microfinance (Continued) Freedom From Hunger is committed to achieving full education services to the same village bank. These practi- recovery of operating and financial costs with interest rev- tioners limit the number, scope, and time allocation for enue from loans to village banks. However, Freedom From topics to be addressed, maintaining that the marginal cost Hunger and other credit with education practitioners of education is low enough for reasonable interest charges to (Catholic Relief Services, Katalysis, Plan International, and cover the costs of both financial and educational services. World Relief, among others) differ in their attitudes toward They also believe that the marginal cost of education can be covering the costs of education with revenue from credit legitimately charged to the village banks, because it makes operations. Some have separate field agents to provide the members better financial service clients by improving financial services and educational services to the same village their family survival skills, productivity, and management bank. They regard education as a social service that can and skills. should be subsidized by grants or government taxes and Practitioners of credit with education believe that they also believe that the skills required for delivering effec- there may also be a cost to not providing education, such tive education and financial services are too many and too as epidemic outbreaks, loan growth stagnation, and mem- varied for a single field agent. Other practitioners try to ber dropout due to illness of children. Comparison stud- minimize the marginal cost of education by training and ies are currently going on to determine the impact of supporting each field agent to provide both financial and credit with education programs. Source: Freedom From Hunger 1996. As in the case of enterprise development services, the I Latin American solidarity group lending delivery and management of social services should be kept I Village banking as distinct as possible from the delivery and management I Self-reliant village banks. of financial intermediation services. This does not mean While microfinance approaches generally include that social services cannot be provided during group meet- some specific product design issues, a primary means of ings, but they must be clearly identified as separate from differentiating one approach from another is in the credit and savings services. Furthermore, it is not reason- choice of products and services provided and the man- able to expect that revenue generated from financial inter- ner in which provision is made. Any approach must be mediation will always cover the costs of delivering social based on the target market and its demand for financial services. Rather, the delivery of social services will most intermediation, as well as other products, contextual fac- likely require ongoing subsidies. MFIs that choose to pro- tors in the country, and the objectives and institutional vide social services must be clear about the costs incurred structure of the MFI. These approaches do not operate and must ensure that the donors supporting the MFI in isolation. Many of them rely on other institutions understand the implications of providing these services. (either public or private) to provide additional services Freedom From Hunger provides an example of an so that the targeted clients can effectively utilize finan- international NGO that provides social services in conjunc- cial services. tion with financial intermediation (box 3.18). The Grameen Trust supports replicators of the Grameen Bank model around the world with funding and technical assistance. However, the Grameen Trust will Appendix 1. Microfinance Approaches only support MFIs that replicate exactly the model for their first two years as it is implemented in Bangladesh. The following pages provide information on some of the Grameen believes that their model is successful because of most well-known microfinance approaches. The all of its elements and that if some are rejected or changed, approaches presented are: the MFI might fail. After two years of operations replica- I Individual lending tors are allowed to introduce changes if they feel that some I Grameen Bank solidarity lending elements must be adapted to the local context. PRODUCTS AND SERVICES 83 Interest rates are higher than formal sector loans but Box A3.1.1 Training for Replicators in West Africa lower than informal sector loans. Most MFIs require IN WEST AFRICA AN INNOVATIVE TRAINING CYCLE HAS some form of collateral or cosignatories. Compulsory sav- been set up for African replicators (Itinéraire de Formation- ings usually are not required. Diffusion). The total cycle lasts for seven months, during Detailed financial analysis and projections are often which time replicators receive a month of formal training included with the loan application. The amount and and spend four months on three different MFI sites, where terms are negotiated with the client and the credit offi- they participate fully in the institution's activities. During cer's supervisor or other credit officers. Documentation the last three months, the trainees design their own MFI, is required, including a loan contract; details regarding drawing on the experience they have acquired in the field the clients references; if applicable, a form signed by the and the knowledge they have of their home market. With cosigner and his or her personal information; and legal such a process the newly established MFIs can become a deeds to assets being pledged and credit history. Credit synthesis of several successful programs. After the initial officers are often recruited from the community so that training the replicators receive ongoing technical assistance and follow-up from the model programs. they can base their analysis on their knowledge of the client's creditworthiness (character-based lending). Source: Contributed by Cecile Fruman, Sustainable Banking with The loan is usually disbursed at the branch office. A the Poor Project, World Bank. visit is often made to the client's place of business to verify that the client has made the purchases specified in the loan contract. Periodic payments are made at the Other practitioners are strongly in favor of adapta- branch or through preapproved payments. tion. They believe that there is no single model appro- priate for all situations and that every model must be P R O D U C T S . Loan sizes can vary from US$100 to adapted to the local context to fit local needs. This US$3,000 with terms between six months and five requires that MFIs be flexible and creative. ACCION years. Savings may or may not be provided depending has been successful in adapting their solidarity group on the institutional structure of the MFI. Training and lending model to varying contexts. technical assistance may be provided by credit officers; sometimes training is provided on a per-fee-basis or is Individual Lending mandatory. Individual lending is defined as the provision of credit to SIGNIFICANT EXAMPLES. These include ADEMI in the individuals who are not members of a group that is joint- Dominican Republic; Caja Municipales in Peru; Bank ly responsible for loan repayment. Individual lending Rakyat Indonesia; Agence de Credit pour l'Enterprise requires frequent and close contact with individual Privée in Senegal; Alexandria Business Association in clients to provide credit products tailored to the specific Egypt; Self-Employed Women's Association in India; needs of the business. It is most successful for larger, Fédération des Caisses d'Epargne et de Crédit Agricole urban-based, production-oriented businesses and for Mutuel in Benin; and Caja Social in Colombia. clients who have some form of collateral or a willing cosigner. In rural areas, individual lending can also be APPROPRIATE CLIENTS. Clients are urban enterprises or successful with small farmers. small farmers, including both men and women, and may be medium-income small businesses, microbusi- METHOD. Clients are individuals working in the informal nesses, and production enterprises. sector who need working capital and credit for fixed assets. Credit officers usually work with a relatively small Grameen Solidarity Group Lending number of clients (between 60 and 140) and develop close relationships with them over the years, often pro- This lending model was developed by the Grameen viding minimal technical assistance. Loan amounts and Bank of Bangladesh to serve rural, landless women wish- terms are based on careful analysis by the credit officer. ing to finance income-generating activities. This model 84 MICROFINANCE HANDBOOK is prevalent mostly in Asia but has been replicated in Latin American Solidarity Group Lending other contexts. Grameen Trust has more than 40 repli- cators in Asia, Africa, and Latin America. The solidarity group lending model makes loans to individual members in groups of four to seven. The METHOD. Peer groups of five unrelated members are members cross-guarantee each other's loans to replace self-formed and incorporated into village "centers" of traditional collateral. Clients are commonly female up to eight peer groups. Attendance at weekly meetings market vendors who receive very small, short-term and weekly savings contributions, group fund contribu- working capital loans. This model was developed by tions, and insurance payments are mandatory. Savings ACCION International in Latin America and has been must be contributed for four to eight weeks prior to adapted by many MFIs. receiving a loan and must continue for the duration of the loan term. The group fund is managed by the group METHOD. Customers are typically informal sector microbusi- and may be lent out within the group. Group members nesses, such as merchants or traders who need small mutually guarantee each other's loans and are held amounts of working capital. Group members collectively legally responsible for repayment by other members. guarantee loan repayment, and access to subsequent loans is No further loans are available if all members do not dependent on successful repayment by all group members. repay their loans on time. No collateral is required. Payments are made weekly at the program office. The Mandatory weekly meetings include self-esteem build- model also incorporates minimal technical assistance to the ing activities and discipline enforcement. borrowers, such as training and organization building. Loans are made to individuals within the group by Credit officers generally work with between 200 and 400 the local credit officer at the weekly meetings. However, clients and do not normally get to know their clients very only two members receive loans initially. After a period well. Loan approval is by the credit officers based on mini- of successful repayment, two more members receive mal economic analysis of each loan request. Loan disburse- loans. The final member receives her loan after another ment is made to the group leader at the branch office, who period of successful repayment. Grameen typically pro- immediately distributes to each individual member. Credit vides precredit orientation but minimal technical assis- officers make brief, occasional visits to individual clients. tance. Loan appraisal is performed by group members Group members normally receive equal loan amounts, and center leaders. Branch staff verify information and with some flexibility provided for subsequent loans. Loan make periodic visits to client businesses. Credit officers amounts and terms are gradually increased once clients usually carry between 200 and 300 clients. demonstrate the ability to take on larger amounts of debt. Loan applications are simple and are reviewed quickly. PRODUCTS. Credit is for six months to one year and Savings are usually required but are often deducted from payments are made weekly. Loan amounts are usually the loan amount at the time of disbursement rather than from US$100 to US$300. Interest rates are 20 percent requiring the clients to save prior to receiving a loan. a year. Savings are compulsory. Savings serve primarily as a compensating balance, guar- anteeing a portion of the loan amount. SIGNIFICANT EXAMPLES. These include Grameen Bank and Bangladesh Rural Advancement Committee in PRODUCTS. Initial loan amounts are generally between Bangladesh; Tulay sa Pag-Unlad, Inc. and Project US$100 and US$200. Subsequent loans have no upper Dungganon in the Philippines; Sahel Action in Burkina limit. Interest rates are often quite high and service fees Faso; and Vietnam Women's Union. are also charged. Savings are usually required as a por- tion of the loan; some institutions encourage establish- APPROPRIATE CLIENTELE. Clients are from rural or urban ing intragroup emergency funds to serve as a safety net. (densely populated) areas and are usually (although not Very few voluntary savings products are offered. exclusively) women from low-income groups (means tests are applied to ensure outreach to the very poor) SIGNIFICANT EXAMPLES. These include ACCION affili- pursuing income-generating activities. ates: PRODEM, BancoSol Bolivia; Asociación Grupos PRODUCTS AND SERVICES 85 Solidarios de Colombia; and Genesis and PROSEM in Regular weekly or monthly meetings are held to collect Guatemala. savings deposits, disburse loans, attend to administra- tive issues, and, if applicable, continue training with the APPROPRIATE CLIENTELE. Clients are mostly urban and MFI officer. include both men and women who have small to medi- um incomes (microbusinesses, merchants, or traders). PRODUCTS. Members' savings are tied to loan amounts and are used to finance new loans or collective income- Village Banking generating activities. No interest is paid on savings. However, members receive a share from the bank's Village banks are community-managed credit and sav- relending or investment profits. The dividend distributed ings associations established to provide access to financial is directly proportional to the amount of savings each services in rural areas, build a community self-help individual has contributed to the bank. group, and help members accumulate savings (Otero and Loans have commercial rates of interest (1 to 3 per- Rhyne 1994). The model was developed in the mid- cent per month) and higher rates if from an internal 1980s by the Foundation for International Community account. Some banks have broadened service delivery to Assistance (FINCA). Membership in a village bank usu- include education about agricultural innovations, nutri- ally ranges from 30 to 50 people, most of whom are tion, and health. women. Membership is based on self-selection. The bank is financed by internal mobilization of members' funds as S IGNIFICANT EXAMPLES . These include FINCA in well as loans provided by the MFI. Mexico and Costa Rica; CARE in Guatemala; Save The Children in El Salvador; Freedom From Hunger in METHOD. A village bank consists of its membership and Thailand, Burkina Faso, Bolivia, Mali, and Ghana; and a management committee, which receives training from Catholic Relief Services in Thailand and Benin. The the sponsoring MFI. The sponsoring MFI lends seed original model has been adapted in a variety of ways. In capital (external account) to the bank, which then lends FINCA in Costa Rica committee members take on the on the money to its members. All members sign the tasks of bank teller and manager. Catholic Relief loan agreement to offer a collective guarantee. The loan Services works through local NGOs. Freedom From amount to the village bank is based on an aggregate of Hunger in West Africa works directly with credit unions all individual members' loan requests. Although the in order to help them increase their membership among amount varies between countries, first loans are typical- women. Its clients graduate to the credit union. ly short term (four to six months) and are small amounts ($50), to be repaid in weekly installments. APPROPRIATE CLIENTELE. Clients are usually from rural or The amount of the second loan is determined by the sparsely populated but sufficiently cohesive areas. They savings a member has accumulated during the first loan have very low incomes but with savings capacity, and are period through weekly contributions. The methodology predominantly women (although the program is also anticipates that the members will save a minimum of adequate for men or mixed groups). 20 percent of the loan amount per cycle (internal account). Loans from the internal account (member Self-Reliant Village Banks (Savings and Loans savings, interest earnings) set their own terms, which Associations) are generally shorter, and their own interest rates, which are generally much higher. Loans to the village Self-reliant village banks are established and managed by banks are generally provided in a series of fixed cycles, rural village communities. They differ from village usually 10 to 12 months each, with lump-sum pay- banks in that they cater to the needs of the village as a ments at the end of each cycle. Subsequent loan whole, not just a group of 30 to 50 people. This model amounts are linked to the aggregate amount saved by was developed by a French NGO, the Centre for individual bank members. Village banks have a high International Development and Research, in the mid- degree of democratic control and independence. 1980s. 86 MICROFINANCE HANDBOOK M ETHOD . The supporting program identifies villages Appendix 2. Matching Enterprise Development where social cohesion is strong and the desire to set up a Services to Demand village bank is clearly expressed. The villagers--men and women together--determine the organization and rules Designing appropriate enterprise development services of their bank. They elect a management and credit com- requires an understanding of the many systems (cultural, mittee and two or three managers. Self-reliant village legal, political, macroeconomic, local, marketplace) within banks mobilize savings and extend short-term loans to which microenterprises operate. Microentrepreneurs and villagers on an individual basis. The sponsoring program small business operators may be less aware than their larger does not provide lines of credit. The bank must rely on counterparts are of the degree to which those systems can its savings mobilization. impinge upon their business activity. After a year or two the village banks build up an One can visualize these contexts as a series of concentric informal network or association in which they discuss circles, with the person operating the business at the center current issues and try to solve their difficulties. The asso- (figure A3.2.1). ciation acts as intermediary and negotiates lines of credit with local banks, usually an agriculture development Circle 1. The Business Operator bank. This links the village banks to the formal financial sector. Because management is highly decentralized, The enterprise owner's self is the first context in which he central services are limited to internal control and audit- or she operates. Each individual brings his or her own cir- ing, specific training, and representation. These services cumstances, energies, character, skills, and ideas to the are paid for by the village banks, which guarantees the enterprise. At this level, constraints on the success of the financial sustainability of the model. business often have to do with intangibles such as the pres- ence of or lack of confidence, tenacity, resourcefulness, PRODUCTS. These include savings, current accounts, and ambition, adaptability, and the capacity to learn. Likewise, term deposits. Loans are short-term, working-capital the business operator's level of basic skills (literacy and loans. There is no direct link between loan amounts and numeracy) and their applications in the business (book- a member's savings capacity; interest rates are set by each village according to its experience with traditional sav- ings and loans associations. The more remote the area, Figure A3.2.1 The Entrepreneur's Context the higher the interest rate tends to be, because the opportunity cost of money is high. Loans are paid in one installment. Loans are individual and collateral is Beyond the local marketplace necessary, but above all it is village trust and social pres- The subsector sure that ensure high repayment rates. Management committees, managers, and members all receive exten- The business sive training. Some programs also provide technical assistance to microentrepreneurs who are starting up a business. The business operator S I G N I F I C A N T E X A M P L E S . These include Caisses Villageoises d'Epargne et de Crédit Autogérées in Mali (Pays Dogon), Burkina Faso, Madagascar, The Gambia (Village Savings and Credit Associations or VISACA), São Tomé, and Cameroon. APPROPRIATE CLIENTELE. Clients are in rural areas and include both men and women with low to medium Source: Author's schematic. incomes and some savings capacity. PRODUCTS AND SERVICES 87 keeping, being able to read the maintenance instructions At this level the service provider has to understand what on a machine) will obviously make a difference. Some of the product or service is, how it is produced, what tech- these needs are addressed through the provision of social nology is used, what human labor is involved, and what intermediation or social services. the inputs are. For example, if a tailor requires zippers or Beyond the entrepreneur's self is the realm of knowl- buttons, how are they supplied and how are they inven- edge and skills that are directly related to the conduct of toried, if at all? If storage is required, how is that made business. On the cusp of the second circle (the business available? Is time being lost because regular machinery itself) the emphasis is still on the knowledge possessed by maintenance is not scheduled properly or because parts the operator. Business skills training aimed at imparting that wear out on a predictable basis are not stocked in basics such as bookkeeping and record keeping usually anticipation of need? What about the employees or other involves fairly generic skills and thus lends itself to class- human input that the business depends on? Are they room teaching. family members? Are they paid? How are they paid? Do For start-up businesses an enterprise development ser- they stay on? Or do they leave to become competitors? vice provider can introduce concepts of the market and Can production be improved by changing the flow courses that provide opportunities for practice in of work itself or by changing the physical organization bookkeeping and accounting, legal issues (registration), of the work space? Is there waste that can be eliminated business planning, marketing, and pricing, as well as cus- or reduced? Would changes in the timing of production tomer orientation. or selling make a positive difference in the owner's bot- A program could, for example, help a woman shop- tom line? keeper devise an inventory system and a simple worksheet Many changes that can be made within the enterprise to monitor operating costs. She also could maintain a need to be brought to the attention of the entrepreneur book that lists those products her clients ask for that she by an outsider. That outsider can be an "expert" or advi- does not currently sell. sor working for the enterprise development service It is very important to know which entrepreneurs have provider or a colleague in the same type of business. a chance to become successful. Accordingly, some tests of The many methods used in enterprise development ser- commitment and ambition can be imposed. One meth- vices at the business level range from on-site visits by a reg- ods is to have people pay or agree to pay later a portion of ular business advisor to the creation of local walk-in their profits to the service provider. Another is to not offer business advisory service offices and enterprise clusters or the service until a certain level of business activity has industry networks. They may, even in the case of literate been reached. urban entrepreneurs, include advice that comes in printed It is also important to differentiate between skills that form or that entrepreneurs can learn themselves, for exam- can successfully be imparted by teaching (the generic ple, by going through a stock set of questions that provide business skills discussed above) and those that need to be them with the equivalent of an "outsider's viewpoint." imparted one-on-one in the business setting (see below). The more one moves away from one-on-one direct enterprise development services to indirect service provi- Circle 2. The Business Itself sion the more likely it is that costs per client will go down. But in the case of indirect service provision, outcomes sim- The second concentric circle surrounding the entrepre- ilar to those possible with direct service cannot be auto- neur is the business itself. This is where all operations matically achieved. Indirect services at the business level take place in the enterprise, from buying inputs to mak- can also include industrial estates and business incubators. ing and selling the product or service. At this level the These can be set up on a commodity subsector basis, that problems and constraints encountered begin to have as is, by concentrating enterprises that are in the same or much to do with forces external to the operator as they related subsector (for example, woodworking) or that share do with the operator's knowledge. the trait of being start-ups. The first benefit is that start- Enterprise development services provided directly to ups share the costs of receiving subsidies to cover the costs the business tend to be less generic in nature and more of infrastructure and basic services such as electricity, custom-tailored. Thus the cost of providing service rises. water, storage space, and machinery. By concentrating 88 MICROFINANCE HANDBOOK businesses in the same place, enterprises also benefit from The subsector approach emphasizes (Rhyne and Holt external economies--the emergence of suppliers who pro- 1994, 33): vide raw materials and components at a bulk price, new I Market access. The subsector approach aims to identi- and second hand machinery and parts, and the emergence fy both the growing portions of subsector markets and of or access to a pool of workers with sector-specific skills. the barriers that must be overcome if small enterprises Sometimes the industrial estate or business incubator are to gain access to those growth markets. provides intangible but equally important benefits to I Leverage. The approach aims for cost effectiveness by start-ups by enabling the entrepreneurs to speak with each identifying interventions that can affect large numbers other informally. This often builds confidence and fosters of enterprises at one time, often through indirect the discovery of new solutions to problems. More tangi- rather than one-on-one mechanisms. bly, as start-ups mature and business grows, subcontract- I Releasing constraints. Many of the most important con- ing arrangements within the incubator often develop. straints it addresses are highly industry specific and Unfortunately, industrial estates and business incuba- thus are missed by "generic" enterprise development. tors have often failed. Many microenterprise operators Subsector analysis requires versatility in design and try to avoid regulation and taxation, and they often see flexible access to a wide range of skills. Institutions that the visibility involved in moving to an industrial estate or identify subsector analysis as a means of intervention will incubator as disincentives. Microenterprises often depend need to consider many issues. on walk-in clients, so moving out of low-income neigh- Common constraints at the subsector level are: borhoods can mean losing customers. These are prob- I Rising competition, as low barriers to entry allow new lems that can be solved by paying more attention to the microenterprises into the marketplace location of such incubators and estates. Overall, the I Space for the market's physical infrastructure, includ- problems with these failed efforts often have had more to ing warehousing do with the fact that they have been started by govern- I The imposition of fees or solicitation of bribes by ment agencies without careful preparation, proper incen- local authorities tives for government workers, or the participation of the I Transport availability, reliability, and cost eventual users in their planning. I Input availability, if a commodity that the enterprise needs for production is not available regularly Circle 3. The Subsector I Lack of appropriate technology I Limits (permanent, cyclical, seasonal) in customer All businesses operate in the context of other businesses purchasing power in the same subsector and sellers and suppliers of inputs I Lack of product diversity and services to the subsector. A subsector is identified I Dominance by middlemen. by its final product and includes all firms engaged in For example, the structure of ownership of transport the supply of raw materials, production, and distribu- may be something a group of enterprises could correct if tion of the product (Haggeblade and Gamser 1991). assisted by an enterprise development service provider. A Subsector analysis describes the economic system of the U.S. NGO working in rural Zaire in the mid-1980s businesses related to the final product and involves the helped local farmers pool resources to form a shareholder- provision of both financial and nonfinancial services. A owned trucking company, which served producers who subsector is defined as the full array of businesses were unable to make money because they faced exorbitant involved in making and selling a product or group of transport fees. The service provider undertook the legal products, from initial input suppliers to final retailers. work involved in setting up this company and helped to Typical areas of intervention include technology devel- put together the finance plan for the first vehicle. opment, skill training, collective marketing or purchas- Often the way things are purchased is a constraint ing of inputs, and, in some cases, policy advocacy on efficiency. For example, in the second-hand clothing (Rhyne and Holt 1994). Examples of subsectors that business in much of Africa the quality of the bales MFIs have developed are dairy, wool, sericulture, palm varies randomly. The buyer does not know what is in oil, and fisheries. the bale and cannot look inside before purchasing it. PRODUCTS AND SERVICES 89 Accordingly, the buyer is often reluctant to pool-pur- works can change the parameters of the market. chase bales with other sellers. However, if an enterprise Encouraging entrepreneurs to work together and to development service provider were to come up with a combine strengths is essential for the development of way of decreasing this risk (for example, by pooling microenterprises. Cooperation between firms, mutual purchases and then creating a secondary distribution learning, and collective innovation can often expand a channel whereby the contents of each repackaged bale microenterprise's chances of increasing its market share would be known), this constraint could be overcome. and profits. Some examples: Inputs are often a major problem, as are parts or the I Sharing information on markets or suppliers sourcing of equipment. People may not know of avail- I Setting up joint cooperatives able alternative sources. I Bidding on contracts together Technology itself is often a constraint. New technolo- I Sharing costs of holding a stand at a fair or market gies can be developed that are custom-tailored to a specif- I Reducing costs of warehousing by splitting a facility. ic level of the subsector. Pricing can also be a constraint. In some subsectors, particularly retail petty trade in the M ARKET INFRASTRUCTURE IMPROVEMENT . The World urban developing world, price can be the only feature dif- Bank's Southwest China Poverty Reduction Project ferentiating adjacent sellers (personality traits of the sellers includes construction of farmers' markets to improve the excepted). The goods are likely to be the same, and as one links between low-income household producers in poor seller is literally six inches away from the next, location is counties in mountainous southwest China and transport not a factor. And yet price itself cannot vary much, if at companies from booming markets on the south China all. By studying the supply of the commodity traded coast. (shoes, for example) and the nature of its wholesaling, it may be possible to form a buyers' coop through which Circle 4. Beyond the Local Marketplace petty traders can reduce the price they pay and thus pass on that savings to the consumer (although this competi- The last of the concentric circles surrounding the indi- tive advantage may be short-lived). vidual microentrepreneur is the largest and furthest away The timing of production of some commodities from his or her enterprise. Here at the macro level the affects price. Retail traders in fresh foods are often at the forces exerting constraints on the enterprise are indeed mercy of the agricultural production cycle and do not large. These can be at the country level, the regional know how to balance the relationship between their sup- level, and even the world trade level. They extend ply and the customers' demand. Storage or value added beyond the marketplace and economic forces in general (such as through drying) give the seller an edge on price. to include: I National social and political forces (corruption, insta- EXPANSION OF THE RANGE OF SUPPLIERS AND bility, tribal rivalries, the admission or exclusion of CUSTOMERS. This can be done by clustering and net- nonnative traders, investors, and suppliers) and working as well as by creating intermediaries or associa- geopolitical forces (for example, the end of the influ- tions. Indirect options in which the enterprise ence of cold war politics on Africa). development service provider can help are the organiza- I Regulatory and policy forces and constraints. Laws tion of trade fairs, collective promotional efforts, trade may work for or against a sector (or subsector) or for coops, showroom and exhibition space, industrial only one channel in a particular sector and against estates, business incubation, and so forth. Pooling another channel in the same sector. Likewise, laws employees or temporarily exchanging employees can affect the way in which lenders operate (the presence loosen constraints in the labor market, and enterprise of a thorough and reliable collateral law will affect the development service providers can act as brokers for nature of lending throughout the financial sector). subcontracting between and among enterprises. Licensing requirements for businesses and taxation can be important constraints. NETWORKING AND ENTREPRENEURS CLUBS. Cross-visits I Macro-level market forces, such as the dynamism or and other forms of network clustering or industrial net- weakness of other exporters of the same commodity 90 MICROFINANCE HANDBOOK is also difficult and requires not just technical ability but Box A3.2.1 Policy Dialogue in El Salvador often political skill as well. IN THE MID-1980S A U.S. NONGOVERNMENTAL ORGANIZA- Although there have been very few enterprise devel- tion working with subsector studies in El Salvador identi- opment service providers who have operated at this fied a government policy that inadvertently created import level of intervention, the payoffs and multiplier effects competition to rural small enterprises making henequen can, in theory, be great. A service provider working in fibre rope. The government simply was not aware that this way is clearly providing an indirect service to enter- domestic needs for cheap rope could be met with domestic prises, which the individual entrepreneur and even production. After this study was brought to the attention of groups of individuals banding together cannot provide the government through "policy dialogue," import regula- for themselves. tions were changed to favor local producers. The initially The enterprise development service provider can high cost of undertaking the studies proved to be amortized undertake studies of the sector to identify hidden con- quickly once the lot of many local producers improved. straints that are invisible to the individual entrepreneur or Source: Contributed by Thomas Dichter, Sustainable Banking that the entrepreneur knows about but cannot change with the Poor Project, World Bank. without help. in other countries (for example, Ghana was a major Sources and Further Reading exporter of palm oil until surpassed by Malaysia) or the availability of uniform containers in a country for Bennett, Lynn. 1993. "Developing Sustainable Financial Systems the micro-level retailers (in Kenya, for example, the for the Poor: Where Subsidies Can Help and Where They production and selling of honey in the late 1980s was Can Hurt." Talking notes for the World Bank AGRAP negatively affected by the lack of readily available, Seminar on Rural Finance, May 26, Washington, D.C. ------. 1994. "The Necessity--and the Dangers--of uniformly sized small containers). Combining Social and Financial Intermediation to Reach I National and global labor market forces, including the Poor." Paper presented at a conference on Financial changes in migrant labor patterns (for example, Services and the Poor at the Brookings Institution, Filipinos are no longer going to Saudi Arabia to work, September 28­30, Washington, D.C. thus changing the nature of competition in the start-up ------. 1997. "A Systems Approach to Social and Financial microenterprise arena in the Philippines); the nature of Intermediation with the Poor." Paper presented at the the workers' prior education, especially in technical Banking with the Poor Network/World Bank Asia Regional skill areas; or the influence of cultural factors on peo- Conference on Sustainable Banking with the Poor, ple's willingness to take certain kinds of jobs. November 3­7, Bangkok. I Contiguous financial sector constraints and enablers, Bennett, L., P. Hunte, and M. Goldberg. 1995. "Group- such as a viable insurance industry that can price Based Financial Systems: Exploring the Links between insurance at a rate affordable for small businesses Performance and Participation." World Bank, Asia Technical Department, Washington, D.C. investing in capital equipment or vehicles. Biswas, Arun, and Vijay Mahajan. 1997. "SEWA Bank." Case In general, the constraints that are the furthest away Study for the Sustainable Banking with the Poor Project, are the hardest for entrepreneurs to do anything about World Bank, Washington, D.C. and can be the most binding. They are the forces that Bratton, M. 1986. "Financing Smallholder Production: A set limits and that may ultimately keep the transforma- Comparison of Individual and Group Credit Schemes in tion of the microenterprises in a given sector from tak- Zimbabwe." Public Administration and Development (UK) 6 ing place. The enterprise development service provider (April/June): 115­32. must understand these constraints even though they are Caskey, John P. 1994. "Fringe Banking--Check-Cashing the hardest and most expensive to study, requiring a spe- Outlets, Pawnshops, and the Poor." Russel Sage Foun- cial capacity to analyze large forces and an investment in dation, New York. rigorous research. Once conclusions about these macro- Committee of Donor Agencies for Small Enterprise Development. 1998. "Business Development Service for level constraints are drawn, implementing real solutions PRODUCTS AND SERVICES 91 SMEs: Preliminary Guidelines for Donor-Funded Sustainable Banking with the Poor Project, World Bank, Interventions." Private Sector Development, World Bank, Washington, D.C. Washington, D.C. Nelson, Candace, Barbara McNelly, Kathleen Stack, and CGAP (Consultative Group to Assist the Poorest). 1997. Lawrence Yanovitch. 1995. Village Banking: The State of "Introducing Savings in Microcredit Institutions: When Practice. The SEEP Network. New York: Pact and How?" Focus Note 8. World Bank, Washington, D.C. Publications. Freedom From Hunger. 1996. "The Case for Credit with Otero, Maria, and Elizabeth Rhyne. 1994. The New World of Education." Credit with Education Learning Exchange Microenterprise Finance: Building Healthy Financial Institutions Secretariat, Davis, California. for the Poor. West Hartford, Conn.: Kumarian Press. Fruman, Cecile. 1997. "FECECAM­Benin." Case Study for Paxton, Julia. 1996a. "A Worldwide Inventory of Microfinance the Sustainable Banking with the Poor Project, World Institutions." World Bank, Sustainable Banking with the Bank, Washington, D.C. Poor Project, Washington, D.C. Goldmark, Lara, Sira Berte, and Sergio Campos. 1997. ------. 1996b. "Determinants of Successful Group Loan "Preliminary Survey Results and Case Studies on Business Repayment: An Application to Burkina Faso." Ph.D. diss., Development Services for Microentrepreneurs." Inter- Department of Agricultural Economics and Rural American Development Bank, Social Programs and Sociology, Ohio State University, Columbia. Sustainable Development Department, Microenterprise Puhazhendhi, V. 1995. "Transaction Costs of Lending to the Unit, Washington, D.C. Rural Poor: Non-Governmental Organizations and Self-help Haggeblade, J., and M. Gamser. 1991. "A Field Manual for Groups of the Poor as Intermediaries for Banks in India." Subsector Practitioners." GEMINI Technical Note. U.S. Foundation for Development Cooperation, Brisbane, Australia. Agency for International Development, Washington, D.C. Rhyne E., and S. Holt. 1994. "Women in Finance and Huppi, M., and Gershon Feder. 1990. "The Role of Groups Enterprise Development." Education and Social Policy and Credit Cooperatives in Rural Lending." World Bank Discussion Paper 40. World Bank, Washington, D.C. Research Observer (International) 5 (2): 187­204. Von Pischke, J.D. 1991. Finance at the Frontier: Debt Capacity Krahnen, Jan Pieter, and Reinhard H. Schmidt. 1994. and the Role of Credit in the Private Economy. World Bank, Development Finance as Institution Building: A New Economic Development Institute. Washington, D.C. Approach to Poverty-Oriented Lending. Boulder, Colo.: Waterfield, Charles, and Ann Duval. 1996. CARE Savings and Westview Press. Credit Sourcebook. Atlanta, Ga.: CARE. Ledgerwood, Joanna, and Jill Burnett. 1995. "Individual Women's World Banking. 1996. Resource Guide to Business Micro-Lending Research Project: A Case Study Review-- Development Service Providers. New York. ADEMI." Calmeadow, Toronto, Canada. Yaron, Jacob, McDonald Benjamin, and Gerda Piprek. 1997. McLean, Doug, and Matt Gamser. 1995. "Hi-Tech at the "Rural Finance Issues, Design and Best Practices." World Frontier: Reducing the Transaction Cost of Lending in Bank, Agriculture and Natural Resources Department, Swaziland with the Smart Card." Seminar Abstract 5, Washington, D.C. CHAPTER FOUR The Institution I n chapter 3 we introduced the "Systems Framework" ernments, donors, and international NGOs. When in which different institutions can be involved in pro- deciding whom to cooperate with, the objectives of both viding the services demanded by low-income clients. the MFI and the potential development partner must be In this chapter the focus is on the primary institution that considered. Partnerships affect the structure of the MFI, provides financial services, its institutional structure, and its funding sources, and its activities. The first section of its capacity to meet the demands of its clients. this chapter gives an overview of why institutions (and The majority of MFIs are created as nongovernmental partner institutions) are important and outlines the vari- organizations (NGOs). However, as the field of microfi- ous institutional types. The remaining sections focus on nance develops, the focus is changing from the delivery of the institutional issues that arise as MFIs grow and credit services to a true process of financial intermediation, expand their operations. including the provision of savings and other financial ser- This chapter will be of interest to donors or consul- vices demanded by the working poor. Furthermore, the tants who are examining or evaluating MFIs and to shrinking resource base (donor funds) to support the ever- practitioners who are considering transforming their increasing demand implies that MFIs will eventually need institutional structure, examining issues of governance, to support themselves. Accompanying this change of per- or developing new funding sources. Practitioners, spective is a better understanding of the implications of donors, and international NGOs in the process of institutional structure for achieving the ends of greater ser- selecting partner institutions will also find this chapter vice, scale, and sustainability. interesting. Many MFIs are now beginning to look at the advan- tages and disadvantages of different institutional struc- tures. This chapter examines the range of institutional The Importance of Institutions types that are appropriate for microfinance and for meet- ing the objectives of MFIs, including formal, semi-for- An institution is a collection of assets--human, financial, mal, and informal structures. It also addresses other and others--combined to perform activities such as granti- institutional issues, including: ng loans and taking deposits overtime.1 A one-time activity I Institutional growth and transformation such as a "project" is not an institution. Thus by its very I Ownership and governance nature an institution has a function and a certain perma- I The accessing of new sources of funding nence. However, when one looks at specific providers of I Institutional capacity. financial intermediation to low-income women and men in Most MFIs have established or are establishing part- a given country, one can easily see differences in the degree nerships with other development agencies, such as gov- to which they really are institutions, that is, the degree to 1. This section and the following section on institutional types were written by Reinhardt Schmidt, Internationale Projekt Consult (IPC) GmbH. 93 94 MICROFINANCE HANDBOOK which they have a well-defined function and are set up and operations. This is important for two reasons. One is that run to perform their function on a permanent basis. Clearly, only a growing institution can meet the demand that its permanence is important. Poor men and women need per- clientele typically exhibits. The other is that many devel- manent and reliable access to savings and credit facilities, opment finance institutions that cater to poor clients are so and only stable institutions can assure permanence. small that the unit costs of their operations are too high. Growth is an efficient way to reduce costs. Attributes of a Good Institution Experts sometimes claim that an MFI should not depend on external support but rather should be subsidy A good institution has three attributes: independent. This requires that the revenues from its 1. It provides services to the relevant target group. operations be sufficient to cover all of its costs, including I Appropriate services include the offering of loans that loan losses, the opportunity cost of equity, and the full, match client demand. This refers to loan size and inflation-adjusted cost of debt. This is a requirement that maturity, collateral requirements, and the procedures any mature institution must meet, just like any commer- applied in granting loans and ensuring repayment. A cial business. However, in most practical cases the real good MFI must have or be willing to adopt an appro- question is not whether an MFI is already financially self- priate credit technology, which will enable quality ser- sufficient--most of them are not--but rather, to what vices to be designed and distributed so that they are extent their costs exceed their revenues and whether and attractive and accessible to the target group. how fast their dependence on external support decreases I The scope of services must be consistent with the situa- over time. An MFI must be able and determined to tion of the clientele. In some cases, simply offering make visible progress toward financial self-sufficiency. loans of a specific type may be all that is needed; in oth- Not all aspects of stability can be readily expressed in ers, an array of different loan types may be necessary, or numbers. One such aspect is organizational stability. A it may be more important to offer deposit and payment sound MFI must have an organizational and ownership transfer facilities or a combination of all these services. structure that helps to ensure its stability both in a finan- I Prices that the clients have to pay for the services of cial sense and with respect to its target group. One can- the institution are not generally a point of major not call an institution stable if it simply turns away from importance, according to practical experience. its original target group of poor clients as soon as it starts However, low transaction costs for clients, a high to grow and become more efficient and professional. degree of deposit liquidity, and rapid availability of This temptation is great, as poor people are not the most loans are extremely important features for a target- profitable target group or the easiest to deal with. group-oriented institution to provide. Therefore it is all the more important that an MFI have 2. Its activities and offered services are not only demand- an ownership and governance structure--that is, a divi- ed but also have some identifiable positive impact on the sion of roles between the management and the board of lives of the customers. directors or the supervisory board--that prevents the 3. It is strong, financially sound, and stable. institution from "drifting away into social irrelevance." Because the people who belong to the target group need a reliable supply of financial services--that is, access The Importance of Partner Institutions to credit facilities--and an institution to which they can entrust their deposits, it is of paramount importance that Most MFIs, with the possible exception of some commer- the MFI be a stable or sustainable institution or at least cial banks, work with one or more development agencies or be clearly on the way to becoming one. partners. These development agencies may be international Stability has, first and foremost, a financial dimension. NGOs, governments, or donors that provide technical A stable institution is one whose existence and function is assistance, funding, and training to the MFI itself rather not threatened by a lack of funds for making necessary than the MFI's clients. (While some partner institutions payments; it must be solvent at any given moment and may in fact provide services directly to the clients of an also in the foreseeable future. In addition, a sustainable MFI, here we are considering partners that help to develop institution must be able to maintain or expand its scale of the ability of an MFI to provide financial intermediation.) THE INSTITUTION 95 Table 4.1 Key Characteristics of a Strong Microfinance Institution Key areas Characteristics Vision I A mission statement that defines the target market and services offered and is endorsed by management and staff. I A strong commitment by management to pursuing microfinance as a potentially profitable market niche (in terms of people and funds). I A business plan stating how to reach specific strategic objectives in three to five years. Financial services and delivery methods I Simple financial services adapted to the local context and in high demand by the clients described in the mission statement. I Decentralization of client selection and financial service delivery. Organizational structure and human resources I Accurate job descriptions, relevant training, and regular performance reviews. I A business plan spelling out training priorities and a budget allocating adequate funds for internally or externally provided training (or both). I Appropriate performance-based incentives offered to staff and manage- ment. Administration and finance I Loan processing and other activities based on standardized practices and operational manuals and widely understood by staff. I Accounting systems producing accurate, timely, and transparent infor- mation as inputs to the management information system. I Internal and external audits carried out at regular intervals. I Budgets and financial projections made regularly and realistically. Management information system I Systems providing timely and accurate information on key indicators that are most relevant to operations and are regularly used by staff and management in monitoring and guiding operations. Institutional viability I Legal registration and compliance with supervisory requirements. I Clearly defined rights and responsibilities of owners, board of direc- tors, and management. I Strong second level of technically trained managers. Outreach and financial sustainability I Achievement of significant scale, including a large number of under- served clients (for example, the poor and women). I Coverage of operating and financial costs clearly progressing toward full sustainability (as demonstrated in audited financial statements and financial projections). Source: Fruman and Isern 1996. Both practitioners and donors should consider the impor- same rights to determine what the partners can do and tance of partnerships, because the need to strengthen local want to do together. institutions and build their capacity is essential to meeting Partnerships are formed when: the growing demand of low-income clients. "organizations seek to mutually strengthen and The term "partner" does not suggest a biased rela- sustain themselves. It [the partnership] is an tionship in which the more powerful and resourceful empowering process, which relies on trust and side imposes its will and ideas upon the other. Instead, confidence, solidarity of vision and approach, and partners discuss and jointly define objectives and ways of it acknowledges mutual contribution and equality. attaining them. In a partnership both sides have the Both partners have complementary roles, estab- 96 MICROFINANCE HANDBOOK lished through negotiations and subject to change partners if a permanent institution with a lasting impact as the partnership grows and circumstances for low-income women and men is to be created. change." (Long, Daouda, and Cawley 1990, 124) Foreign partner agencies can bring funding, technical Local institutions can bring to the partnership the assistance, and training to the partnership. Often foreign advantage of knowing more about local circumstances: partner agencies are familiar with microfinance "best prac- the target group or clientele, their situation and demand tices" and have at their disposal information and sources to for financial services, the local financial market, and the which the local partner may not readily have access. local laws and habits. Furthermore, there are ethical and An American international NGO, Freedom From political considerations that make it advisable for a for- Hunger, focuses on linking its MFI operations in Africa with eign institution to have a local partner. Last but not least, local bank and credit unions. This has allowed it to focus pri- a local partner is needed when an institution has been marily on technical assistance while creating sustainable insti- established and the foreign partner is planning to with- tutions in the countries in which it works (box 4.1). draw from the MFI and leave it in the hands of local Partner selection implies cooperation, which is only people. The failure of thousands of development finance possible if there is a common purpose, mutual respect projects in the past demonstrates that it is necessary to and openness, and a willingness by both partners to con- have not only an institutional basis but also strong local tribute to the achievement of the common goal. Thus Box 4.1 Freedom From Hunger: Partnering with Local Institutions FREEDOM FROM HUNGER (FFH) WORKS EXTENSIVELY IN chance that Freedom From Hunger's partners are largely Africa with local partners. Rather than going into a country credit unions or development banks--institutions with a and setting up its own microfinance activity, Freedom From strong social or development mission. How such an Hunger selects local formal financial institutions, trains them approach might play out with more profit-seeking partners its own lending model, and provides technical advice. The has not been tested. organization has identified the advantages of this partnership Determinations that are of subsidiary but significant for both parties: importance are : I Freedom From Hunger reduces the cost of delivering ser- I The financial health of the lending institution vices, because the local institution can often fund its own I The source of financing for the portfolio (in some portfolio. Freedom From Hunger's expertise in financial instances it is the bank or credit union, in others cases management greatly helps in the start-up and ongoing the international NGO, which may need to supplement management. The local partner offers clients access to the the partner's available resources) formal financial sector. Credit unions, as partners, bring I The source of funding for operational costs particular advantages because they have a network of con- I The selection of staff delivering program services tacts, know their clients, and offer other local knowledge. I Responsibility for loan loss. I The local partner gains access to creditworthy clients When these arrangements are worked out successfully, using Freedom From Hunger's model, which features a the international NGO can see its activities grow beyond high repayment rate and group loan guarantees. The new any scale that it could have reached with its own resources; it borrowers are also savers who build the bank's loan capi- can achieve a sustainable initiative, whether under its own tal and profitability. Finally, reaching out to the clients control or under that of the local financial institution. In the targeted by Freedom From Hunger represents an invest- second case it has also achieved the establishment of microfi- ment in the welfare of the poor from which the bank can nance within the formal financial system in a way that avoids derive public relations and marketing benefits. many of the difficult challenges inherent in other models. Making this model work requires special attention to a The financial services are offered under the aegis of a regulat- variety of issues. The greatest challenge for the international ed institution that can provide security and stability to its NGO is finding financial partners who have an interest in clients, and the organization does not have to transform the poverty and microbusiness sector and will honestly enter itself or create a new financial institution to deliver services the relationship with a long-term commitment. It is not by to those it most wants to serve. Source: SEEP Network 1996a. THE INSTITUTION 97 there is no standard answer to the question, What makes a good local or international partner institution? It Box 4.2 Types of Financial Institutions depends on who the other partner organization is, what Formal institutions its ideas and objectives are, what it can contribute, and I Public development banks the way in which it is willing to act as a partner. I Private development banks I Savings banks and postal savings banks I Commercial banks Institutional Types I Nonbank financial intermediaries. Semiformal institutions I Credit unions The following is a brief discussion of each institutional I Multipurpose cooperatives type and includes its suitability as a partner to a donor, I NGOs international NGO, or government. The suitability of I (Some) self-help groups. donors, international NGOs, and governments as part- Informal providers ners is not discussed. I (Pure) moneylenders Formal institutions are defined as those that are sub- I Traders, landlords, and the like (as moneylenders) ject not only to general laws and regulations but also to I (Most) self-help groups specific banking regulation and supervision. Semiformal I Rotating savings and credit associations (work groups, institutions are those that are formal in the sense of being multipurpose self-help groups) registered entities subject to all relevant general laws, I Families and friends. including commercial law, but informal insofar as they Source: Author's findings. are, with few exceptions, not under bank regulation and supervision. Informal providers (generally not referred to as institutions) are those to which neither special bank nology and organizational structures as they had used law nor general commercial law applies, and whose before, their success was extremely modest. operations are also so informal that disputes arising from Rural development banks and specialized small business contact with them often cannot be settled by recourse to development banks that escaped closure during the recent the legal system. wave of financial sector reforms that swept over many countries have some attractive features: almost all of them Formal Financial Institutions have a broad network of branches; many are by now active collectors of deposits; and quite often they are the The following are the most typical types of formal finan- only large institutions in rural areas to offer at least a min- cial institutions. imum of financial services to poor clients (box 4.3). Development banks can be considered good partners for PUBLIC DEVELOPMENT BANKS. Development banks are or foreign development agencies--not in the traditional sense until quite recently have been, a special type of large, of providing funds, but rather in improving and reshaping centralized, government-owned bank. Most of them their organizational structures, financial management, and were set up with ample financial support from foreign the way in which they design and provide their services to and international organizations. They were created to poorer clients. The crucial questions to be asked before provide financial services to strategic sectors such as agri- undertaking such a partnership are whether there is an culture or industry. Most are the product of traditional openness to change and a willingness to give up the ample interventionist approaches, which place greater emphasis powers that the traditional approach certainly entailed and on disbursements at low, subsidized interest rates than whether a foreign organization has enough resources to sup- on the quality of lending. port such a change if it is desired not only by the people in Their traditional clients are large businesses. In the the bank but by relevant politicians in the country. 1970s several development banks started to offer their services to small farmers and small businesses. But P RIVATE DEVELOPMENT BANKS . Private development because they tried to do this with the same credit tech- banks are a special category of banks that exist in some 98 MICROFINANCE HANDBOOK Box 4.3 Development Banks D EVELOPMENT BANKS CAN LEND EITHER DIRECTLY OR lage cooperatives and groups of farmers gave better results. through intermediaries, as shown in these examples. In more recent years the Banque Nationale has started lend- Direct lending. The Bank for Agriculture and Agricultural ing to microfinance institutions, such as the caisses villa- Cooperatives (BAAC) in Thailand is a state-owned develop- geoises--self-managed village banks--in three regions of the ment bank created in 1966 to provide financial assistance to country. These institutions in turn on-lend to their clients. farmers and agriculture-related activities. Since the financial This linkage has been extremely successful, with high repay- reforms of 1989 the bank's efforts have been directed pre- ment rates and very low costs for the bank. The bank now dominantly at the low- to medium-income range. At first it wishes to become directly involved in setting up self- lent mostly through agricultural cooperatives, but repayment managed village banks in other areas of the country. If its problems led the bank to start lending directly to farmers. sole motivation is to channel more funds to the rural econo- Lending through intermediaries. The experience of the my, this may prove to be a dangerous strategy. Indeed, the Banque Nationale de Développement Agricole (BNDA), a success of the caisses villageoises in Mali is based on their state-owned bank in Mali, is quite different. This bank ownership by the villagers themselves--and this sense of ceased lending directly to clients in rural areas in the early ownership might be weakened if the Banque Nationale is 1990s because of very low repayment rates. Lending to vil- too strongly involved. Source: Contributed by Cecile Fruman, Sustainable Banking with the Poor Project, World Bank. developing countries. Their aim is broad economic of deposits is normally invested in government securities development directed to fill capital gaps in the produc- or simply transferred to the treasury. Most post office tive sector that are considered too risky by commercial savings banks are part of the postal administration of standards. They often have lower capital requirements their country (box 4.5). than commercial banks and enjoy some exemptions, With a few noteworthy exceptions, post office sav- either in the form of tax breaks or reduced reserve ings banks are in poor shape, both financially and oper- requirements (box 4.4). ationally. Their financial situation is poor because they have to pass their net deposits on to the treasury, which SAVINGS BANKS AND POSTAL SAVINGS BANKS. The legal often does not want to acknowledge its indebtedness to status and ownership of savings banks varies, but they are the post office savings banks. Their organizational weak- typically not owned by the central government of their nesses derive from the fact that they are managed by the country, and they often have a mixture of public and pri- postal service, which has no genuine interest in their vate owners. Some, such as those in Peru, were set up success. This state of affairs is all the more regrettable in and owned by municipalities; others, such as the rural view of the fact that deposit facilities and money trans- banks in Ghana, are simply small local banks owned by fer services within easy reach of the majority of poor local people. people are "basic financial needs," which other institu- As the name indicates, savings banks tend to empha- tions tend not to provide with comparably low transac- size savings mobilization more than other banks. Their tion costs. main strength is that they are decentralized, rooted in Savings banks could be attractive partners for vari- the local community, and interested in serving local ous kinds of microfinance activities. Their main small business. strengths are that they are decentralized, rooted in the In many countries there are post office savings banks local community, and interested in serving local small modeled on patterns imposed by former colonial powers. business. Some savings banks, such as those in Peru, A typical post office savings bank operates through the are among the most successful target-group-oriented counters of post offices, and therefore, has a very large financial institutions, while others may be good part- network of outlets; it does not offer credit but only takes ners for activities aimed at restructuring the savings deposits and provides money transfer services. A surplus bank system. THE INSTITUTION 99 C OMMERCIAL BANKS . Commercial banks are formal lending to established businesses. A typical commercial financial institutions that focus on short- and long-term bank has little experience in providing financial services Box 4.4 Tulay sa Pag-Unlad's Transformation into a Private Development Bank TULAY SA PAG-UNLAD, INC. (TSPI), IN THE PHILIPPINES, With these realities in mind, one option under considera- determined that it would create a private development bank tion is a special transitional structure, with the new bank and the based on the current regulatory environment and its own original NGO in one integrated organization with shared per- institutional capacity. sonnel and a unifying mission. The bank and the NGO would I New banking regulations permit the establishment of maintain separate financial statements and transparent inter- private development banks with much lower capital company transactions but would share many of the head office requirements (US$1.7 million, as opposed to US$50 functions and some of the operational management. The NGO million for commercial banks). However, the new private would be responsible for the Kabuhayan product aimed at the development banks are only allowed to operate outside poorest people as well as those branches within Metropolitan Metropolitan Manila, meaning that TSPI will need to Manila that cannot be legally transferred to the bank. The bank consider the location of its branches carefully so as to stay would be responsible for the most profitable loan products, close to its desired client base. located in those branches most ready for bank status. I While TSPI has managed to create a diversified product The transition period would take two to three years as line, one of its four loan services, Sakbayan, which pro- the bank grew and became large enough to operate within vides loans to tricycle and taxi drivers, is significantly Manila and take over the rest of TSPI's branches. At that stronger than the others and will become the anchor point the bank and the NGO would separate all functions product for the bank. This product, however, has limited and staff. The decision to merge Kabuhayan into the bank growth potential because taxi routes are regulated. would be made at that time or later. If the conclusion was Strengthening the other product lines is thus a priority that it would never reach commercial viability, the NGO task for TSPI. would continue with a separate head office structure. In the I TSPI recognizes that it still faces some capacity- meantime, the unified structure provides TSPI with several building issues, including human resource and sys- clear advantages: tems development. However, the pressure to source I It controls administrative costs. additional loan capital and the potentially short regu- I It avoids potential cultural conflicts between the NGO and latory "window of opportunity" are demanding pro- the bank about the mission. gress toward the creation of a private development I It ensures coherent policies, performance measurement, bank. and expectations. Source: Calmeadow 1996. Box 4.5 The Savings Bank of Madagascar F ROM 1918 TO 1985 THE POSTAL ADMINISTRATION IN The Savings Bank of Madagascar provides its clients Madagascar collected savings. In 1985 the Savings Bank of with savings services only. Clients earn interest on their Madagascar (Caisse d'épargne Madagascar) was created savings depending on the length of time they are held. under the financial supervision of the Ministry of Finance Once a year all clients give their transactions books to the and the technical supervision of the Ministry of Post and bank's agency or postal office to calculate the accrued Telecommunications. The Savings Bank of Madagascar interest. began with 46 existing postal offices. In June 1996 it was The Savings Bank of Madagascar focuses on poor clients, operating out of 174 windows located in both main cities as evidenced by an average savings balance of less than and remote small towns. For use of the postal outlets the US$32. This represents 28 percent of GDP. The total num- bank pays the postal administration 0.8 percent on savings ber of clients in June 1996 was 372,291, with 46 percent of collected. them women. Source: Galludec 1996. 100 MICROFINANCE HANDBOOK to microentrepreneurs and small farmers; however, some commercial banks recently recognized that it might be Box 4.6 Caja Social: A Colombian Commercial Bank beneficial to do business with these clients (box 4.6). Reaching the Poor They also understand that it might be necessary to pro- CAJA SOCIAL, FOUNDED IN 1911 IN COLOMBIA AS THE ST. vide services in ways that differ from the traditional Francis Xavier Workmen's Circle and Savings Bank, is an commercial banking approach. example of a commercial bank that has a significant micro- There are 12 basic principles with which banks must finance portfolio. Its original mandate was to promote sav- comply if they choose to focus part of their operations on ings among the country's poor. Its success in deposit low-income clients (Yaron, Benjamin, and Piprek 1997): mobilization led to an expansion of branch savings banks 1. Ensure appropriate governance throughout Colombia. In 1991 Caja Social was officially transformed into a commercial bank. Its activities, overseen 2. Define the institution's strategies and objectives by the holding company, Fundacion Social, range from 3. Learn from the competition in the informal sector mortgage credit to small business loans and leasing. 4. Find out what services clients really want As of September 1995 Caja Social had 1,159,204 active 5. Establish appropriate modes of delivery savings accounts (averaging US$338) and 173,033 out- 6. Contain transaction costs standing loans (averaging US$2,325). While Caja Social 7. Cover costs with appropriate, positive on-lending serves numerous middle-income clients, a significant por- interest rates tion of its business is targeted at low-income clients. Nearly 8. Customize loan terms and conditions for the target 20 percent of its clients earn less than US$3,000 annually, clientele which is 75 percent of the country's average income per 9. Monitor and maintain the quality of the assets economically active adult. Given the enormous scale of the 10. Manage and diversify risks bank, this translates into well over 225,000 low-income clients. In addition, the Caja has been successful in target- 11. Mobilize savings resources in the market ing female clients. Approximately 43 percent of its clients 12. Motivate staff and invest in them (with information are women. and incentives). While Caja Social always has maintained the objective of Some banks choose to develop a range of services for assisting the country's poor, it has found this market niche to microentrepreneurs and rural peasants and then to ser- be financially rewarding. In fact, it has been one of the most vice them directly by going to the communities. Branch profitable banks in the Colombian financial system. Arrears banking is the more traditional approach, with bank rates remain manageable at 4.5 percent, and the bank has tellers providing services in a bank facility. A highly achieved operational and financial self-sufficiency without decentralized network of branches is necessary, which dependence on subsidies from either national or internation- must include very small branches with minimal fixed al organizations. Caja Social's experience has shown that assets, located in central villages where they can serve banking with the poor can be sustainable and profitable. poor clients in nearby rural areas or in urban slums. This Source: Contributed by Julia Paxton, Sustainable Banking with is the most likely approach for banks, but it is costly. the Poor Project, World Bank. Other banks may rely more on intermediaries, acting as a wholesaler rather than a retailer. "Downscaling" is the technical term used to describe nical assistance to build up departments for small and projects that aim to introduce new approaches into a micro lending. commercial bank that has so far not tried to provide ser- vices to a poorer clientele. Downscaling is, for instance, NONBANK FINANCIAL INSTITUTIONS. In some countries spe- the centerpiece of what the Inter-American Development cial regulation has been established for nonbank financial Bank is practicing in its "micro global" programs in institutions. These institutions are set up to circumvent Latin America. Another example is the Russia Small the inability of some MFIs to meet commercial bank stan- Business Fund established by the European Bank for dards and requirements due to the nature of their lending. Reconstruction and Development. In both cases com- Examples of MFIs regulated as nonbank financial institu- mercial banks are carefully selected and analyzed and, tions are Caja de Ahorro y Prestamo Los Andes in Bolivia after a positive evaluation, may receive funding and tech- and Accion Comunitaria del Peru (MiBanco) (box 4.7). THE INSTITUTION 101 Box 4.7 Caja de Ahorro y Prestamo Los Andes CAJA DE AHORRO Y PRESTAMO LOS ANDES WAS ESTABLISHED were already in place. Consequently, Los Andes has been able as the first Bolivian private financial fund in 1995. Los to produce the required reports to the central bank and the Andes grew out of the nongovernmental microlender Pro- Bolivian superintendency with few modifications. Credito, with the technical assistance of the private consult- In contrast to Los Andes, BancoSol (Banco Solidario ing firm Internationale Projekt Consult GmbH (IPC), S.A.) entered the regulated financial sector as a commercial financed by the German development agency, GTZ. The bank in 1992, because the private financial funds did not minimum capital requirement for a private financial fund is yet exist. As a result, BancoSol became the first private com- US$1 million. The majority of the total US$600,000 of mercial bank in the world dedicated solely to providing paid-in capital to Los Andes came from Pro-Credito. financial services to the microenterprise sector. This helps to From its inception as Pro-Credito in 1991, the institution explain why the Bolivian Superintendency of Banks is con- planned to enter the regulated financial sector. No organiza- sidered to be one of the most innovative in Latin America. tional changes were implemented once it became a private It is making significant strides in creating a competitive financial fund. Sophisticated management information sys- financial market and is committed to opening the financial tems, which integrate data on savings and credit operations, sector to microenterprises. Source: Rock 1997. Finance companies or financiers are nonbank financial Financial cooperatives provide savings and credit ser- intermediaries that channel equity funds, retained earn- vices to individual members. "They perform an active ings, and other borrowed capital to small, unsecured financial intermediation function, particularly mediating short-term loans. Finance companies are often associated flows from urban and semi-urban to rural areas, and with consumer credit and installment contracts. They are between net savers and net borrowers, while ensuring often not allowed to mobilize savings; however, their activities vary by their charters. For example, some are able to mobilize time deposits but not demand deposits. Box 4.8 Accion Comunitaria del Peru Their form is now being adapted in some countries to ACCION COMUNITARIA DEL PERU ORIGINALLY CONSIDERED provide a regulated vehicle for microfinance with lower transforming itself into a finance company (which requires barriers to entry than a commercial bank (box 4.8). a minimum capital base of US$2.7 million) or a commer- cial bank (which requires a minimum capital base of US$6 Semiformal Financial Institutions million.) However, the Peruvian development bank, COFIDE, proposed the formation of a new category of nonbank financial institutions to meet the financing needs The most common types of semiformal financial institu- of small and microenterprises. In December 1994 the tions are financial cooperatives and financial NGOs. superintendency issued a resolution creating a new struc- ture called an Entidades de Desarrollo para la Pequena y CREDIT UNIONS, SAVINGS AND LOAN COOPERATIVES, AND Microempresa (EDPYME). As such an entity, an MFI OTHER FINANCIAL COOPERATIVES. There are a great many with a minimum capital requirement of US$265,000 can forms of cooperative financial institutions (often identified access capital markets, additional bank funding, and spe- as credit unions or savings and loan cooperatives). Such insti- cial rediscount credit facilities from COFIDE. An tutions play a significant role in the provision of financial EDPYME is required to maintain a loan loss reserve equiv- services to poor target groups. In several countries coopera- alent to 25 percent of capital, and at least 10 percent of tive financial institutions are structured in accordance with after-tax profits must be transferred to this reserve annual- the models of the American and Canadian credit union or ly. The resolution specifies that it provide financing to per- sons engaged in activities characterized as small or micro cooperative systems. Other cooperative institutions, such businesses. as Raiffeisen (Germany), Credit Mutuel (France), Alternative Bank (Switzerland), and Triodos (Holland), Source: Montoya 1997. are inspired by models developed in Europe. 102 MICROFINANCE HANDBOOK that loan resources remain in the communities from ble to cooperatives in general. In some countries specif- which the savings were mobilized" (Magill 1994, 140). ic regulations for credit unions are being developed. In Financial cooperatives are organized and operated the West African Economic and Monetary Union according to basic cooperative principles: there are no (with eight countries as members) a law for savings and external shareholders; the members are the owners of loan associations was passed in 1994, placing these the institution, with each member having the right to institutions under the supervision of the Ministry of one vote in the organization. In addition to holding Finance in each country. shares redeemable at par, members also may deposit Characteristics of financial cooperatives include: money with the organization or borrow from it. I Clients who tend to come from low-income and Membership is usually the result of some common lower-middle-income groups. bond among members, often through employment or I Services that are almost exclusively financial in nature. membership in the same community. The policy-mak- I Self-generated capital, typically without any depen- ing leadership is drawn from the members themselves, dence on outside funding to cover operating costs, and members volunteer or are elected for these posi- which are generally kept low. tions. Financial cooperatives are rarely subject to bank- Individual financial cooperatives often choose to be ing regulation; instead, they are either unregulated or affiliated with a national league (apex institution), which subject to the special legislation and regulation applica- serves the following purposes: it represents the credit Box 4.9 The Rehabilitation of a Credit Union in Benin THE HISTORY OF THE FÉDERATION DES CAISSES D'ÉPARGNE government, members of the network, and donors that con- et de Crédit Agricole Mutuel dates back to the 1970s. In tributed funds to reimburse depositors, finance studies, and 1975 the Caisse Nationale de Crédit Agricole, or National cover the operating shortfall of the network. The results of Agricultural Credit Bank, was established as a public devel- the first phase were very encouraging insofar as the project opment bank, followed in 1977 by the first local and region- met its principal objectives: restoring the rural population's al credit unions. Over the next 10 years 99 local credit confidence in the network, strengthening management by unions were created. The Caisse Nationale de Crédit elected directors, and improving financial discipline. Agricole then assumed the role of a national federation, tak- In light of the results of the first phase, particularly the ing control of the operations of the entire network and enthusiasm demonstrated by the principal participants, a excluding the elected directors from management and con- second phase of rehabilitation (1994-98) began. The objec- trol. As a result, rules and procedures were developed by the tives of the second phase--currently in progress--are as fol- management of the bank and imposed on elected directors lows: and members. Thus the constituencies being served were I Financial restructuring of the network, both to mobilize unable to voice their concerns and demands. This top-down deposits for on-lending and to generate revenue to cover approach proved to be highly detrimental. operational expenses The financial performance of the network slowly deteri- I A transfer of responsibility to elected directors through orated, and savings were no longer secure. In November the creation of a national federation to govern network 1987 the Caisse Nationale de Crédit Agricole was liquidated. expansion, structure, and training of staff and directors Nevertheless, a study carried out in 1988 showed that the I The creation of the technical secretariat of the federation local credit unions had continued to operate throughout the to supply specific technical support (such as training and crisis and member savings had continued to grow. To inspection) and ensure implementation of general policies. strengthen the autonomy of the network, the Government The second phase of rehabilitation is ongoing, but for the of Benin decided to initiate a rehabilitation program. It most part its objectives have already been met. The net- emphasized the importance of the total freedom of local work's growth has greatly exceeded expectations. The transi- credit unions to define their own policies, including the tion from mere project status to that of an operational, choice of interest rates. The first phase of the rehabilitation autonomous network offering a wide array of credit and sav- program (1989-93) was developed in collaboration with the ings services is nearly complete. Source: Fruman 1997. THE INSTITUTION 103 unions at the national level, provides training and techni- which are the main reason that they are common institu- cal assistance to affiliated credit unions, acts as a central tional types for MFIs. deposit and inter-lending facility, and, in some cases, Financial NGOs should be distinguished from two channels resources from external donors to the national other types of nongovernmental and not-for-profit cooperative system. Contacts with foreign partners are institutions: self-help groups and cooperatives. These typically handled by an apex institution. Affiliation institutions are member-based, whereas NGOs are set involves purchasing share capital and paying annual dues up and managed not by members of the target group, to the national or regional apex. Membership provides the but by outsiders, who are often men and women from right to vote on national leadership and policies and to the middle class of the country who want to support participate in nationally sponsored services and programs. poorer people for social, ethical, and political reasons. Savings services are a key feature for raising capital For example, one of the first Latin American NGOs and are often tied to receiving a loan. Credit is generally providing loans to local microentrepreneurs, a Cali- delivered under the "minimalist" approach. Cooperative based organization called DESAP, was started and man- lending requires little collateral and is based on character aged by a son of one of the wealthiest families in references and cosigning for loans between members. Colombia, who regarded the government's neglect of Without any doubt, there are well-functioning coop- microbusiness as a serious threat to the political stability erative systems. Beyond that, the sheer number of people of his country. who are members of these systems and the aggregate In spite of the fact that they are not member-based, amounts of deposits and loans are impressive. many NGOs are close to the target group, in terms of Nevertheless, many systems do not function as well as both location and understanding. An NGO set up by their basic philosophy would lead one to expect. At the local business people to provide services to microentre- same time, they are very difficult partners for foreign preneurs may indeed be better placed to respond to the institutions. The reason for both of these difficulties is needs of its clients than an institution managed by gov- their structure and the mechanism, which in principle ernment officials or traditional bankers. should make them work. All this indicates that financial NGOs are particularly I The group that constitutes the organization is small promising candidates for foreign institutions in search of a enough so that members know each other well. local partner. However, there are some factors that make I The members regularly change their roles from net NGOs less attractive. The weaknesses of many, though cer- depositors to net borrowers and vice versa. tainly not all, NGOs can be attributed to a combination of If these conditions are not met, a cooperative becomes the following factors (not all of which are unique to them): unstable: the management cannot be monitored, and a I The lack of business acumen with which some NGOs structural conflict arises between borrowers (who prefer are set up and operated low interest rates and little pressure for repayment) and I Overly ambitious aspirations with regard to their net depositors (who prefer high interest and a very cau- social relevance tious application of their deposits). I The limited scale of their operations, which does not permit them to benefit from elementary economies of FINANCIAL NGOS. NGOs are the most common institu- scale tional type for MFIs. The World Bank's "Worldwide I The frequent use of donated funds or soft loans from Inventory of Microfinance Institutions" found that of foreign development organizations the 206 institutions responding to a two-page question- I The influence of people who do not belong to the tar- naire, 150 were NGOs (Paxton 1996). get group and thus are not subject to peer pressure Even more than with other types of institutions, a dis- and are not directly hurt if the institution's money is cussion of NGOs that provide financial services has to eventually lost. start by emphasizing that they are indeed a very diverse All these factors can work together to create a situa- group. The general definition of an NGO is based on tion in which the business and financial side of a finan- what it is not: neither government-related nor profit- cial NGO is not treated with as much care as it ought to oriented. This already indicates their specific strengths, be. Nonetheless, business expertise and stability are indis- 104 MICROFINANCE HANDBOOK pensable if the institution wants to have a permanent tions. For obvious reasons, informal finance comes in many positive effect on the target group. forms and not always in one that can be called a financial Another weakness of many NGOs is their lack of a institution. Some agents in informal financial markets have longer-term perspective and strategy (or of effective busi- at least some aspect of a financial institution about them. ness planning). The questions that a look into the future This applies to moneylenders and money collectors-- would pose are: Where will the institution stand in a few known, for example, as "susu men" in Ghana--and to the years? Will it still exist, will it collapse in the course of millions of rotating savings and credit associations that exist time, or will it survive and even prosper? How is survival and operate in many variants and under many different possible? Will it turn away from its original target group names in almost every country of the developing world. merely to survive financially? According to available evi- Self-help groups are also a type of informal (sometimes dence and experience, for most financial NGOs to sur- semiformal) financial institution. It seems less appropriate vive without losing their target-group orientation, they to apply the term "financial institution" to traders and must first change into professional institutions and then, manufacturers who provide trade credit to their customers, if desired, expand and ultimately transform their institu- and such a classification is completely inappropriate for a tional structure. most important source of informal finance, namely, the network of friends and family members. Informal Financial Providers A certain degree of institutionalization and functional specialization is necessary for an entity in a developing Informal finance is probably much more important for the country to be a potential partner in the context of a financial management of poor households than the provi- finance-related development project. Only very few infor- sion of services by formal and semiformal financial institu- mal institutions can play the role of a local partner. This Box 4.10 CARE Guatemala: The Women's Village Banking Program CARE IS AN INTERNATIONAL NONPROFIT ORGANIZATION The performance of the CARE village banking pro- founded in 1946. It was created to provide assistance and gram has shown steady improvement. All indicators of development programs to the needy. In the middle to late outreach, cost, repayment, sustainability, and profit 1980s CARE International became increasingly interested became more favorable during 1991­95. However, the in microenterprise programs. These programs, notably mere improvement of these figures does not necessarily microfinance, gained worldwide recognition during the indicate that the program is on a sustainable trajectory. 1980s, spawned in part by the success of microfinance pro- The program is still burdened by high overhead costs and grams such as Grameen Bank and Bank Rakyat Indonesia. donor dependency, which, unless rectified, will prevent it While lacking expertise in this field, CARE International from becoming sustainable. became interested in experimenting with microenterprise The village banking program represents a departure from development and hired external consultants to facilitate the other humanitarian programs of CARE Guatemala. For and supervise early experiments in this area. CARE the first time a program is bringing in revenue from the very Guatemala was one of the programs selected for microen- poor it serves. Like many nonprofit organizations that enter terprise development work. into microfinance, dual objectives pull the village banking pro- While CARE Guatemala began providing health, edu- gram in two directions. Often the microfinance programs of cation, and income-generating programs as early as 1959, NGOs worldwide were created to improve the lives of the the women's village banking program of CARE was poor regardless of cost, repayment, or sustainability. As pro- founded in 1989. In its short history, this program has grams have proliferated and donors have become more insis- made inroads into providing microfinance services to some tent on financial performance, a secondary goal of of the most marginalized people in Latin America. Using a sustainability has been introduced. Given its long-standing cul- village bank methodology, the program has provided loans ture of donor dependence and subsidization, the CARE village averaging only US$170 in 1995 to some 10,000 rural banking program has found it difficult to unilaterally reject women in Guatemala. donor assistance and insist on complete financial sustainability. Source: Contributed by Julia Paxton, Sustainable Banking with the Poor Project, World Bank. THE INSTITUTION 105 is regrettable because if they could be made partners, sev- The important exception to be mentioned here is self- eral types of informal providers of financial services might help groups. There are many self-help groups, consisting of work very well, as they are certainly able to reach the tar- self-employed women who informally support the econom- get group and their permanent existence proves that they ic activities of their members by providing mutual guaran- are sustainable. But many informal arrangements, such as tees that facilitate their members' access to bank loans, by rotating savings and credit associations, would probably borrowing and lending among themselves, or by encourag- only be destabilized by efforts to enlist them as partners, ing each other to save regularly (box 4.11). Yet, at the same because the mechanisms on which they are based depend time these groups are visible or "formal" enough to estab- on their unfettered informality. lish contact with some development agencies. There are Box 4.11 The Use of Self-Help Groups in Nepal THERE ARE MANY INTERNATIONAL NGOS AND GOVERNMENT appraisal is generally done by the group with some guidance programs in Nepal that utilize local self-help groups or sav- from the staff. Interest rates on loans to end-borrowers are ings and credit cooperatives in the provision of microfinance. between 10 and 36 percent, with the higher rates charged by They provide some or all of the following services: nongovernment-funded programs. Some programs of inter- I Revolving funds for on-lending national NGOs, particularly those providing microfinance as I Grants to cover operating costs, including staff and other a minor activity, charge very low rates of interest, sometimes operating expenses as low as 1 percent. Not surprisingly, repayment of these I Matching funds, whereby the international NGO match- loans is very poor, because borrowers tend to view them as es (or provides a multiple of) the amount of savings col- grants rather than loans. lected by the savings and credit cooperative from its Savings are usually compulsory and are required on a members weekly or monthly basis. For the most part these savings I Technical assistance, including program development, cannot be withdrawn, resulting in significantly higher effec- group formation, staff and client training, and financial tive borrowing rates. Interest rates paid on these savings management. range from 0 percent to 8 percent, with the majority paying Target Market. Many international NGOs supporting 8 percent. (Interest paid is usually simply added to the sav- savings and credit cooperatives focus on reaching the "poor- ings held.) Savings are often managed by the group itself, est of the poor" with financial services. In addition, many resulting in access to both "internal loans" (those made from specifically target women, believing that the benefits of the group savings) and "external loans" (those made by the increased economic power will be greater for women because cooperative). It is worth noting that interest rates on internal they are generally responsible for the health and education of loans, which are set by the group, are often substantially their children and the welfare of the community itself. higher than those on external loans, even within the same However, these organizations frequently combine the deliv- groups. This indicates that borrowers can pay higher rates of ery of financial services with social services. This often results interest and do not require subsidized interest rates. It also in lower rates of repayment, because the social services are indicates that the "ownership" of funds greatly influences the usually delivered free of charge while financial services are interest rate set on loans and the repayment of loans, in turn not. contributing to financial sustainability. Methodology. In programs supported by both government The success of savings and credit cooperatives providing and international NGOs, the savings and credit cooperatives microfinance services supported by government and interna- are responsible for the delivery of financial services to clients. tional NGOs is greater than that of government-mandated The cooperatives use groups for lending and savings collec- programs that do not use the cooperatives. Experience over tion. Loan sizes from the cooperatives are generally small and the past five years indicates that the savings and credit coop- no physical collateral is required. Instead, group guarantees eratives can deliver financial services at a much lower cost are provided. The savings and credit cooperatives, with few than the government itself to reach the intended target mar- exceptions, provide loans for relatively short terms, generally ket. However, these programs still suffer from high loan loss- less than one year, over half for less than six months. Loan es and inefficient management. Source: DEPROSC and Ledgerwood 1997. 106 MICROFINANCE HANDBOOK some identifiable characteristics of a good self-help group: it cussion of governance and ownership and accessing exhibits a high degree of coherence and strong common capital markets--issues that arise as the MFI expands. interests among its members; it has a democratic structure and, at the same time, a strong leadership; and empower- Expansion Within an Existing Structure ment of its members is an important objective. However, development projects rarely use these groups Depending on the objectives of the MFI and the contex- as their main partners. Instead, some of them try to estab- tual factors in the country in which it works, an NGO lish a network composed of various partners, including or cooperative may be the most appropriate institutional self-help groups and institutions that provide financial structure, providing the MFI can continue to grow and services to the self-help groups or their members. Only in meet the demands of the target market. Existing struc- very few exceptional cases could self-help groups of poor tures may be most appropriate, because formalizing their people with a focus on financial matters be partners in the institutions can require substantial capital and reserve context of projects aiming to create semiformal financial requirements. As formal financial intermediaries, they institutions such as cooperatives or NGOs. may become subject to usury laws or other regulations that limit the MFI's ability to operate. For example, Centro de Creditos y Capacitacion Institutional Growth and Transformation Humanistica, Integral y Sistematica para la Pequena y Microempres, a microfinance NGO in Nicaragua, For the most part, MFIs are created as semiformal insti- wanted to expand its microlending services (see box tutions--either as NGOs or as some form of savings 4.12). It considered creating a formal bank but found and credit cooperative. Given these institutional struc- that the regulatory environment in Nicaragua was not tures, they are often limited by a lack of funding suitable. This example demonstrates how local contex- sources and the inability to provide additional products. tual factors can constrain as well as facilitate the choice Unless the MFI is licensed and capable of mobilizing of institutional structure. deposits or has achieved financial sustainability and is thus able to access commercial funding sources, access Creating an Apex Institution is often restricted to donor funding. Furthermore, as MFIs grow the need for effective governance and the Some MFIs, particularly those partnering with interna- question of ownership arise. tional NGOs, may choose to create an apex institution as For MFIs that are structured as formal financial insti- a means of managing growth and accessing additional tutions, that is, development or commercial banks, the funding. An apex institution is a legally registered whole- issues of growth and transformation are not as signifi- sale institution that provides financial, management, and cant. Generally, growth within a formal institution can other services to retail MFIs. These apex institutions are be accommodated within the existing structure. Trans- similar to apex institutions for financial cooperatives. formation is rarely an issue. An exception to this is when However, in this case the MFI or more likely its foreign a commercial or development bank creates an MFI sub- partner or donor determines that a second-tier organiza- sidiary, such as the Banco Desarrollo in Chile; however, tion is required to provide wholesale funding and facili- this is a rare occurrence. Often formal institutions have tate the exchange of information and "best practices." the systems in place to accommodate this change. Rather than being member based, the apex institution is The following discussion is adapted from SEEP set up and owned by an external organization. Apex Network 1996a and focuses on semiformal institutions institutions do not provide services directly to microen- and ways in which they manage their institutional trepreneurs; rather, they provide services that enable growth. Three options are presented: maintaining the retail MFIs (primary institutions) to pool and access existing structure and managing growth within that resources. An apex institution can: structure; forming an apex institution to support the I Provide a mechanism for more efficient allocation of work of existing MFIs; and transforming to a new, for- resources by increasing the pool of borrowers and malized financial institution. This is followed by a dis- savers beyond the primary unit THE INSTITUTION 107 Box 4.12 Using the Nongovernmental Organization as a Strategic Step in Expansion C ENTRO DE C REDITOS Y C APACITACION H UMANISTICA , Banks are also required to pay a 30 percent tax on net profits Integral y Sistematica para la Pequena y Microempres or 4 percent on net worth, whichever is higher. (CHISPA), a program of the NGO Mennonite Economic Credit unions are regulated by the Ministry of Labor in Development Association (MEDA), targets the poorest of Nicaragua and essentially have no requirements regarding the economically active in the productive, service, and com- minimum capital, loan loss provisions, liquidity, legal mercial sectors in and around the secondary city of Masaya reserves, and equity ratios. They do not pay corporate income in Nicaragua. Established in 1991, the loan fund was valued taxes except when they pay dividends. A credit union can at US$740,000, with more than 3,500 active loans in 1996. offer both credit and savings but cannot provide checking The portfolio includes solidarity groups and individual accounts or credit cards. They can only offer services to their clients. Fifty-eight percent of the clients are women. own members but not the general public. However, forming In 1996 the portfolio had a 97 percent repayment rate, a a credit union would be highly risky due to the scale of opera- 27.5 percent nominal interest rate, a cost per dollar loaned tions that CHISPA has already achieved and the unpre- of less than 15 cents, and a 62 percent self-sufficiency level, dictable outcome of board selection. Normally, credit unions including financial costs, with full self-sufficiency expected are formed over a period of gradual growth based on accumu- by mid-1997. The program's business plan calls for scaling lated membership shares and savings, where they all have a up to reach almost 9,000 active clients by the end of 1999, stake in its solvency if not its profitability. increasing the loan portfolio from US$890,000 to US$1.6 NGOs, on the other hand, are currently neither regulat- million, and developing a local institution with a legal ed nor taxed, except by a few municipalities such as Masaya, framework that allows for credit and savings activities. which levies a 2 percent tax on gross profits. They can pro- Based on these growth projections, CHISPA considered vide a range of credit services without restrictions, although the options of forming either a bank or a credit union. While they are not able to offer any type of saving accounts, check- the regulatory environment for banks in Nicaragua is relatively ing accounts, or credit cards. However, there are plans to favorable, the US$2 million paid-in capital requirement is bring nonprofits under the supervision of the superintenden- double it's current portfolio, and to be efficient a bank should cy of banks within a year. If that happens, NGOs would be be leveraged at least four or five times. Therefore, the project- able to capture savings and qualify for some concessional ed scale of operations should really be about US$8 million capital currently available only to the banks. before contemplating a bank. There is also a legal reserve Incorporation as an NGO either as endpoint or interim requirement on savings deposits (15 percent for Cordobas form provides CHISPA with a strong legal and organization- accounts and 25 percent for dollar accounts), which need to al base to advance its mission of serving the economically be placed in non-interest-bearing accounts at the central bank. active poor. Source: Bremner 1996. I Conduct market research and product development institutional capacity building. Furthermore, by pro- for the benefit of its primary institutions viding wholesale funds in the marketplace, apex insti- I Offer innovative sources of funds, such as guarantee tutions remove the incentive for retail MFIs to funds or access to a line of credit from external mobilize deposits. sources There are other potential weaknesses of apex institu- I Serve as a source of technical assistance for improving tions (SEEP Network 1996a): operations, including the development of manage- I Vision and governance issues are made more complex ment information systems and training courses by the number of parties involved. I Act as an advocate in policy dialogue for MFIs. I The level of commitment to expansion and self-suffi- The experience of apex institutions has been mixed. ciency may vary among the members, affecting the Apex institutions that focus on providing funds to pace of expansion and the ultimate scale achieved. retail MFIs, often at subsidized rates, have found lim- The commitment to market-oriented operating prin- ited retail capacity to absorb those funds. What MFIs ciples may also vary, affecting the ability of the group most often need is not additional funding sources but to operate in an unsubsidized fashion. 108 MICROFINANCE HANDBOOK I Differential growth rates among the partners can "For the most part, microfinance apex institutions also strain their relationship, especially when their provide more than just liquidity in the market. needs for resources and technical support widen Usually the apex is set up when everyone agrees dramatically. that there is a drastic shortage of retail capacity. I Monitoring and supervision are essential to good The advertised objective of the apex is to foment performance but are made difficult by the number of the development of stronger retailers capable of partners. If there are weaknesses in financial report- reaching a much more substantial portion of the ing and management at the primary level, these can microfinance clientele." (Rosenberg 1996) adversely affect the second-tier operation. Apex institutions that focus on pooling member- I Unless both the primary institutions and the apex mobilized funds and on-lending these funds at market are efficient, the ultimate costs to the client can be rates to their members represent a far better approach high. There needs to be constant attention to issues than those wholesaling donor or government funds. Also, of productivity and performance. apex institutions should not be creating microfinance While there are many disadvantages to apex institu- retailers. Rather, they are most helpful when many exist- tions, if structured appropriately and set up with clear ing MFIs participate and benefit from doing so. and market-oriented objectives, they can add value and Apex institutions may also be useful in the following aid in the development of microfinance. situations (Von Pischke 1996): Box 4.13 Catholic Relief Services: Using the Apex Model for Expansion CATHOLIC RELIEF SERVICES SUPPORTED THE ESTABLISHMENT management of the credit banks. The corporation will seek a of the Small Enterprise Development Company, Ltd., a 51 percent equity investment in all new credit banks, with finance company in Thailand, to provide financial interme- the counterparts owning between 20 and 49 percent. diation and institutional strengthening services to its NGO Revenues should enable the corporation to achieve financial and (currently five) credit union counterparts implementing self-sufficiency toward the end of the third year. the village bank methodology. The company can facilitate In Peru Catholic Relieve Services is assisting eight mem- the use of collective savings as well as attract donors and bers of COPEME, a local consortium of enterprise NGOs, bankers. It is expected to achieve financial self-sufficiency by to form an Entidades de Desarrollo para la Pequena y 1999, when it will be managing a US$910,000 portfolio and Microempresa or finance company to serve the departments charging a 6 to 10 percent spread. In turn, the counterparts of Lima, Callao, Lambayeque, Arequipa, Piura, and Trujillo. will be lending funds to 175 village banks with almost Operating as a for-profit institution funded by commercial 11,000 members. investments, the company will provide financing services In Indonesia Catholic Relief Services is supporting the enabling the partners to provide 20,000 loans, averaging development of formal financial services at two levels: the between US$50 and US$200 and valued at US$425,000, by creation of 20 subdistrict credit banks, called Bank the end of three years. Perkredital Rakyats, and the formation of a national-scale Catholic Relief Services identifies four key attributes of limited liability company, called the Self Reliance these models: Corporation. Catholic Relief Services' traditional target I They are based on consistent involvement of experienced groups, the Usaha Bersamas (rural savings and credit NGOs, which, in every case, remain key actors with sig- groups), will receive financial services directly from the sub- nificant decisionmaking power. district credit banks. As for profit subsidiaries of Catholic I All parties have demonstrated a willingness to experiment Relief Services' NGO partners, the subdistrict credit banks with new speculative models that take advantage of many will help them increase the clients served from 16,000 to regulatory innovations. more than 30,000 in 5 years, with loans averaging US$150. I They avoid the large capital requirements of the com- Self sufficiency of the banks is expected in 24 months. The mercial bank form. Self Reliance Corporation will function as an Indonesian I Lending through diverse partners also mitigates risks registered company and will assist with the start-up and associated with these transformations. Source: SEEP Network 1996a. THE INSTITUTION 109 I When they are lenders of last resort, not necessarily in provide additional financial services to the target mar- situations of crisis but on the basis of cost. Retail ket. However, creating a formal financial institution lenders regard an apex institution as a good source of also implies additional costs and restrictions as the expensive funds and presumably use them sparingly MFI becomes regulated and supervised (see chapter 1). and only for highly important and profitable programs. Capital requirements may be much higher than antici- I When the apex institution fills in seasonally. pated, and unless the MFI has reached financial self- Agricultural-based retail lenders might want to borrow sufficiency it will be difficult (and costly) to attract seasonally as a means of managing cash flow. Again, equity investors and commercial debt (see the section the funds should not have to be provided by the apex below on accessing capital markets). And finally, the institution at anything less than commercial rates. MFI must develop the institutional capacity to manage I When the apex institution becomes a shareholder in a number of different products and services, mobilize the retail MFIs, in the expectation that this would resources (both debt and equity as well as human provide an attractive overall return. In this case the resources), and enhance management information sys- apex institution would expect to add value by provid- tems to adhere to regulatory reporting requirements ing expertise and oversight as well as funds in the and manage additional products (see the section below form of equity and quite possibly debt. on institutional capacity building). While transforma- I When retail lenders are not permitted to take tion may seem like an ideal path, MFIs should consid- deposits. Apex institutions could then play a useful er the substantial changes that are required when role if they added value through their terms and con- transforming to a formal financial institution. They ditions and behaved commercially. must also thoroughly examine their institutional capac- An apex institution appears to be most successful ity and determine if it meets the requirements for when a critical mass of strong MFI retailers already exists transformation (Appendix 1 provides a framework for and when it is focused on working with existing formal conducting this analysis). financial institutions that are "downscaling" to meet the The costs to convert into a formal financial institu- demands of low-income clients. (For further information tion are often exceedingly high: they include feasibility about apex institutions, see Gonzalez-Vega 1998.) and pre-start-up work, capital requirements, as well as the changes in management and systems that must be Creating a Formal Financial Intermediary implemented. Each MFI must consider the various types of formal institutions described above and deter- Recently the field of microfinance has focused on the mine which one best suits their needs (box 4.15). For transformation of financial NGOs into formal financial example, PRODEM's choice of the commercial bank institutions. This approach involves the transfer of the model for BancoSol was due to its great leverage poten- NGO's or cooperative's operations to a newly created tial and ability to offer savings and other financial prod- financial intermediary, while the original institution is ucts, matched by the right combination of political and either phased out or continues to exist alongside the new financial support. On the other hand, Accion intermediary. In most cases, the original MFI's assets, Comunitaria del Peru and COPEME, an association of staff, methodology, and systems are transferred to the NGOs partnered with Catholic Relief Services, decided new institution and adapted to meet the more rigorous to establish finance companies in accordance with spe- requirements of a financial intermediary. Various means cial legislation supporting these forms. Although unable may be established to determine a transfer price for the to capture savings, these structures do provide regulated assets and operations of the NGO or cooperative to the vehicles that have leverage potential and can begin new entity. It is imperative, however, that the transfer operations with lower capital (SEEP Network 1996a). price mechanism be transparent (box 4.14). Finally, if an NGO in the transformation process The rationale for developing a formal financial remains a separate institution, the two organizations (the institution is compelling: the potential to access both NGO and new financial intermediary) must develop a savings and commercial funding may help solve an working relationship that benefits both parties and that MFI's funding constraints and increase its ability to solidifies the bridge between them. Some useful bridging 110 MICROFINANCE HANDBOOK Box 4.14 Transformation from a Nongovernmental Organization to Financiera Calpía IN 1988 THE GERMAN AID INSTITUTIONS CREATED A ROTAT- of rapid growth, the fund was transformed into Financiera ing credit fund for the Association of Micro and Small Calpía, a small, strictly targeted group-oriented formal bank. Entrepreneurs (AMPES) of El Salvador. The objective of this The transformation was undertaken with the full support of project was to improve the supply of credit for small business the Association of Micro and Small Entrepreneurs, which, people on a lasting basis. Plans were made with the association through its Fundación Calpía, is now the main shareholder of to create a stable and professional institution that could ulti- Financiera Calpía. The other shareholders are local and inter- mately be transformed into a small target-group-oriented for- national development institutions. mal financial institution. Transformation and formalization were necessary for two From the beginning, credit operations were organized reasons. To satisfy the high demand for small and very small separately from the association's other business, both to loans, the institution needed the right to accept savings assure professionalism and also to limit undue personal deposits from its customers and to have access to the inter- influence. While the association was the owner of AMPES- bank market. And because of the challenges and dangers of Servicio Crediticio, it did not run the operations. Also, the the extremely dynamic expansion of its operations, the institu- donor institutions insisted that the purchasing power of the tion needed--and wanted--the increased stability and tighter fund be maintained. Accordingly, interest rates were set at an control that the new status provides by making Calpía still unusually high level (effective rates between 25 percent and more independent from the association and putting it under 40 percent). Finally, strict lending policies were implement- the formal control of the superintendency of banks. ed and a system of financial incentives for loan officers was Currently, Calpía is the most important specialized created, which kept loan losses below 1 percent. lender to the small business community of El Salvador. It With external funding from the Inter-American has opened eight branches to date. Because of its high posi- Development Bank the loan portfolio and the number of bor- tive (inflation-adjusted) return on equity and its institutional rowers doubled every year. At the end of 1996 the number of stability, other banks in the country regard it as highly credit borrowers was around 20,000 and the loan portfolio had worthy, improving the availability of loan funds to on-lend grown to the equivalent of US$14 million. During this phase to the target population. Source: Contributed by Reinhard Schmidt, Internationale Projekt Consult GmbH (IPC). mechanisms include the following (SEEP Network Box 4.15 Catholic Relief Services' Guatemala 1996a): Development Bank I Overlapping directorates I Cost sharing of head office functions, branch space, CATHOLIC RELIEF SERVICES' PARTNER IN GUATEMALA, Cooperative Association for Western Rural Development, insurance, and so forth is planning to establish a development bank, because its I Joint resource mobilization, with the NGO attract- organizing principles are more amenable than a commer- ing the social investment capital that the financial cial bank to the social development mission that the asso- intermediary may not qualify for directly ciation supports. This mission involves: I Coordinating policies for product development, tar- I A focus on the very poor, the core constituency of get areas, and populations. Catholic Relief Services and its partners I An emphasis on savings I Accessibility for NGOs and popular organizations in Governance and Ownership terms of the initial investment requirements and administrative operations MFIs, particularly those set up as financial NGOs, are I Easy replicability. often formed by a visionary. These visionaries are not This last characteristic is seen as essential both to sup- usually interested in profits but rather have a social port rapid scale-up and to encourage diversification rather objective to improve the lives of low-income members than financial concentration. of their community. Only after the MFI begins to grow Source: SEEP Network 1996a. does the visionary reach the conclusion that the MFI needs to adopt a more business-like approach. This is THE INSTITUTION 111 often the result of limited donor funds, demands of the planning for the executive. In the transition from a target market, and a general shift in the field of microfi- small, growing entrepreneurial organization to an nance toward formalized institutions. established institution, effective governance ensures As the MFI grows and management systems are devel- that the company survives. Governance moves an oped, the need for governance arises to ensure effective institution beyond dependency on the visionary. management of the MFI and, potentially, to attract people The board should comprise members who have a with much-needed skills (usually from the private sector). number of different skills, including financial, legal, While governance does not always result in a change of and managerial. In particular, representation from the vision for the MFI, it does establish a means of holding private sector is important. Also, the ability to critically management accountable. Furthermore, as the MFI grows analyze management's plans as well as provide effective the issue of ownership becomes apparent. This is particu- guidance is paramount when selecting a board. An MFI larly important as the MFI begins to create a more formal- must define the following: ized structure. I The role of board members both within the board and with regard to external alliances GOVERNANCE. (Adapted from Clarkson and Deck 1997.) I The desired areas of expertise Governance refers to a system of checks and balances I The existence of committees to oversee specific areas whereby a board of directors is established to oversee the of operation management of the MFI. The board of directors is I Term limits for board seats responsible for reviewing, confirming, and approving I The process for replacing board members the plans and performance of senior management and I The role of the executive director in selecting board ensuring that the vision of the MFI is maintained. members Management is responsible for the daily operations of I The optimum number of board members putting the vision into action. I Mechanisms to evaluate the contribution of individ- The basic responsibilities of the board are: ual members. I Fiduciary. The board has the responsibility to safe- Board members must be provided with and agree to guard the interests of all of the institution's stake- clear and common objectives. Furthermore, it is important holders. It serves as a check and balance to ensure that members be independent from the MFI and be chosen the MFI's investors, staff, clients, and other key for their expertise rather than their own interests or political stakeholders that the managers will operate in the agendas or those of the senior management. Board mem- best interest of the MFI. bers should act in such a way that they create accountability I Strategic. The board participates in the MFI's long- and enable stakeholders to trust one another. Governance term strategy by critically considering the principal gives shareholders, donors, governments, and regulators risks to which the organization is exposed and approv- confidence that managers are being appropriately super- ing plans presented by the management. The board vised. Thus board members should not receive any person- does not generate corporate strategy but instead al or material gain other than the approved remuneration. reviews management's business plans in light of the institution's mission and approves them accordingly. OWNERSHIP. (Adapted from Otero 1998.) It is the owners I Supervisory. The board delegates the authority for of the MFI that elect (or at times compose) the governing operations to the management through the executive body of the institution. Owners, through their agents on director or chief executive officer. The board super- the board, hold management accountable. Ownership is vises management in the execution of the approved an important but often nebulous issue for MFIs, particu- strategic plan and evaluates the performance of man- larly as many are funded with donor contributions. agement in the context of the goals and time frame Neither formal financial institutions nor NGOs have outlined in the plan. owners per se. Formal MFIs have shareholders who I Management development. The board supervises the own shares that give them a residual claim to the assets selection, evaluation, and compensation of the of the MFI if there is anything remaining after it has senior management team. This includes succession discharged all of its obligations. Shareholders have the 112 MICROFINANCE HANDBOOK right to vote their shares to elect board members, who in turn control the company. Having shareholders Box 4.16 BancoADEMI Ownership Structure results in clear lines of accountability between the board ON SEPTEMBER 11, 1997, THE MONETARY BOARD OF members and the MFI. THE Dominican Republic approved a banking license for NGOs do not generally have shareholders; rather, the Banco de Desarollo ADEMI, S.A. The new develop- management usually elects the board members. This can ment bank focuses on the small and medium-size enter- (but does not always) result in a conflict of interest if prise sector, making loans in the range of US$35,000 to management selects board members who will conform US$300,000. BancoADEMI also offers savings facilities to the interests of senior management. Furthermore, to the general public through time deposit accounts and NGO board members do not usually fulfill the board's financial certificates. About 72 percent of the shares in BancoADEMI fiduciary role by assuming responsibility for the institu- are owned by the Association for the Development of tion's financial resources, especially those provided by Microenterprises (ADEMI), the original NGO that donors. spun off its small enterprise loans to the bank and As MFIs formalize their structures (that is, change continued to serve the microenterprise sector. Another from being an NGO into a formal financial institu- 10 percent of the shares are owned by association tion) and begin to access funding beyond the donor employees and the remainder by several individuals community, the "owners" or those that have a financial who have been active in the development of the stake in the institution can change. If the NGO association. remains as a separate entity, it often owns a majority of The bank has an authorized capital of $150 million the shares of the new institution. In spite of this major- pesos, or about US$10.7 million, and an initial paid-up ity ownership, it is important that the relationship capital of just over $100 million pesos (US$7.1 million). between the MFI and the NGO be kept at arm's With the transfer of assets from ADEMI, the new insti- tution was projected to make a profit in its first year of length and include a transparent and clear system of operations. transfer pricing. "Owners" of formalized MFIs can generally be Source: Benjamin and Ledgerwood 1998. divided into four categories: I NGOs I Private investors return or positive impact in the lives of the clients I Public entities (table 4.2). I Specialized equity funds. Owners may be heavily involved in the operations of All of these owners are concerned with receiving an the MFI or may take on a more passive role. Owners adequate return on their investment. However, NGOs that play a more active role must have adequate skills and public entities may have other priorities as well. and the ability to spend the time required. Owners are For example, they may be concerned with a social often also called upon to access additional capital, par- Table 4.2 What Is at Stake for Microfinance Owners? Nongovernmental Private Public Specialized organization investors entities equity funds Moral responsibility Return on investment Political concern Return on Institutional mission Capital preservation Entry into the field investment Return on investment Sense of social Return on investment Institutional mission Long-term concern responsibility Good project Long-term concern Institutional creditability or image Source: Otero 1998. THE INSTITUTION 113 ticularly equity capital. As owners they may wish to "research and development" for commercial lenders con- maintain their stake as the MFI grows and takes on sidering entry into a new market (either lending to MFIs additional capital. This implies that they also must have or directly to clients). Guarantee funds are usually continued access to capital. designed to leverage resources. For example, a guarantee fund of US$5 million may encourage banks to lend Accessing Capital Markets US$10 million to microentrepreneurs, thereby leveraging the fund's resources by a factor of two to one. The more The majority of MFIs fund their activities with donor or effective the guarantee mechanism the higher the lever- government funding through grants or concessional age factor becomes (the guarantee fund backs an increas- loans. However, it is becoming evident that donor fund- ingly large loan portfolio). Ultimately, the aim is to ing is limited. As MFIs expand and reach a critical stage gradually transfer the risk from the guarantee mechanism of growth, they find that they cannot sustain their to the participating financial institutions. growth with only donor support. Some are beginning to For example, ACCION's Bridge Fund provides its access capital markets. There are various ways that an NGO affiliates with access to commercial loans that they MFI can access new capital, including: otherwise might not have. The Bridge Fund puts cash I Debt accessed through guarantee funds, loans, and deposits, securities, or letters of credit with commercial deposit mobilization banks, which in turn on-lend to the NGO affiliate at I Equity market rates. Eventually, some affiliates have been able to I Equity investment funds borrow directly without the use of the guarantee scheme. I Socially responsible mutual funds However, the costs of guarantee schemes must be I Securitization of the loan portfolio. considered relative to the benefits (box 4.17). Vogel and Adams (1997) claim that there are three categories of ACCESSING DEBT. For the most part it is necessary to be costs that accompany loan guarantee programs: the costs financially self-sufficient to access commercial sources of of setting up the program, the costs of funding the sub- funds. However, with the backing of guarantees by sidy needed to energize and sustain the program, and donors or international NGOs, some MFIs, if they earn the additional cost incurred by the financial system of enough revenue to at least cover their cash costs, may be running and participating in the guarantee program. able to leverage their donor funds with commercial loans The benefits of a loan guarantee program are the addi- in amounts equal to the donor funds. tional lending induced by the transfer of part of the Guarantee funds are financial mechanisms that reduce lender's risk to the guaranteeing organization. Both bor- the risk to a financial institution by ensuring repayment rowers and society benefit from the increases in net of some portion of a loan (adapted from Stearns 1993). income realized by borrowers. Vogel and Adams suggest Guarantee funds are used to encourage formal sector that the benefits are hard to measure, because it is diffi- bank lending to the microenterprise sector. They can be cult to determine if the guarantee scheme resulted in used to guarantee a loan made by a commercial bank to true additionality (that is, additional lending to the tar- an MFI, which then on-lends funds to its clients, or get market). Furthermore, substitution can also occur loans from a bank directly to microentrepreneurs. There whereby the bank simply transfers part or all of the qual- are three types of guarantee funds that cover the risks of ifying portion of its existing loan portfolio to the guar- making loans. These risk-related design features are antee programs or to an NGO, in this way benefiting some of the most important determinants of the accep- from the subsidized guarantee program that takes bor- tance and use of a guarantee mechanism by banks. They rowing clients away from other NGOs not benefiting are guarantees covering: from the subsidies. I A percentage of the loan principal At the very least, few guarantee schemes are financial- I A percentage of the loan principal and the interest lost ly sustainable without some form of subsidy. However, I A certain amount of the loan (say, the first 50 percent). they appear to be more beneficial than direct subsidies to By reducing risk and transaction costs faced by the clients. (For further information on guarantee schemes financial institution, guarantee funds can function as for microenterprises see Stearns 1993, and on guarantee 114 MICROFINANCE HANDBOOK schemes for small and medium-size enterprises see MFIs can access commercial debt by borrowing Gudger 1997; Levitsky 1997). directly from commercial banks or by issuing financial As an MFI continues to expand and eventually reach- paper in the market place (box 4.19). es financial self-sufficiency (revenue earned covers all costs, including inflation and an imputed cost of capital; A CCESSING E QUITY . Capital markets can also be ac- see chapter 8), its ability to leverage its equity (donor cessed by selling shares of ownership (equity) of the funds and retained surpluses) increases. If it becomes a MFI. For this to be possible the institution must be a formally regulated financial institution, it can achieve formal financial intermediary with shareholders. Un- leverage ratios of up to 12 times its equity by borrowing like debt instruments, equity does not have a fixed directly from commercial sources or mobilizing deposits yield or maturity. Rather, equity investors invest in (if regulated to do so). However, mobilizing savings the profitable future of the MFI and expect a share of requires substantial changes to the MFI, which are dis- the returns. Because the microfinance industry has not cussed fully in chapter 6. yet developed to the point where there are many in- An MFI is ready to access commercial financing when vestors who are aware of the potential of MFIs, it may it has built an equity base through past donor grants and has a positive net worth (see box 4.18). Box 4.18 Key Measures for Accessing Commercial Financing Box 4.17 Guarantee Scheme in Sri Lanka MFIS ARE READY TO ACCESS COMMERCIAL FINANCING IF THE GUARANTEE SCHEME IN SRI LANKA WAS STARTED IN they: 1979 and operated by the central bank. The scheme cov- I Have perfected their service delivery methods and ered 60 percent of the loan amount and was instituted as product design to respond to the demands of their an incentive for the commercial banks to participate in market in a rapid and efficient way, ensuring an the credit line channeled through the National increased volume of operations and repeat borrowing Development Bank, financed by the World Bank. I Have a strong sense of mission and a sound governing Between 1979 and 1995 the World Bank approved structure that is free from political interference, so that four loans for small and medium-size enterprises for a they can make policy decisions that protect their total of US$105 million. Repayment rates on the financial health subloans given from the first two World Bank loans were I Have a management team that focuses on efficient ser- low (70­73 percent), but this improved significantly to vice delivery and productivity, on profits rather than more than 90 percent in the later bank loans. volume, and sets productivity goals and incentive In the earlier years of operation claims were high--7 schemes percent of guarantees claimed in 1979­80 and 13 percent I Have information systems that produce clear, accu- in 1982­85. Substantial changes were made and claims rate, timely, and relevant information for management dropped to 2 percent in 1991­95. decisionmaking and that focus on well-developed loan In 1979­96, 18,500 guarantees were given, and there tracking and financial reporting systems, reporting on were subsequently 696 claims for guarantee payment (89 costs and income both on a profit-center basis and for rejected and 417 paid out) up to the end of 1994, for a the MFI as a whole total of US$1.36 million. Outstanding guarantees for I Have a record of achieving high levels of financial per- which the central bank has a contingent liability were esti- formance, of incorporating appropriate pricing poli- mated at a maximum total claim of US$0.88 million. The cies based on the full cost of delivering the services, actual and potential total loss on guarantees paid on bad and of maintaining the value of donated equity debt was US$2.2 million. With 18,850 guarantees this I Maintain low levels of delinquency (well below 5 per- was a cost to the state of US$118 per entrepreneur assist- cent to 8 percent of outstanding portfolio, with loan ed. This amount was estimated to be more than offset by loss rates below 2 percent) to ensure optimum income additional tax generated. and prevent asset erosion. Source: Levitsky 1997. Source: Clark 1997. THE INSTITUTION 115 Box 4.19 Accessing Capital Markets by Issuing Financial Paper ISSUING FINANCIAL PAPER REFERS TO A FORMAL IOU GIVEN should reflect a methodology proven by microlending by the MFI to investors in exchange for their funds. activities deployed since 1987 by BancoSol's predecessor, Financial paper is issued with a fixed date on which the debt PRODEM. This long track record of success is evidenced will be paid (maturity) and a preestablished interest pay- by a consistent historical loss record of less than 1 percent ment (yield). The financial market then evaluates the paper of portfolio. On that basis the yield offered by BancoSol in terms of yield versus risk, comparing it with every other paper was highly attractive relative to its real rather than investment opportunity available. To succeed in the finan- perceived risk. cial market an MFI must offer at least an equally attractive It is not necessary to become a regulated financial insti- yield to alternatives that are perceived to have the same level tution before issuing financial paper. For example, in 1995 of risk. Accordingly, for a new entrant in the market whose the ACCION affiliate in Paraguay, Fundación Paraguaya risk profile is unknown, a higher yield may be necessary. de Cooperacion y Desarrollo, as an NGO issued 350 mil- Establishing the correct risk is especially important for lion guaranies in debt through the Asuncion Securities an MFI operating in an industry that is itself unknown. Exchange. For example, in 1994, when ACCION was placing paper In addition, it may not even be necessary for the issuing issued by BancoSol to North American financial institu- institutions to be 100 percent financially self-sufficient to tions, investors began with an expectation of returns issue paper, although it is clearly preferable, as in the case of reflecting both country and venture capital risk. While the Grameen Bank in Bangladesh. In the debt market it is Bolivian country risk was unquestionably a relevant consid- possible to find buyers as long as there is the firm expecta- eration, ACCION argued successfully that the appropriate tion that the cash flow (whether generated by operations or business risk, instead of being compared to venture capital, donations) is sufficient to service the debt. Source: Chu 1996b. be some time before equity investors play an im- EQUITY INVESTMENT FUNDS. Equity investment funds portant role in the funding mechanisms of MFIs. provide equity and quasi-equity (subordinated debt) to However, some equity investment funds are being selected organizations. ProFund is one such investment developed specifically to invest in growing, sustainable fund that was set up for the sole purpose of investing in MFIs. expanding MFIs in Latin America (box 4.20). Efforts Box 4.20 ProFund--an Equity Investment Fund for Latin America P RO F UND IS AN INVESTMENT FUND INCORPORATED IN (ProFund Internacional 1996­97). ProFund's sponsors Panama and administered from Costa Rica. It was created to (original investors) are Calmeadow, ACCION International, support the growth of regulated and efficient financial inter- FUNDES (a Swiss foundation), and Société d'Invest- mediaries that serve the small business and microenterprise issement et de Développement International (SIDI), a market in Latin America and the Caribbean. It operates on a French NGO consortium. Other investors include the profit basis, providing equity and quasi-equity to eligible International Finance Corporation, the International financial institutions so that they can expand their opera- Development Bank's Multilateral Investment Fund, and the tions on a sustainable and large-scale base. By supporting the Rockefeller Foundation. growth of efficient intermediaries, ProFund aims to achieve By 1997 ProFund had made seven investments, with superior financial returns for its investors through long-term total commitments of US$12.2 million. MFIs invested in capital appreciation of the MFIs in which it invests. include Accion Comunitaria del Peru, Banco Empresarial At the end of 1997 ProFund was capitalized by a group (Guatemala), BancoSol (Bolivia), Banco Solidario (Ecuador), of 16 investors who have subscribed a total of more than Caja de Ahorro y Prestamo Los Andes (Bolivia), Finansol US$22 million with 63 percent of the capital paid in (Colombia), and Servicredit (Nicaragua). Source: ProFund Internacional 1997. 116 MICROFINANCE HANDBOOK are under way to establish similar funds in Africa and Asia. Box 4.22 DEVCAP--a Shared-Return Fund DEVCAP (DEVELOPMENT CAPITAL FUND) IS A MUTUAL SOCIALLY RESPONSIBLE MUTUAL FUNDS. There are two fund designed to provide financial returns to its investors types of socially responsible mutual funds: screened and as well as provide its members with revenue. DEVCAP shared-return funds. With screened mutual funds, man- represents a consortium of MFIs that invested initial capi- agers screen companies for social criteria. Profits are paid tal to develop an asset base for the fund. Additional capi- to shareholders who choose to invest in these funds tal is provided by public investors. The assets of the fund because they want to support socially responsible compa- are invested in a leading socially screened portfolio (social nies. The Calvert Group mutual fund is an example of a equity portfolio) based on the Domini Social Index, which consists of the stocks of approximately 400 U.S. screened mutual fund (box 4.21). companies chosen for their corporate and social responsi- Shared return funds are mutual funds owned by mem- bility as well as their financial performance. The return ber organizations (MFIs). Shareholders agree to donate earned on the investments is shared between the investor (share) a percentage of the return to the member organi- and DEVCAP. The tax-deductible donation made by the zations. DEVCAP (Development Capital Fund) is an investor ranges from 50 percent to 100 percent of the example of a shared mutual fund owned by MFI mem- return earned. ber organizations (box 4.22). The return earned by DEVCAP is shared among its member MFIs. Members include Appropriate Technology SECURITIZATION. Securitization links microfinance insti- International, Catholic Relief Services, Save the Children, tutions to capital markets by issuing corporate deben- and the Seed Capital Development Fund. tures backed by (and serviced by) the MFI's portfolio Source: Development Capital Fund. (adapted from Chu 1996a). The structure requires the creation of a single purpose corporation, which buys the microenterprise portfolio and capitalizes itself by issuing its face value. The amount of discount is based on the debentures into the capital market. quality of the portfolio. For example, if the single pur- The purpose of creating a single purpose corporation pose corporation purchases US$105 of portfolio for is to acquire the microenterprise portfolios of known only US$100, the additional US$5 is held in a reserve to entities without taking on other risks. The equity of the protect the corporation against any portfolio losses. This single purpose corporation comes from the microfinance reserve amounts to roughly 5 percent of the portfolio. In organization and its partners. The single purpose corpo- this case the historic rate of default must be lower than 5 ration uses its funds to purchase the portfolio from an percent and likely less than 2 percent (the reserve should MFI, which it does at a discount--that is, for less than be greater than the historic loan loss to provide comfort for investors). In addition, the single purpose corpora- tion has substantial equity so anyone who buys paper Box 4.21 The Calvert Group--a Screened Mutual Fund issued by it is highly protected. ONE TO 3 PERCENT OF THE CALVERT GROUP'S ASSETS ARE The single purpose corporation then sells commercial invested in "high social impact" instruments such as MFIs. paper through its financial partner, a leading brokerage Eligible institutions must have loan capital of at least firm in the local financial market. For securitization to US$300,000 from diverse sources and demonstrate the succeed, the partner must be significant and well need for more capital. The cost of borrowing by the MFIs respected with an established network of buyers. The is generally below market rates by 3 percent or 4 percent. partner receives a management fee and an investment Renewable loan terms are for one year, and the average loan banking fee. amount is US$100,000. Investors risk receiving a lower The object of securitization is to increase the availabil- return than from other mutual funds, depending on the ity of funds and, at the same time, reduce the cost of success of the overall investment strategy of the fund. funds. Grameen Bank and ACCION International and Source: Calvert Group. its affiliate, the Fundación Ecuatoriana de Desarollo (FED) in Ecuador, are two MFIs that are either consider- THE INSTITUTION 117 ing or have successfully achieved securitization of their improvements. For more formalized institutions this may portfolios. (For an explanation of the structure that FED require a greater focus on client needs through improved used for securitizing its portfolio, see Chu 1996a.) product development and human resource management. Because both debt and equity represent a willingness For NGOs this may include the adoption of a formal to accept financial risk, the fundamental factor that deter- business planning process, improved financial and pro- mines the sustainability of access to capital markets is the ductivity management, and new funding sources. institution's credibility as perceived by investors, whether Since the purpose of this handbook is to improve the they are lenders or shareholders. To link with capital mar- institutional capacity of MFIs, it is useful to briefly high- kets MFIs must provide clear and solid answers to such light the institutional capacity issues that many MFIs are critical questions of governance as (Chu 1996b): facing and to identify where these issues are addressed in I Can MFIs that are NGOs or have NGO owners pro- this and following chapters. vide financial markets with the assurance that they will Institutional capacity issues include (SEEP Network make decisions with the same standards of prudence as 1996a): enterprises that have traditional shareholders with com- I Business planning. An MFI needs to be able to trans- mercial monies at risk? late its strategic vision into a set of operational plans I Will NGOs be able to resist the temptation to let based on detailed market and organizational analysis, nonfinancial considerations--whether lofty or base-- financial projections, and profitability analyses. overwhelm return considerations, because, in the Appendix 2 provides an outline for business absence of commercial shareholders, they are ulti- planning. mately accountable only to their institutional mission? I Product development. An MFI must be able to diversi- I Can boards of directors of NGOs effectively control fy beyond its original credit products (usually indi- management? vidual, solidarity group lending, or both) into other I Will donor agencies be able to distinguish between areas such as savings, which can provide desired ser- those NGOs that can live up to the required stan- vices to clients and accommodate their growth. An dards and those that cannot? MFI must also be able to price its products on the I In some cases, particularly in Asia, clients become basis of operating and financial costs and demand in shareholders as part of the methodology. Will such the marketplace. This requires periodic review to atomized owners be able to participate in a meaning- ensure appropriate pricing. (For more on the devel- ful way, and, if so, will the cumulative effect of opment and pricing of credit and savings products, minuscule portions of ownership lead to commercial- see chapters 5 and 6 respectively.) ly sound results? I Management information systems. As MFIs grow one I Will NGOs that have generated their equity through of their greatest limitations is often their man- grants and donations be able to exhibit the same disci- agement information system. It is imperative that an pline and rigor as a private investor? For example, can MFI have adequate information systems for financial an institution like ACCION invest in BancoSol with and human resource management. For example, funds provided by donor agencies and behave like a establishing an effective incentive system depends commercial investor? critically on the existence of a good management Answers to these questions will become apparent as information system, which then allows management the field of microfinance grows and more MFIs begin to to track the various indicators of performance that it access capital markets. wants to reward. (For a more detailed discussion of management information systems, see chapter 7.) I Financial management. Improvements in accounting Institutional Capacity Building and budgeting are often required to monitor loan portfolio quality, donor subsidies, and the growing Regardless of institutional type and ways of managing volume of operations. Additional skills are required in: growth, all MFIs need to periodically review their institu- I The adjustment of financial statements for subsi- tional capacity and consider where they might make dies and inflation 118 MICROFINANCE HANDBOOK IPortfolio risk management, including appropriate tion about efficiency and productivity enhancement see risk classification, loan loss provisioning, and chapter 10.) write-offs The credit unions in Guatemala present a good I Performance management, including profitability example of the importance of periodic reviews and a and financial viability focus on institutional capacity building (box 4.23). I Liquidity and risk management, focused on effec- tively administering portfolios largely based on short-term loans Appendix 1. MFI Operational Review I Asset and liability management, including appro- priate matching of amortization periods in rela- The following outline is from SEEP Network 1996b and tion to the loan portfolio and asset management was developed by Calmeadow, Toronto, Canada. and capital budgeting processes that are inflation sensitive. Definition (For a discussion of financial management see chapters 8 through 10.) The operational review is a tool used to evaluate the insti- Efficiency and productivity enhancement. MFIs must tutional maturity of a microfinance organization. An be able to operate in a way that best combines standard- operational review helps an NGO planning to transform ization, decentralization, and incentives to achieve the into a self-sufficient microfinance intermediary to evaluate greatest output with the least cost. They must also devel- its readiness. This tool draws on lessons from the formal op systems for staff recruitment and selection, compen- banking sector, from Calmeadow's own observations sation, training, and motivation to support staff working with microfinance institutions, and from recent commitment and accountability. (For further informa- publications on institutional development. Box 4.23 Credit Unions Retooled I N MID -1987 THE U.S. A GENCY FOR I NTERNATIONAL ing, encouraged borrowing, and forced those who saved to Development contracted the World Council of Credit subsidize those who borrowed. Over time, the response of Unions to provide technical and financial assistance to the many members was to find a way to borrow-out their share National Credit Union Federation. During the course of the savings with no intention of ever repaying the loan. project a major restructuring of the Guatemalan credit union In the new model created through the project, the focus movement took place. New financial tools and disciplines was shifted from share savings to voluntary deposits with a were developed and tested in 20 community credit unions competitive rate of return. Subsidized external financing was countrywide. Their use transformed the credit unions into eliminated, and because all funding came from internally modern, effective financial intermediaries, capable of com- mobilized deposits, loans were priced based on market rates. peting in commercial financial markets. Loan approvals were based on the repayment capacity of the Guatemalan credit unions had followed a traditional devel- borrower. Earnings were capitalized rather than distributed opment model for 25 years. This model was based on the the- to members, so that the credit unions' capital maintained its ory that the rural poor lacked the resources necessary to save value. The credit unions undertook a market-based, results- and thereby fuel their development potential. International oriented business planning process. And finally, improved donors responded to the lack of local resources by providing financial reporting control and evaluation was instituted the credit unions with external capital at subsidized rates for through intensive training. on-lending to their members. Loan sizes were based on a mul- While the process was slow and difficult and required sig- tiple of the amount of shares a member held in the credit nificant changes in the attitudes of elected leaders, employ- union rather than on the member's repayment capacity. ees, and the member-owners, the result was greater financial Shares earned very low returns, so the only purpose in saving independence, stronger and safer credit union operations, was to access a loan. The traditional model discouraged sav- and superior, member-driven financial services. Source: Richardson, Lennon, and Branch 1993. THE INSTITUTION 119 The Tool I Indicators of health--Standards against which to eval- uate the organization. As we gain experience with The operational review provides the user with guidelines applying the tool, these standards will become more on how to evaluate a microcredit program in seven key quantitative. areas: corporate governance, market and clients, credit I Transformation issues--Issues requiring thought and methodology, distribution, human resource management, attention if the organization is contemplating trans- computerization, and financial management. Each topic forming into a regulated financial intermediary. includes three sections: documentation, indicators of health, and transformation issues. 1. CORPORATE GOVERNANCE Documentation How the Tool Works I History, institutional type/incorporation, vision and goals The review involves a thorough documentation of the I Board of directors (names and professions/areas of operating procedures and institutional characteristics of expertise, regularity of meetings; committees of the the organization and an evaluation of these factors board, term of members, and board renewal proce- against performance levels currently being achieved by dures) leading micro-finance institutions worldwide. I Managing director and senior management (names An operational review takes about 2 weeks (or 10 and experience; years of service) working days) of on-site field work to gather the informa- tion required from an intermediate level organization Indicators of health (more than 2,000 clients). This includes collecting and I Vision and goals clearly articulated and consistent synthesizing a large amount of data and materials as well as I Evidence of a strong sense of mission interviews with all senior managers, a comprehensive sam- I Board well constituted, with a mixture of financial ple of other head office staff and field staff, and clients. services and other business Two useful products can result from an operational I Legal, marketing, fundraising, and community devel- review: opment skills 1. A detailed documentation of the organization that can I Board members well informed about the global be given to visitors, researchers, funders, and groups microfinance sector and committed to the vision of who may want to start a microcredit program and are the organization looking for lessons from current practitioners. I Board members who are active in committees 2. An evaluation of the organization's readiness for rapid I Depth of experience and strong leadership demonstrat- growth and transformation, comprising of two sec- ed by the managing director and senior management tions--one on institutional capacity in general and I Strong commitment by senior management toward one on issues particular to transformation. the vision of the organization. Sufficient breadth and depth of senior management so organization is not Operational Review: An Outline vulnerable to the loss of any one individual. I A business planning process and strategic plans in 1. Corporate governance place. 2. Markets and clients I If microfinance is just one activity of the organization, 3. Credit methodology the financial services are run and monitored indepen- 4. Distribution dently, unencumbered by other nonfinancial activities. 5. Human resource management 6. Computerization Transformation issues 7. Financial management I What new corporate structures are being contemplat- For each topic above, this outline includes three sections: ed? Do these make sense? Does transformation make I Documentation--A checklist of items to be collected sense given the institutional maturity of the organiza- and documented during the review. tion? Is transformation being contemplated for the 120 MICROFINANCE HANDBOOK right reasons or should other options be considered at I New product development is viewed as a necessary and this point in time? ongoing activity in the organization and accountability I Will the new structure require a different mission, or is for product enhancements and new product develop- the new structure just a different vehicle for imple- ment is clearly assigned to appropriate personnel. menting the same mission? (Business versus social development mission.) Transformation issues I How will ownership of the new organization be struc- I The size of target markets should be sufficiently large tured? Will new partners be required? What type of to sustain a self-sufficient microfinance institution. partners are most desirable? What combination of for- I Will transformation require a redefinition of the tar- profit and not-for-profit institutions? get market or a departure from the current target? I Do all board members fully understand and support Will it threaten the current market focus? the goal of transformation? I Will the board need to be reconstituted for the trans- 3. CREDIT METHODOLOGY formed organization? Documentation I Do all senior managers fully understand and support I Product design (loan amounts (sizes), terms, and the goal of transformation? repayment frequency; pricing; interest and fees; savings products and requirements; insurance or other funds) 2. MARKETS AND CLIENTS I Eligibility criteria (gender, age, years of experience in Documentation business, collateral, guarantors, peer groups, other col- I State of small and microenterprise sector in the coun- lateral substitutes; legal, cultural, or other regulatory try (how large? how sophisticated? how literate?) requirements) I Barriers faced (cultural and political factors) I Credit delivery procedures (promotions, screening, I Suppliers of microcredit in the country; competitors orientation, group formation, applications and dis- I Target markets (size, type and subsector, gender, geo- bursements, repayments, renewals) graphic coverage) I Delinquency management (reports and monitoring; I Policies on client graduation enforcement/follow-up procedures; incentives and I Profile of current borrowers (size of microenterprises sanctions: i) rewards for timely repayment, ii) penal- served; three most common activities by gender; dis- ties for late payment, iii) consequences of default, tribution by sector: trade, manufacturing, service; insurance/use of savings, policies on rescheduling, proportion of new versus repeat borrowers; propor- refinancing, and death of borrower) tion of female borrowers) I New product development (how done and who is Indicators of health accountable) I The basic methodology appears to be well-tested and stable. Indicators of health I Product design appears to suit business needs of bor- I Target markets are clearly articulated; ideally only one rowers. or two primary targets. If more than one target, the I Eligibility criteria are easy to meet--not too restrictive. different markets should work well together within I For group-based programs, group cohesion appears to one delivery structure. be strong. I Profile of current borrowers should match the targets. I Delivery procedures minimize inconvenience and I Low client drop out rates/high renewal rates. Clients transaction costs for client. express a high level of satisfaction with the service. I Delivery procedures minimize chances for fraud. New clients approach the organization voluntarily. I Branch procedures are routine/standardized and facili- I A strong "client service" culture is evident. tate loan officer work flow--they make it easy for I Client and market research is conducted on a regular loan officers to conduct their work effectively. basis; innovation in meeting the evolving needs of I Strong relationship is evident between clients and clients is valued. loan officer. THE INSTITUTION 121 I Delinquency management is tightly enforced. Pro- I Apparent culture of organization--hierarchical or cedures for late payment follow-up are triggered auto- participatory. matically after certain time periods. Loan officers not I Apparent employee satisfaction levels; turnover lev- involved in collections once client is no longer negoti- els--head office and field offices. ating in good faith. Incentives and sanctions appear to I Existence of an employee union or association. be effective. Late payments are minimal. I Existence of an employee newsletter. I Repayment/recovery rates are calculated according to I Appraisal system. international standards and are above 85% (for cumu- I Promotion policies/succession planning. lative repayment, including arrears) I Other innovative personnel policies. I Portfolio at risk rates are calculated according to inter- national standards and are <10% Indicators of health I Head office size and structure that is appropriate for Transformation issues the size of operations currently and in the near future. I Will transformation require any change to the current I Branch organization structure is streamlined, mini- methodology? How will this be perceived by the client? mizes overhead, and does not put too much stress on managers. 4. DISTRIBUTION I Span of control for managers is close to the ideal of Documentation five to seven direct reports each. I Type, number, and location of field offices (diagram I Sufficient number of qualified candidates applying to if desirable) advertised positions. Appropriate "incubators" for I Outreach / branching / growth strategy staff identified, for example, technical colleges. Time I Nature and strength of relationships with related or spent screening candidates is minimized. Progressive supporting institutions that assist with operations and hiring policies in place (that is, gender sensitive). outreach (such as banks) I Job descriptions that are available, clearly written, and emphasize expected outputs rather than activities. No Indicators of health overlaps or underlaps in accountabilities between I Branching strategy is clearly articulated and makes departments and individuals. sense--urban versus rural--builds economies of I Training materials are thorough and well designed, scale--not too sparse covering philosophy and client service concepts as I Strong relationships with supporting institutions, that well as technical skills. A period of coaching/mentor- is, banks, collection agencies ing built into the training. Training period screens out inappropriate candidates. Length and cost of Transformation issues training minimized. I How will the transformation affect relationships with I Compensation policies are set to attract the right cal- other institutions? Will contracts need to be revised? iber of staff. I Incentive schemes are designed to reward portfolio 5. HUMAN RESOURCE MANAGEMENT quality as well as volume, and incentive amounts are set Documentation relative to the extra income earned through enhanced I Organigram of head office and field personnel structures. performance. I For primary operations staff (hiring criteria and pro- I Employees who express a high level of satisfaction cedures, job descriptions, training program; compen- with their work. Low employee turnover. No antago- sation policies, incentive scheme design, other nistic employee unions or associations. recognition and reward mechanisms). I Dynamic use of employee newsletter, with contribu- I For head office positions (any unusual or noteworthy tions from employees. Strong internal communications hiring or training policies, compensation policies, and transparency. A team spirit evident in all activities. incentive scheme design, other recognition and I Appraisal systems, promotions policies, and other reward mechanisms). policies are perceived as fair. 122 MICROFINANCE HANDBOOK Transformation isues Transformation issues I Will all head office and field operations go to the new I What systems changes will be required to accommo- institution or will some remain? date new reporting requirements for the new institu- I Will split be clean or will there be complex relation- tion? ships between the NGO and the new institution? I Will the current system be able to handle rapid growth I Will head office functions be shared between institu- and increased complexity over the next 18 months? tions? I How do staff members feel about the proposed 7. FINANCIAL MANAGEMENT transformation? Are they aware of it, do they under- Documentation stand the reasons for it, are they supportive of it? I Financial statements (when produced? when audited? What are their biggest concerns regarding transfor- how consolidated?) mation? I Relationship with auditors (availability of profit and I If the organization is being divided into two, will the cost center information and statements, availability of same human resources policies apply to both? Will multiple year financial forecasting models) there be a differential in compensation levels? How I Budgeting process will this be handled? I Cash flow management--cash flow forecasting, liquid- ity monitoring, minimizing idle cash, bank account 6. COMPUTERIZATION management Documentation I Delinquency management--repayment rates, portfolio I Current state of automation aging, arrears aging, provision and write-off policies head office and field offices I I Internal controls to avoid fraud, negligence, and theft, loan portfolio system, accounting system, other I internal audit procedures systems? I Sources of capital and relationship with I Hardware and software used, development tools used, bankers/funders, average cost of capital network design I Size of equity base in relation to total assets I Existence of functional specifications I Self-sufficiency management--operational efficiency, I Level of staff satisfaction with current systems productivity, and self-sufficiency ratios. I Profile and depth of information systems department I Procedures used for auditing systems. Indicators of health I Financial statements are produced and audited at least Indicators of health annually I Degree of automation in line with size and complexi- I Statements should be produced for the financial ser- ty of organization, degree of automation in line with vice activities of the organization unconsolidated from dependability of communications infrastructure in the any other nonfinancial activities country I Auditing firm should be respected and ideally have I Diversification of risk--avoid dependence on a single some financial sector experience/expertise programmer to develop and maintain the systems, I Income statements and key ratios are available for and avoid dependence on one expensive firm to make each primary operating unit--branch/region/total small updates to the systems I A three- to five-year financial forecasting model is in I Hardware and software used should be well supported place and actively used to assist with pricing and other locally, development tools should be well known policy setting and steer the organization toward locally profitability. I Apparent high level of staff satisfaction with the sys- I An annual budgeting process is in place and is used to tems control expenses. I Strong information systems manager and depart- I Cash flows are well managed--liquidity problems are ment--knowledgeable, reliable, service-oriented avoided, idle cash is minimized, and bank charges are I Effective procedures in place for auditing systems. minimized. THE INSTITUTION 123 I Delinquency and default are closely monitored and available after transformation? How will the higher tightly managed; appropriate provision and write-off costs affect self-sufficiency levels? (consider operating policies are in place and properly implemented. cost subsidies and financial cost subsidies.) I Sufficient sources of loan capital are in place to fund I How should the capital structure for the new organi- expected growth. The average cost of capital is kept to a zation be designed? How much control needs to be minimum. The relationship with banks is approached maintained? What kind of shareholders are desirable? as a business relationship, not as a public relations or How much capital is required to support the new donations relationship. Over time, the organization is institution for the first few years of its existence? What moving away from subsidized sources of funding and is an optimum debt to equity ratio for the new insti- developing a good reputation with commercial funders. tution? I The equity base is sufficient in relation to overall I What will be the dividend policy for the first three to assets and performing assets. five years of the transformed institution? I Key ratios of operational efficiency, productivity, and self-sufficiency are regularly monitored and guide the decisionmaking of board and senior man- Appendix 2. Manual for Elaboration of a agement. Key ratios are steadily improving, with Business Plan full self-sufficiency likely to be reached within 12 to 24 months. The following outline is from SEEP Network 1996b and I Management information reports are timely, accurate, was developed by CENTRO ACCION Microempresarial, and relevant. Bogota, Colombia, and ACCION International, Cam- I Key reports include: bridge, Mass. I Actual payments against expected payments--for field staff Definition I Client status report--for field staff I Bad debts report--for field staff and head office This manual provides guidelines for creating a business I Aging of arrears--for field staff and head office plan. It is divided into nine chapters, each of which I Aging of portfolio--for head office covers a distinct part of a business plan, from market I Performance indicators report--for head office analysis to portfolio projections to estimating opera- and board of directors. tional costs. The core chapters provide instructions for I Break-even analyses are available and guide decision- data gathering and analysis that feed directly into the making. construction of standard financial statements. Because I Pricing is carefully managed. Calculations are available the manual was written for ACCION affiliates, it for effective costs to the client and for effective yields to focuses on planning for the expansion of credit ser- the organization. Rates are established in order to cover vices. operating costs, loan losses, and financial costs and to achieve a profit at certain levels of operation. The Tool Transformation issues The original tool is a 45-page manual, in spanish, with I How will financial statements need to be revised for the following chapters: Institutional Analysis, Defining the new organization(s)? Market Size, Identifying the Competition, Projecting I Will new auditors need to be appointed? the Portfolio, Market Penetration, Financing Expansion, I How will bank accounts and liquidity management Operational Costs, Priority Projects, Projective Finan- be affected? cial Statements. I How will internal audit procedures need to be The tool included here is an English summary of the changed? manual, which highlights the main themes in each chap- I Are current subsidies in place today, due to the orga- ter and provides some examples of the many illustrative nization's non-profit status, which will no longer be tables and charts found in the original. 124 MICROFINANCE HANDBOOK How the Tool Works an integration of all elements of the plan, as well as func- tional integration of the managers who formulate and over- The business plan is a document that explicitly establish- see it. Using a financially based planning tool helps es an institution's trajectory over a three-year period. It is managers adapt to changing conditions and know more pre- the tool by which an institution's mission gets translated cisely how their decisions will affect financial performance. into measurable targets. These targets are set in function The plan serves as a reference tool to clarify priorities, unify of market demand and the competition. They inform efforts, avoid contradictions and evaluate the performances income and expense projections that take sustainability of both the institution and individual employees. from theory to reality. Moreover, for affiliates in the ACCION network, the In light of its mission, an institution sets targets for business plan is an essential tool in building relationships the market share it will achieve by the end of the three- with the formal financial sector. Without a plan, finan- year planning period. Market share, defined by geo- cial markets will not be disposed to increase lines of cred- graphic area, translates into a certain number of branch it or negotiate more favorable terms. offices and field staff, which in turn determines the insti- tution's cost structure. From these targets, the portfolio 1. INTRODUCTION size can be deduced, which then points to the amount of What is business planning? funding required. Internal coherence is then determined It is a process by which an organization defines a path by combining the product pricing, the cost structure, from its current situation to where it wants to be at a and the financing plan to form the income statement. determined point in the future. The Business Plan process begins with an ACCION team (usually two people, one of whom is often the presi- How is it used? dent) making a visit that includes a formal presentation of In the course of business planning, the organization the state of the microfinance program and of its medium- determines its destination (within a three-year horizon) term goals. The business plan tool is presented and staff and designs the road to get there. It establishes numeri- learn how to apply it. This takes place over a two-day cally defined objectives, strategies, and actions. This period in the presence of all members of the institution's diagnosis helps to answer two important questions: top management and, in some cases, with the attendance I Can the organization successfully travel along this path? of board members. While concepts related to the mission I What additional resources will it need and how will it are discussed, most of the time is spent on issues, includ- secure them? ing the type and size of the market, the relative strength of the competition, setting realistic growth targets based on a Who participates in the business planning process? number of key assumptions, and definition of a handful The entire management team must participate to ensure of priority projects that will have to be carried out if the that all of the pieces of the planning puzzle fit, resulting proposed objectives are to be achieved. At the end of the in a consistent, coherent plan. two-day period, a time line for completing different drafts of the business plan is agreed upon. What are the characteristics of a successful business plan? Most of the work is then carried out by the affiliates I Involves entire management team themselves, over a period of two to four months. The I Identifies key elements that differentiate the program work is divided among departments, with each manager from other similar organizations and explain its success responsible for involving his/her staff in the process to I Analyzes the environment, including the market and the fullest extent. As drafts of the plan are produced, they competition are sent to ACCION for a critical review of the assump- I Defines the organization's objectives tions and internal consistency. I Identifies strategies to overcome obstacles to achiev- The strength of the business plan lies in the way in ing objectives which all elements, and every decision, have a financial I Analyzes the financial implications of these strategies implication with an impact on financial performance and I Translates the objectives, strategies and resources into the bottom line. This requires analytical rigor. It also forces numerical terms that are void of inconsistencies. THE INSTITUTION 125 2. BUSINESS PLAN FORMAT ton around which the rest of the plan is elaborated. Box A4.2.1 presents a suggested format for a business Specific elements of the institutional analysis follow: plan and shows the links to financial status. History of the institution. 3. INSTITUTIONAL ANALYSIS Mission: The raison d'être of the enterprise, the mission This chapter directs the reader to the elements of an insti- statement defines the institution's values and priorities. tutional analysis that cover its history, mission, and com- The mission clearly defines the product and services to parative advantages. The institutional analysis is one of be offered, its target customer base and the location of the most important parts of the business as it is the skele- this clientele. Box A4.2.1 Suggested Business Plan Format Executive Summary Chapter 1: Institutional Analysis (Institutional Mission; Market Share; Diagnostic-- Weaknesses and Strengths) RELATIVE TO FINANCIAL STATUS Chapter 2: Basic Premises Relative to Macroeconomic Indices, National Financial System Chapter 3: Portfolio Revenue Volume: Size and Composition, Organic Growth, Increase in New Clients I Segmented by Currency: National/Dollar I Segmented by Industrial Market minus I Segmented by Region Price: Active Interest Rate Policy Strategies: Volume and Price Chapter 4: Financing Expansion Financial costs Volume: Size and Composition I By Source/Type of Instrument minus Price: Cost of Funds Strategy: Volume and Price Chapter 5: Operational Costs Operational costs Projected versus Past Analysis I By Component/Specific Department equal I Increase in Personnel, by positions Chapter 6: Priority Projects Earnings (Consistent with the Business Plan) Description of Each Priority Project minus I Timeline and Estimated Costs Chapter 7: Investment Plan Investments (for example, systems, personnel) Projective Financial Statements equal Three-year history, Estimate of Current Year, Three-year projection Statistical Annex Cash flow 126 MICROFINANCE HANDBOOK In defining the mission, planners must move the mis- and projections in three main areas: macroeconomics, sion from a concept to the articulation of its quantitative the informal sector, and the financial system. implications, which express the institution's objectives. Consequently the business plan identifies all elements 5. PORTFOLIO ANALYSIS AND PROJECTIONS necessary and sufficient to achieve those objectives and The projected portfolio is the "backbone" of the business outlines the specific actions to assure that these elements plan, determining costs and income. This chapter of the are in place when they are needed. manual contains six sections that offer guidelines for esti- mating the market, defining the competition, establishing Comparative advantage: Those characteristics that differen- portfolio projections, determining market penetration, tiate the institution from the others and explain its success. classifying the portfolio, and setting interest rate policies. Strengths and weaknesses (internal). Determining the market. To express an institution's mis- sion in quantitative terms, it is necessary to define the Opportunities and threats (external). customer base and estimate the size of the market this customer base represents. This provides a yardstick 4. BASIC PREMISES against which to measure both the reality of growth pro- This brief section of the manual describes the general jections and progress toward achieving the stated mis- economic analysis that should be part of the business sion. plan. This economic analysis should incorporate analysis The manual concedes that there is no one accepted, established procedure for determining market size. Given the quality and quantity of available data on the informal Box A4.2.2 Estimating the Market: An Example from sector, efforts to measure it can only result in estimations Ecuador drawn from available statistics, census data, sociological Steps research and other sources. The manual does offer the 1. Find the economically active population (EAP) from following guidelines and an example of one method of the most recent census data estimation in box A4.2.2. 2. Define the informal sector (for each city) I Focus on the geographic area of operation and its 3. Estimate the percentage of the EAP that is informal 4. Estimate the percentage that is subject to credit potential for growth. The steps outlined in box A4.2.2 5. Project growth. should be followed for each city or region of operation. I Within the informal sector, define the segment of the The case of Ecuador market that is eligible for products and services that 1. EAP (1989 census) 38% of total population the institution provides. 2. Formal sector 28% of EAP registered I Growth projections should be based on both his- with social security torical patterns and current factors such as migra- Unemployment 13% tion, government policies relevant to the sector, Independent professionals 2% and so forth. 43% of EAP is in the Although only approximated, this definition of the formal sector market is extremely important because it is a vision of Informal sector = Total population x 38% x (1­43%) the business environment that must be shared by man- 3. We estimate that 15% of the informal sector would agement in order to serve as a basis for decisionmaking. not be interested in credit. Of the targeted 85%, only 38% own their business and, therefore, would be can- Defining the competition. To position the institution in didates for loans. the marketplace, the following information about the competition should be collected. This information will Customer base = Informal sector x 0.85 x 0.38 help to identify the institution's strengths and weakness- Source: SEEP Network 1996b. es and influence strategies for the three-year planning horizon. THE INSTITUTION 127 I Who are the competitors? I Indicators of portfolio quality: delinquency and default. I Size Interest rate policies. The final part of portfolio projec- I Funding sources tions is setting the interest rate(s), which will be based on I Client base the following factors: I Market niche I Market rates (researched as part of competition I Marketing methods analysis) I Portfolio size and quality I Operating costs (explained in chapter 8). I Methodology I Financial costs (a weighted average) I How solid are they? I Capitalization objectives (that is, targeted level of I Pricing retained earnings generated by the portfolio deter- I Time to deliver service mined during the institutional analysis). I Marketing methods Although there are various methods for setting inter- I Numbers. est rates, one simple example is illustrated below: It can be helpful to construct a "Chart of Comp- Interest rate charged 20% etition," placing competitors on a graph defined by ­ Financing costs 10% interest rates and non-monetary costs. In addition to = Financial margin 10% locating the institution in relation to its competitors, the ­ Operating costs 5% graph can be used to show its movement in time. = Operating margin 5% Desired margin 8% Projecting the portfolio. Projections of portfolio growth With estimated financing and operating costs, the must be made on two levels: loan volume of existing interest rate can be set to achieve the desired margin. In clients in the current portfolio and loan volume attrib- this example, the institution has not achieved its desired uted to new clients. While the institution has the option margin leaving it with the option to raise its interest rate to take new clients or not, it must attend to the needs of or reduce costs. This is a dynamic process in which the existing clients. (Inability to do so would trigger wide- projections and parameters will need to be adjusted to spread delinquency and even default.) Additional factors secure the desired results. to consider in projecting the portfolio include: I Client drop-out rate 6. FINANCING EXPANSION I Growth pattern of average loan size over time-- This chapter identifies various funding sources--their influenced by average loan term, cyclical nature of characteristics, costs, and conditions. Each is analyzed for demand, influence of inflation on initial loan size, accessibility and inherent risks. Here are two sources of and new loan products with different characteris- funding: internal (retained earnings) and external (dona- tics. tions, shareholder capital, soft loans, commercial loans, Estimating market penetration. The relationship of the and deposits). With the portfolio growth and the desired portfolio projections to the estimated market for credit rate of capitalization established, external funding needs yields the market penetration (MP ). Calculating the can be determined. Assuming multiple sources at varying institution's market penetration serves as a reality check cost, it is necessary to establish a weighted average of the for the projections: sources to estimate the institution's financial costs. Table A4.2.1 provides a hypothetical example. The weighting Customer base (defined in projected portfolio) MP = factor for each funding category is the product of the pro- Potential market portion of the total external capital and the effective inter- Classifying the portfolio. The portfolio must be disaggre- est rate. The sum of the weights is the weighted average gated or classified by a number of factors that have an interest rate the organization pays on its external capital. impact on cash flow and costs. These include: I Loan size and number 7. OPERATIONAL COSTS--PAST AND PROJECTED I Loan maturities Offices and staff by city or geographic area. Disaggregate I Life of the client with the program costs by office. Personnel, the most significant cost, 128 MICROFINANCE HANDBOOK Table A4.2.1 Financial Cost as a Weighted Average Proportion Interest rate Weighted Funding Source Amount (percent) (percent) interest rate Donations 250,000 5.0 0 0 IDB loan 500,000 10.0 1 0.001 Government loan 1,000,000 20.0 20 0.04 Bank loan 1,750,000 35.0 35 0.1225 Deposits 1,500,000 30.0 25 0.075 Total 5,000,000 100.0 0.239 Source: Fruman and Isern 1996. should be detailed by type and number. Analyze histori- to be addressed, and the benefits to be gained--both cal performance, going back three years, including cur- qualitative and quantitative. rent year projections; Estimate future costs according to projections and selected productivity ratios. Every new Costs. In addition to the impact on operating costs, the plan office and additional staff must be consistent with must account for special costs associated with these projects, accepted levels of productivity and fully justified in the such as the staff training required to operate a new system. business plan. Timeframe for investments. A detailed plan of tasks and Productivity ratios of the time line for their accomplishment outlines the process for purposes of control and to identify points Clients/credit Officer/credit officer/ Portfolio when funds will be needed to finance the project. employees credit officer Clients/employees Credit officer/agency Portfolio/ Sources of financing for the project. Precise knowledge of employee the possible sources of funds for the project is essential to Clients/agency Employees/agency Portfolio/ guarantee results. This involves identifying the sources, agency their characteristics, conditions, terms, and process by which to secure funds. Personnel costs. Use historical data to estimate salaries, benefits, and annual increases. 9. FINANCIAL STATEMENTS With these projections completed, all the elements are in Specific planning for new facilities or renovations. Once the place to construct projected financial statements includ- number of new offices or the expansion of existing ones has ing the Balance Sheet, the Profit and Loss Statement, and been established, the process must be fully identified in the the Cash Flow Statement. business plan, including associated projects, investments, and expenses such as legal costs, publicity, and additional maintenance. Sources and Further Reading 8. PRIORITY PROJECTS Benjamin, McDonald, and Joanna Ledgerwood. 1998. "The Some projects merit separate treatment in the business Association for the Development of Microenterprises planning process. Examples include the introduction of (ADEMI): `Democratising Credit' in the Dominican savings services, the creation of a formal financial interme- Republic." Case Study for the Sustainable Banking with diary or modernization of the institution's systems. A chap- the Poor Project, World Bank, Washington, D.C. ter in the business plan on priority projects should outline: Bremner, Wayne. 1996 "CHISPA." In "Institutional Structures for Micro Enterprise Finance: Implications for Sus- Project objectives. Description, background, the problem THE INSTITUTION 129 tainability and Expansion." In Moving Forward: Emerging Churchill, ed., Establishing a Microfinance Industry. Strategies for Sustainability and Expansion. The SEEP Washington, D.C.: MicroFinance Network. Network, New York: PACT Publications. Development Project Service Centre (DEPROSC) and Joanna Calmeadow. 1996. "Planning for Institutional Trans- Ledgerwood. 1997. "Critical Issues in Nepal's Micro- formation--Analysis of a Microfinance Organization." In Finance Circumstances." University of Maryland at Moving Forward: Emerging Strategies for Sustainbility and College Park, Institutional Reform and the Informal Expansion. The SEEP Network, New York: PACT Sector, College Park, Md. Publications. Edgcomb, Elaine, and Jim Cawley, eds. 1993. An Institutional CGAP (Consultative Group to Assist the Poorest). 1998. Guide for Enterprise Development Organizations. New York: "Apex Institutions." Occasional Paper, World Bank, PACT Publications. Washington, D.C. Fruman, Cecile. 1997. "FEECAM--Benin." Case Study for ------. 1997a. "Anatomy of a Micro-finance Deal: A New the Sustainable Banking with the Poor Project, World Approach to Investing in Micro-Finance Institutions." Bank, Washington, D.C. Focus Note 9. World Bank, Washington, D.C. Fruman, Cecile, and Jennifer Isern. 1996. "World Bank ------. 1997b. "The Challenge of Growth for Micro-Finance Microfinance Training." World Bank, Washington, D.C. Institutions: The BancoSol Experience." Focus Note 6. Galludec, Gilles. 1996. "CGAP Appraisal Mission." World World Bank, Washington, D.C. Bank, Consultative Group to Assist the Poorest, Wash- ------. 1997c. "Effective Governance for Micro-Finance ington, D.C. Institutions." Focus Note 7. World Bank, Washington, D.C. Garber, Garter. 1997. "Private Investment as a Financing ------. 1997d. "State-Owned Development Banks in Micro- Source for Microcredit." The North South Agenda Papers Finance." Focus Note 10. World Bank, Washington, 23. University of Miami, North-South Center, Miami, Fla. D.C. Gonzalez-Vega, Claudio. 1998. "Microfinance Apex Christen, Robert Peck. 1997. Banking Services for the Poor: Mechanisms: Review of the Evidence and Policy Managing for Financial Success. Washington, D.C.: Recommendations." Ohio State University, Department of ACCION International. Agricultural, Environmental, and Development Economics, Christen, Robert, Elizabeth Rhyne, and Robert Vogel. 1995. Rural Finance Program, Columbus, Ohio. "Maximizing the Outreach of Microenterprise Finance: An Gudger, Michael. 1997. "Sustainability of Credit Guarantee Analysis of Successful Microfinance Programs." Program Systems." In SME Credit Guarantee Schemes 4 (1 and 2): and Operations Assessment Report 10. U.S. Agency for 30­33. Published for the Inter-American Development Bank International Development, Washington D.C. by The Financier. Chu, Michael. 1996a. "Securitization of a Microenterprise Holtmann, Martin, and Rochus Mommartz. 1996. A Portfolio." In Craig Churchill, ed., An Introduction to Key Technical Guide for Analyzing Credit-Granting NGOs. Issues in Microfinance: Supervision and Regulation, Saarbricken, Germany: Verlag fir Entwicklungspolitik. Financing Sources, Expansion of Microfinance Institutions. Jackelen, Henry R., and Elisabeth Rhyne. 1991."Toward a Washington, D.C.: MicroFinance Network. More Market-Oriented Approach to Credit and Savings for ------. 1996b. "Reflections on Accessing Capital Markets." the Poor." Small Enterprise Development 2 (4): 4­20. ACCION International summary of paper delivered at Krahnen, Jan Pieter, and Reinhard H. Schmidt. 1994. Fourth Annual MicroFinance Network Conference, Development Finance as Institution Building: A New November, Toronto. Approach to Poverty-oriented Banking. Boulder, Colo.: Churchill, Craig, ed. 1997. Establishing a Microfinance Westview Press. Industry: Governance, Best Practices, Access to Capital Levitsky, Jacob. 1997. "Best Practice in Credit Guarantee Markets. Washington, D.C.: MicroFinance Network Schemes." In SME Credit Guarantee Schemes. 4 (1 and 2): ------. ed. 1998. Moving Microfinance Forward: Ownership, 86­94. Published for the Inter-American Development Competition, and Control of Microfinance Institutions. Bank by The Financier. Washington, D.C.: MicroFinance Network. Long, Carolyn, Daouda Diop, and James Cawley. 1990. Clark, Heather. 1997. "When a Program is Ready to Access "Report of Program Working Group." In Elaine Commercial Funds: A Discussion for Commercial Edgecomb and James Cawley, eds., An Institutional Guide Standards." Paper prepared for the Microcredit Summit, for Enterprise Development Organizations. New York: February 3, Washington D.C. PACT Publications. Clarkson, Max, and Michael Deck. 1997. "Effective Magill, John H. 1994. "Credit Unions: A Formal-Sector Governance for Micro-Finance Institutions." In Craig Alternative for Financing Microenterprise Development." In 130 MICROFINANCE HANDBOOK Maria Otero and Elisabeth Rhyne, eds., The New World of Rosenberg, Rich. 1996. "Comment on DevFinance Network Microenterprise Finance: Building Healthy Financial May 15, 1996." Internet discussion group: devfinance Institutions for the Poor. West Hartford, Conn.: Kumarian @lists.acs.ohio-state.edu Press. Schmidt, Reinhard H., and Claus-Peter Zeitinger. 1994. Montoya, Manuel. 1997. "EDPYME in Peru." In Craig "Critical Issues in Small and MicroBusiness Finance." Churchill, ed., Establishing a Microfinance Industry. Internationale Projekt Consult GmbH (IPC), Frankfurt. Washington, D.C.: Microfinance Network. SEEP (Small Enterprise Education and Promotion) Network. Nataradol, Pittayapol. 1995. "Lending to Small-Scale Farmers: 1996a. Moving Forward: Emerging Strategies for Sustainability BAAC's Experience." Bank of Agriculture and Agricultural and Expansion. New York: PACT Publications. Cooperatives, Finance against Poverty Conference. ------. 1996b. Moving Forward: Tools for Sustainability and Reading, U.K. Expansion. New York: PACT Publications. Otero, Maria. 1998. "Types of Owners for Microfinance Stearns, Katherine. 1993. "Leverage or Loss? Guarantee Funds Institutions." In Craig Churchill, ed., Moving Microfinance, and Microenterprises." Monograph Series 8. ACCION Forward Ownership, Competition, and Control of International, Washington, D.C. Microfinance Institutions. Washington, D.C.: Microfinance Vogel, Robert C., and Dale Adams. "Costs and Benefits of Network. Loan Guarantee Programs." In SME Credit Guarantee Otero, Maria, and Elisabeth Rhyne, eds. 1994. The New World of Schemes 4 (1 and 2): 22­29. Published for the Inter- Microenterprise Finance: Building Healthy Financial Institutions American Development Bank by The Financier. for the Poor. West Hartford, Conn.: Kumarian Press. Von Pischke, J.D. 1991. "Finance at the Frontier: Debt Paxton, Julia. 1996. "A Worldwide Inventory of Microfinance Capacity and the Role of Credit in the Private Economy." Institutions." World Bank, Sustainable Banking with the World Bank, Economic Development Institute Poor Project, Washington, D.C. Development Studies, Washington D.C. ProFund Internacional. 1997. S.A. Annual Report 1996­97. ------. 1996. "Comment on DevFinance Network, May 14, San Jose, Costa Rica. 1996." Internet discussion group: devfinance@lists. Rhyne, Elisabeth. 1996. "Human Resource Development: acs.ohio-state.edu Microfinance as a Breakthrough Service." In An Wesley, Glenn D., and Shell Shaver. 1996. "Credit Union Introduction to Key Issues in Microfinance. Washington, Policies and Performance in Latin America." Paper pre- D.C.: MicroFinance Network. sented at the 1996 Annual Meeting of the Latin Richardson, David C., Barry L. Lennon, and Brian L. Branch. American and Caribbean Economics Association, 1993. "Credit Unions Retooled: A Road Map for Financial October, Mexico City. Stabilization." World Council of Credit Unions, Madison, Yaron, Jacob. 1992. Successful Rural Finance Institutions. World Wisc. Bank Discussion Paper 150. Washington, D.C. Rock, Rachel. 1997. "Regulation and Supervision Case Yaron, Jacob, McDonald Benjamin, and Gerda Piprek. 1997. Studies--Bolivia." In Regulation and Supervision of "Rural Finance Issues, Design and Best Practices." World Microfinance Institutions: Case Studies. Microfinance Bank, Agriculture and Natural Resources Department, Network Occasional Paper 2. Washington, D.C. Washington, D.C. Part II­ Designing and Monitoring Financial Products and Services CHAPTER FIVE Designing Lending Products B y definition, MFIs provide credit. Regardless of Cash Patterns, Loan Terms, and Payment the approach selected (see chapter 4), the actual Frequency loan products need to be designed according to the demands of the target market. This involves estab- This section covers credit fundamentals, including lishing appropriate loan amounts, loan terms, collateral cash patterns of borrowers, loan amounts, loan terms, requirements (or substitutes), interest rates and fees, and repayment schedules. Underlying each topic is an and, potentially, compulsory savings or group contribu- emphasis on understanding the behavior and credit tion requirements. needs of clients. Successfully designed credit products that meet the needs of microentrepreneurs are a necessity for any Client Cash Patterns and Loan Amounts MFI. It is important that the people who provide and evaluate lending services understand the different ele- To design a loan product to meet borrower needs, it is ments of lending products and the way in which these important to understand the cash patterns of borrowers. elements affect both borrowers and the viability of the Cash inflows are the cash received by the business or MFI. This chapter illustrates how different design ele- household in the form of wages, sales revenues, loans, or ments can result in lending products that are specifically gifts; cash outflows are the cash paid by the business or tailored to both the target market and the capacity of household to cover payments or purchases. Cash pat- the MFI. terns are important insofar as they affect the debt capac- This chapter emphasizes designing financial products ity of borrowers. Lenders must ensure that borrowers to meet client needs, based on the belief that microentre- have sufficient cash inflow to cover loan payments when preneurs value access to financial services and act in a they are due. responsible manner if they are treated as clients rather Some cash inflows and outflows occur on a regular than as beneficiaries.1 The chapter will be of interest to basis, others at irregular intervals or on an emergency or practitioners who wish to modify or refine their loan seasonal basis. Seasonal activities can create times when products and to donors and consultants who are evaluat- the borrower generates revenues (such as after a harvest ing the credit products of MFIs, in particular their finan- period) and times when there is no revenue (revenue cial aspects. The appendixes to the chapter provide may be received from other activities). However, loan details on more technical topics such as setting a sustain- terms often extend over several seasons, during which able interest rate and calculating the effective rate on there can be gaps in revenues. loans using the internal rate of return and taking into Loans should be based on the cash patterns of bor- account varying cash flows. rowers and designed as much as possible to enable the 1. Portions of this chapter are based on materials originally published in Ledgerwood (1996). 133 134 MICROFINANCE HANDBOOK client to repay the loan without undue hardship. This ing, resulting in a four-month business cycle. Her net helps the MFI avoid potential losses and encourages revenue over the four months (after purchasing sup- clients both to manage their funds prudently and to plies for 1,000 and incurring all other expenses but build up an asset base. (This does not mean that only before loan repayment) is 1,600 (400 per month). cash flows from the specific activity being financed are ALTERNATIVE 1: Four-month loan matching her considered; all cash flows are relevant.) business cycle. She borrows 1,000 for four months at 3 The appropriate loan amount is dependent on the pur- percent monthly interest, with monthly payments of pose of the loan and the ability of the client to repay the 269 (calculated on the declining balance method, loan (that is, debt capacity). When determining the debt which is explained in detail in the section on loan pric- capacity of potential clients, it is necessary to consider their ing, below). Her total payments are then 1,076 (inter- cash flow as well as the degree of risk associated with this est expense of 76). Revenue of 1,600 less the loan cash flow and other claims that may come before repay- repayment of 1,076 leaves her a net income of 524. ment of a loan to the MFI. Adjusting the debt capacity of Cash flow over the four months is as follows: a borrower for risk should reflect reasonable expectations about adverse conditions that may affect the borrower's Period Business Loan Net income enterprise. Adjustment for adversity has to reflect the 0 (1,000) 1,000 -- lender's willingness to assume the risks of borrowers' 1 400 (269) 131 inability to repay. The greater the MFI's capacity to 2 400 (269) 131 assume risk, the higher the credit limits the lender can 3 400 (269) 131 offer (Von Pischke 1991). 4 400 (269) 131 Often MFIs have a maximum loan size for first-time Total 600 (76) 524 borrowers, which increases with each loan. This is designed both to reduce the risk to the MFI and to create an incen- In this scenario, the dressmaker has extra income tive for the clients to repay their loans (namely, the promise to consume or reinvest as additional working capital if of a future larger loan). In addition, increasing loan sizes she chooses. The loan term is appropriately matched enable the client to develop a credit history and an under- to her business cycle and cash flow patterns. standing of the responsibilities associated with borrowing. ALTERNATIVE 2: Two-month loan shorter than her business cycle. She borrows 1,000 for two months at 3 How Does the Loan Term Affect the Borrower's Ability to percent per month with monthly payments of 523 Repay? and sales of 400 per month. Total payments are 1,046 (interest expense of 46). Revenue of 1,600 less the The loan term is one of the most important variables in loan repayment of 1,046 leaves 554. microfinance. It refers to the period of time during which Cash flow is as follows: the entire loan must be repaid. The loan term affects the Period Business Loan Net income repayment schedule, the revenue to the MFI, the financing costs for the client, and the ultimate suitability of the use 0 (1,000) 1,000 -- of the loan. The closer an organization matches loan terms 1 400 (523) (123) 2 400 (523) (123) to its client's needs, the easier it is for the client to "carry" 3 400 0 400 the loan and the more likely that payments will be made 4 400 0 400 on time and in full. The following example provides three Total 600 (46) 554 alternative loan terms with installment payments:2 N For example, a dressmaker purchases cloth and sup- With a two-month loan term and a four-month plies every four months to benefit from bulk purchas- business cycle, the borrower does not generate enough 2. This loan term analysis was developed by Barbara Calvin, codirector, International Operations, Calmeadow, and was originally published in Ledgerwood (1996). DESIGNING LENDING PRODUCTS 135 revenue in the first two months to make the loan pay- The preceding three alternatives demonstrate that ments. If she had no savings to begin with or no cash flow in part determines the debt-servicing capacity access to other income or credit to support the loan of borrowers. This influences the appropriate loan terms payments, she would not have been able to repay the and loan amounts, which in turn determine the debt-ser- loan. vicing requirements. MFIs should design the loan terms ALTERNATIVE 3: Six-month loan longer than her and loan amounts to meet the debt-servicing capacity of business cycle. She borrows 1,000 for six months at 3 their clients. percent per month with monthly payments of 184.60 This becomes less of a concern the more profitable the and sales of 400 per month for four months only. business becomes because greater revenue can potentially (Assume that because she does not have a full 1,000 result in enough additional income to build up sufficient built up in working capital at the end of the four savings, so that the client no longer needs to borrow months, she cannot buy more inventory at the "bulk unless she or he wants to expand the business. (In such a purchase" price and thus has two months of no case the MFI has successfully improved the economic income While this may not always be realistic, it is position of its client and should not consider this client presented here to illustrate the fact that in some cir- "drop out" to be negative or indicative of poor service or cumstances longer loan terms are detrimental to the products.) borrowers, particularly if they cannot access future Depending on cash patterns and loan terms, clients loans until the existing loan is paid back.) Total pay- may at times prefer to prepay their loans. Prepayments ments are 1,107.60 (interest expense of 107.60). have two major advantages for the client. Revenue of 1,600 less the loan repayment of 1,107.60 I They can reduce both the security risk and the temp- leaves 492.40. tation to spend excess amounts of cash. Cash flow is as follows: I They can reduce the burden of the loan installments later in the loan cycle. Period Business Loan Net income Prepayments result in one clear advantage for an MFI: 0 (1,000) 1,000 -- by having the loan repaid earlier, the MFI can revolve the 1 400 (184.6) 215.4 loan portfolio more quickly and thereby reach more 2 400 (184.6) 215.4 clients. 3 400 (184.6) 215.4 4 400 (184.6) 215.4 However, prepayments are difficult to monitor, and if 5 0 (184.6) (184.6) they are significant they can disrupt the cash flow of an 6 0 (184.6) (184.6) MFI (or its branches). This may affect the ability to Total 600 (107.6) 492.4 accurately predict cash flow requirements. In some MFIs full loan repayment results automatically in a larger loan In this scenario the cash flow in the first four to the client. This may result in the MFI having reduced months is easier for the borrower; however, she may funds available to lend to other clients. be tempted to spend the higher net income in the In addition, when interest is calculated on the declin- early months of the loan, resulting in potential diffi- ing balance, a prepayment usually includes the principal culty making loan payments during the last two owing and any interest due up to the amount of the pre- months. She also has less net income due to the payment. This means that when loans are prepaid, less greater amount of interest paid. interest revenue is received than initially forecast, result- This example illustrates why 12-month loans in ing in decreased income for the MFI (unless the funds busy urban markets often result in delinquency and can be immediately lent out again). defaults toward the end of the loan term. Also, the If a particular subset of borrowers (for example, mar- borrower ends up being underemployed for two ket vendors) tend to prepay their loans on a regular basis, months, because she does not have access to credit to it is advisable to shorten the loan term to suit the needs buy more inventory. If the loan term were shorter and of that client group. Loan terms should be designed to min- she was thus able to borrow again, she could continue imize the need for prepayments. This involves matching to be fully engaged in her business. the loan term to the cash patterns both to help clients 136 MICROFINANCE HANDBOOK budget their cash flows and to lower the likelihood of MFIs may also combine installment loans with lump prepayments or delinquency. sum payments, collecting a minimal amount of the loan Finally, prepayments can also indicate that borrowers (for example, interest) over the loan term, with the are receiving loans from another lender, which may be remainder paid at the end of the harvest season (the providing better service, lower interest rates, or more principal). appropriate terms. If this is the case, the MFI needs to examine its loan products and those of other lenders. Working Capital and Fixed Asset Loans Frequency of Loan Payments The amount of the loan and the appropriate loan term are affected by what the loan will be used for. There are Loan payments can be made on an installment basis generally two types of loans--working capital loans and (weekly, biweekly, monthly) or in a lump sum at the end fixed asset loans. of the loan term, depending on the cash patterns of the Working capital loans are for current expenditures that borrower. For the most part, interest and principal are occur in the normal course of business. Working capital paid together. However, some MFIs charge interest up refers to the investment in current or short-term assets to front (paid at the beginning of the loan term) and princi- be used within one year. Examples are wood purchased pal over the term of the loan, while others collect interest for carpentry, food or goods purchased for market sell- periodically and the principal at the end of the loan term. ing, or chick feed purchased for poultry rearing. A loan Activities that generate ongoing revenue can be made for working capital should have a loan term that designed with installment payments. In this way the matches the business cycle of the borrower (as described client is able to repay the loan over time without having above). Working capital loans from MFIs are generally to save the loan amount (for repayment) over the term of for two months to one year. the loan. The frequency of the loan payments depends Fixed asset loans are those made for the purchase of on the needs of the client and the ability of the MFI to assets that are used over time in the business. These ensure repayment. Some moneylenders collect payments assets typically have a life span of more than one year. every day, particularly if their borrowers are market ven- Fixed assets are usually defined as machinery, equip- dors receiving cash on a daily basis. Other lenders collect ment, and property. Examples of fixed assets include on a monthly basis, because they are not conveniently motorcycles, sewing machines, egg incubators, or rick- accessible to the borrower (that is, the bank branch is shaws. Since the productive activity does not directly use located far from the borrower's business). A balance must up the fixed asset (that is, not sold as part of the prod- be reached between the transaction costs associated with uct), its impact upon profitability is felt over a longer frequent payments and the risk of default through poor period of time. cash management associated with infrequent repayments. A loan made for a fixed asset is generally for a larger For seasonal activities, it may be appropriate to design amount and for a longer term than a working capital the loan such that a lump sum payment is made once the loan (that is, fixed asset loans do not necessarily match activity is completed. Harvesting activities is a good exam- the business cycle). This results in higher risk for an ple. (Note that other household income can be used to MFI, offset somewhat if the organization takes legal repay the loan in small amounts.) However, caution must title of the purchased asset as collateral. Table 5.1 pro- be exercised with lump sum payments, particularly if there vides examples of loan uses for working capital or fixed is risk that the harvest (or other seasonal activity) may fail. assets. If when the loan is due there is no revenue being generat- Although longer loan terms may be required for fixed ed, the risk of default is high. Some MFIs that finance sea- asset purchases, some MFIs find that they do not need sonal activities design their loans with installment to introduce special loan products if they are relatively payments so that at harvest times the borrowers keep most inexpensive and the loan can be repaid over 12 months of their harvest revenue, because the majority of the loan or less. has already been repaid by the end of the harvest. This More and more often MFIs are acknowledging the works to increase the savings (assets) of borrowers. fungibility of money, that is, the ability at various times DESIGNING LENDING PRODUCTS 137 loans. Guarantees are either implicit guarantees, with Table 5.1 Examples of Loan Uses other group members unable to access a loan if all mem- Fixed Working bers are not current in their loan payments, or actual Activity assets capital guarantees, with group members liable if other group Vendor Food stall Merchandise members default on their loans. Refrigerator Plastic bags Some MFIs require group members to contribute to Scale Paper bags a group guarantee fund, which is used if one or more Dressmaker Sewing machine Fabric borrowers fail to repay. Use of the group guarantee fund Sewing table Patterns is sometimes at the discretion of the group itself and Mannequin Needles sometimes decided by the MFI. If it is used at the and thread group's discretion, the group will often lend money Carpenter Saw Lumber from the guarantee fund to the group member who is Lathe Nails unable to pay. The member who "borrows" from the Sander Sandpaper group fund is then responsible for paying the fund back. Poultry Incubator Eggs If use of the group guarantee fund is managed by the hatchery Baskets Electricity owner Wire and holder Wrapping MFI, the fund is seized to the extent of the defaulted Thermometer cloth loan, with other group members making up any short- Shelves Egg fall. Failure to do so means that the entire group no sterilization longer has access to credit. Source: Ledgerwood 1996. CHARACTER-BASED LENDING. Some MFIs lend to people based on a good reputation in the community. Prior to to use funds borrowed for a certain activity for other making a loan the credit officer visits various establish- household expenses. Accordingly, the purpose of the ments in the community and asks about the potential loan is not as important as the borrower's capacity to client's character and behavior. repay it. FREQUENT VISITS TO THE BUSINESS BY THE CREDIT OFFICER. Provided the branch or credit officers are within a reason- Loan Collateral able geographical distance from their clients, frequent visits help to ensure that the client is maintaining the business Generally, MFIs lend to low-income clients who often and intends to repay the loan. Frequent visits also allow have very few assets. Consequently, traditional collater- the credit officer to understand her or his clients' business- al such as property, land, machinery, and other capital es and the appropriateness of the loan (amount, term, fre- assets is often not available. Various innovative means quency of payments, and so forth). Visits also contribute of reducing the risk of loan loss have been developed, to developing mutual respect between the client and the including collateral substitutes and alternative col- credit officer as they learn to appreciate and understand lateral. each other's commitment to their work. Collateral Substitutes R ISK OF PUBLIC EMBARRASSMENT . Often clients will repay loans if they feel that they will be embarrassed in One of the most common collateral substitutes is peer front of their family, peers, and neighbors. This would be pressure, either on its own or jointly with group guaran- the case if public signs, notices printed in the local news- tees (see chapter 3). In addition, there are other frequent- paper, or announcements made at community meetings ly used forms of collateral substitute. list borrowers who do not repay. GROUP GUARANTEES. Many MFIs facilitate the formation RISK OF JAIL OR LEGAL ACTION. Depending on the legal of groups whose members jointly guarantee each other's context in a country, some MFIs have sued or, in rare 138 MICROFINANCE HANDBOOK cases, even jailed clients for nonpayment. Sometimes, stable source of funds because they are illiquid (that is, simply the risk of legal repercussion is enough to encour- savings are not generally available for withdrawal by the age repayment. borrowers while their loans are outstanding). However, it is essential that the MFI act in a prudent manner when Alternative Forms of Collateral on-lending or investing client savings to ensure that the funds will be repaid and can be returned to the client in There are at least three commonly used alternative forms full when necessary. of collateral. ASSETS PLEDGED AT LESS THAN THE VALUE OF THE LOAN. COMPULSORY SAVINGS. Many MFIs require clients to Sometimes, regardless of the actual market value of hold a balance (stated as a percentage of the loan) in sav- assets owned by the borrower, the act of pledging assets ings (or as contributions to group funds) for first or sub- (such as furniture or appliances) and the consequent sequent loans (or both). Compulsory savings differ from realization that they can be lost (resulting in inconve- voluntary savings in that they are not generally available nience) causes the client to repay the loan. It is impor- for withdrawal while a loan is outstanding. In this way tant that the MFI formally seize the assets that have compulsory savings act as a form of collateral. been pledged if the client does not repay the loan. This By being required to set aside funds as savings, borrow- sends a message to other borrowers that the MFI is seri- ers are restricted from utilizing those funds in their busi- ous about loan repayment. ness activities or other income producing investments. Usually the deposit interest rate paid (if any) on the sav- PERSONAL GUARANTEES. While microborrowers them- ings is lower than the return earned by the borrowers if selves do not often have the ability to guarantee their the savings were put into their business or other invest- loans, they are sometimes able to enlist friends or family ments. This results in an opportunity cost equal to the dif- members to provide personal guarantees (sometimes ference between what the client earns on compulsory referred to as cosigners). This means that in the event of savings and the return that could be earned otherwise. the inability of the borrower to repay, the person who Compulsory savings can have a positive impact on has provided a personal guarantee is responsible for clients by smoothing out their consumption patterns and repaying the loan. providing funds for emergencies provided the savings are Many of the foregoing collateral substitutes and alter- available for withdrawal by the borrower. Most compulso- native forms of collateral are used in combination with ry savings are available for withdrawal only at the end of each other. A good example of this is offered by the the loan term, providing the loan has been repaid in full. Association for the Development of Microenterprises Clients thus have additional cash flow for investment or (ADEMI) in the Dominican Republic (box 5.1). consumption at the end of the loan term. Compulsory savings also provide a means of building assets for clients; not all MFIs view compulsory savings as strictly an alter- Loan Pricing native form of collateral. A variation of compulsory savings required by Bank Pricing loans is an important aspect of loan product Rakyat Indonesia is for borrowers to pay additional design. A balance must be reached between what clients interest each month, which is returned to them at the can afford and what the lending organization needs to end of the loan provided they have made full, on-time earn to cover all of its costs. Generally, microfinance payments each month. This is referred to as a "prompt clients are not interest-rate sensitive. That is, microentre- payment incentive" and results in the borrower receiving preneurs have not appeared to borrow more or less in a lump sum at the end of the loan term. This benefits reaction to an increase or decrease in interest rates. For the borrower and provides a concrete incentive to repay the most part, an interest rate far above commercial bank the loan on time, thus benefiting the bank as well. rates is acceptable because the borrowers have such limit- Compulsory savings also provide a source of lending ed access to credit. However, an MFI must ensure that and investment funds for the MFI. They are generally a its operations are as efficient as possible so that undue DESIGNING LENDING PRODUCTS 139 burden is not put on its clients in the form of high inter- average financing costs for an MFI. Many MFIs do not est rates and fees (box 5.1). pay any interest on compulsory savings, and if this is the MFIs can determine the interest rate they need to case the cost of funds for the compulsory savings is zero. charge on loans based on their cost structure. MFIs incur If an MFI funds its loan portfolio with borrowed money four different types of costs: from a bank, for example, at 12 percent and with client I Financing costs savings at 6 percent, the higher the portion of the port- I Operating costs folio funded with client savings, the lower the overall I Loan loss provision cost of funding. (Note that this does not reflect the I Cost of capital. increase in operational costs, for example, for monitor- Each of these costs are discussed further in part III ing that is incurred to collect compulsory savings. This when adjustments to the financial statements are made discussion only refers to the average cost of funds for an and when determining the financial viability of MFIs. For organization and does not include the cost of collecting specific information on how to set a sustainable interest compulsory savings.) rate based on the cost structure of an MFI, see appendix 1. N For example, with a 1,000,000 loan portfolio funded In general, an MFI incurs relatively low financing by 300,000 savings (at 6 percent) and 700,000 bor- costs if it funds its loan portfolio primarily with donated rowed funds (at 12 percent), the average annual cost funds. However, if compulsory savings (discussed above) of funds is: are used to fund the loan portfolio, they can affect the (300,000 x 6%) + (700,000 x 12%) 10.2% = 1,000,000 Box 5.1 The Association for the Development If the compulsory savings component is increased of Microenterprises' Collateral Requirements to 500,000, the average cost of funds is reduced. T HE A SSOCIATION FOR THE D EVELOPMENT OF Using the above example, with a 1,000,000 loan port- Microenterprises (ADEMI) uses a combination of collateral folio funded by 500,000 savings (at 6 percent) and and guarantors to secure its loans. All borrowers are 500,000 borrowed funds (at 12 percent), the average required to sign a legal contract stating their obligation to annual cost of funds is: repay funds at specified terms. Most clients also have a guarantor or cosigner who assumes full responsibility for (500,000 x 6%) + (500,000 x 12%) 9% = the terms of the contract if clients are either unable or 1,000,000 unwilling to repay loans. When feasible, borrowers are The opposite of this is also true. If an MFI funds required to sign deeds for property or machinery as loan its portfolio primarily with grant money for which guarantees. (The association estimates that 80 percent of there is no interest cost, the addition of 6 percent paid the deeds signed represent only 20 percent to 30 percent of the value of the loan.) In some cases, guarantors may be to the clients for compulsory savings increases an asked to pledge security on behalf of the borrower. MFI's overall cost of funds. However, ADEMI strongly advocates that applicants Operating costs include salaries, rent, travel and trans- who are unable to provide collateral not be limited in their portation, administration, depreciation, and so forth. ability to access credit. When first-time borrowers are Depending on the approach selected, these costs vary unable to provide security, credit officers rely heavily on between 12 percent to 30 percent of outstanding loans. information as a collateral substitute. In these situations the Loan loss provisions vary depending on the quality of credit officer investigates the reputation of the applicant in the loan portfolio (discussed in chapter 9) and capital much the same way that an informal moneylender might. costs vary depending on the market rate of interest and Visits are made to neighboring shops and bars and appli- the inflation rate in the country (discussed in chapter 8). cants may be asked to produce financial statements (when Once an MFIs' costs have been considered, there are possible) or receipts for utilities or other payments. Each application is considered on a case-by-case basis. various ways to price loans to generate the estimated rev- enue required. (Note that only loans that are repaid on- Source: Benjamin and Ledgerwood 1998. time and in full earn the full amount of revenue expected. This is discussed further in chapter 9.) The fol- 140 MICROFINANCE HANDBOOK lowing sets out how to calculate interest rates and shows she or he is paying interest on only 500 rather than how service fees can augment interest revenue. 1,000. (Note that with interest paid on the declining balance, a greater portion of the monthly payment is Calculating Interest Rates paid in interest during the early months of the loan and a greater portion of principal is paid toward the end of There are several ways to calculate interest on a loan, of the loan. This results in a slightly larger amount than which two methods are most common: the declining bal- half of the principal remaining outstanding at the mid- ance method and the flat (face-value) method. Interest is point of the loan. In the example in table 5.2, in month generally paid over the term of the loan, although it is six 524.79 is still outstanding, not 500.) sometimes paid up front. To calculate interest on the declining balance, a finan- cial calculator is required. On most financial calculators, THE DECLINING BALANCE METHOD. This method calculates present value and payment must be entered with opposite interest as a percentage of the amount outstanding over signs, that is if present value is positive, payment must be the loan term. Interest calculated on the declining balance negative, or vice versa. This is because one is a cash inflow means that interest is charged only on the amount that and one is a cash outflow. Financial calculators allow the the borrower still owes. The principal amount of a one- user to enter different loan variables as follows: year loan, repaid weekly through payments of principal PV = Present value, or the net amount of cash and interest, reduces or declines every week by the disbursed to the borrower at the beginning amount of principal that has been repaid (table 5.2). This of the loan. means that borrowers have use of less and less of the origi- i = Interest rate, which must be expressed in same time units as n below. nal loan each week, until at the end of one year when they n = Loan term, which must equal the number have no principal remaining and have repaid the whole of payments to be made. loan (assuming 100 percent repayment). N For example, by month 6 of a 12-month loan for 1,000, PMT = Payment made each period. the borrower will only owe approximately 500 if she or N In the example above, a one-year loan of 1,000 with he has paid in regular weekly installments. At that point, monthly payments and 20 percent interest calcu- Table 5.2 Declining Balance Method Loan amount: 1,000; 12-month loan term; monthly loan payments: 92.63; interest rate: 20 percent. Outstanding Month Payments Principal Interest balance 0 -- -- -- 1,000.00 1 92.63 75.96 16.67 924.04 2 92.63 77.23 15.40 846.79 3 92.63 78.52 14.21 768.29 4 92.63 79.83 12.81 688.46 5 92.63 81.16 11.48 607.30 6 92.63 82.51 10.12 524.79 7 92.63 83.88 8.75 440.91 8 92.63 85.28 7.35 355.63 9 92.63 86.70 5.93 268.93 10 92.63 88.15 4.49 180.78 11 92.63 89.62 3.02 91.16 12 92.63 91.16 1.53 0.00 Total 1,111.56a 1,000.00 111.76a -- a. Difference of 0.2 is due to rounding. Source: Ledgerwood 1996. DESIGNING LENDING PRODUCTS 141 lated on the declining balance is computed by enter- weeks in the loan term, divided by 12 or 52 respec- ing the following: tively. PV = ­1,000 (enter as negative amount, as To calculate interest using the flat rate method the it is a cash outflow). interest rate is simply multiplied by the initial amount of i = 20 percent a year; 1.67 percent a month. the loan. For example, if an MFI charges 20 percent n = 12 months. interest using the flat rate method on a 1,000 loan, the interest payable is 200 (table 5.3). Solve for PMT : It is clear that the actual amount of interest charged PMT = 92.63. varies significantly depending on whether the interest is Total payments equal 1,111.56 (12 months at 92.63). calculated on the declining balance or the flat amount. Total interest is 111.56. The flat method results in a much higher interest cost The declining balance method is used by most, if not than the declining balance method based on the same all, formal financial institutions. It is considered the most nominal rate. In the example in table 5.3, interest of 200 appropriate method of interest calculation for MFIs as well. (20 percent flat basis) is 88.44 or 80 percent greater than interest of 111.56 (20 percent declining balance). THE FLAT METHOD. This method calculates interest as a To increase revenue some MFIs will change the interest percentage of the initial loan amount rather than the rate calculation method from declining balance to flat rather amount outstanding (declining) during the loan term. than increase the nominal rate. This may be in reaction to Using the flat method means that interest is always cal- usury laws imposing a maximum rate of interest that is not culated on the total amount of the loan initially dis- high enough to cover the MFI's costs. However, MFIs bursed, even though periodic payments cause the should realize that regardless of the nominal rate quoted, outstanding principal to decline. Often, but not always, clients are well aware of how much interest they are actually a flat rate will be stated for the term of the loan rather paying, based on the amount due each payment period. It is than as a periodic (monthly or annual) rate. If the loan important that all interest calculations be transparent. term is less than 12 months, it is possible to annualize These examples show that with all other variables the the rate by multiplying it by the number of months or same, the amount of interest paid on a declining balance Table 5.3 Flat Method Loan amount: 1,000; 12-month loan term; monthly loan payments: 100; interest rate: 20 percent Outstanding Month Payments Principal Interest balance 0 -- -- -- 1,000.00 1 100 83.33 16.67 916.67 2 100 83.33 16.67 833.34 3 100 83.33 16.67 750.01 4 100 83.33 16.67 666.68 5 100 83.33 16.67 583.35 6 100 83.33 16.67 500.02 7 100 83.33 16.67 416.69 8 100 83.33 16.67 333.36 9 100 83.33 16.67 250.03 10 100 83.33 16.67 166.70 11 100 83.33 16.67 83.37 12 100 83.33 16.67 0.00 Total 1200 1,000.00 200.00 -- Source: Ledgerwood 1996. 142 MICROFINANCE HANDBOOK loan is much lower than that on a loan with interest cal- on the declining balance, the effect of an increase in fees culated on a flat basis. To compare rates of interest calcu- is greater than a similar increase in the nominal interest lated by different methods it is necessary to determine rate (if interest is calculated on the declining balance). what interest rate would be required when interest is cal- N In example 5.2, the an MFI wants to determine its culated on the declining balance to earn the same nomi- future pricing policy. In doing so, it wants to calculate nal amount of interest earned on a loan with a flat basis the effect on the borrower of an increase in the interest calculation. rate and, alternatively, an increase in the loan fee. N In example 5.1, a 1,000 loan with 20 percent interest In example 5.1 a 20 percent interest rate (declin- calculated on a declining balance for one year with ing balance) on a 1,000 loan resulted in 112 in inter- monthly payments results in interest of 112 (rounded est revenue. A loan fee of 3 percent on this loan from 111.56). would result in a fee of 30, making total revenue 142. The same loan with interest calculated on a flat The MFI wants to increase its rate by 5 percentage basis results in interest of 200. points either through the loan fee it charges (from 3 To earn interest of 200 on a loan of 1,000 with percent to 8 percent) or the interest rate it charges interest calculated on the declining balance, the inter- (from 20 percent to 25 percent declining balance). est rate would have to increase by 15 percentage Each increase results in the following: points to 35 percent (additional interest revenue of N A loan fee of 8 percent on a 1,000 loan results in 88): fee revenue of 80. This represents an increase of 50 N Interest on a 1,000 loan at 35 percent declining bal- from the 3 percent fee (30). ance results in monthly payments of 99.96 for one N A interest rate of 25 percent (declining balance) on year or a total interest cost of 200 (rounded from a 1,000 loan results in interest revenue of 140 199.52). (monthly payments of 95). This represents an This example shows that an MFI calculating increase of 28 from a 20 percent interest rate (112). interest on the declining balance would have to Total revenue collected on a 1,000 loan with 20 increase its nominal interest rate substantially to earn percent interest (declining balance) and an 8 percent the same revenue as an MFI calculating interest on a fee equals 192 (112 + 80). Total revenue collected on flat basis. a 1,000 loan with 25 percent interest (declining bal- ance) and a 3 percent fee equals 170 (140 + 30). How Do Fees or Service Charges Affect the Borrower and The effect of a 5 percentage point increase in the the MFI? loan fees from 3 percent to 8 percent is greater than a 5 percentage point increase in the interest rate, pro- In addition to charging interest, many MFIs also charge vided interest is calculated on the declining balance. a fee or service charge when disbursing loans. Fees or ser- This is because the fee is charged on the initial loan vice charges increase the financial costs of the loan for the amount whereas the interest is calculated on the borrower and revenue to the MFI. Fees are often charged declining balance of the loan. as a means of increasing the yield to the lender instead of Although the interest rate may be the same nominal fig- charging nominal higher interest rates. ure, the costs to the borrower--and hence the yield to the Fees are generally charged as a percentage of the ini- lender--vary greatly if interest is calculated on a flat basis tial loan amount and are collected up front rather than or if fees are charged. This will be discussed further in the over the term of the loan. Because fees are not calculated section below on calculating effective rates of interest. Example 5.1 Interest 20% Interest 20% Interest 20% Interest 35% declining balance flat Difference flat declining balance Actual costs 112 200 88 200 200 DESIGNING LENDING PRODUCTS 143 Example 5.2 Service fee Service fee Interest 20% Interest 25% 3% 8% Increase declining balance declining balance Increase Actual costs 30 80 50 (167%) 112 140 28 (25%) Cross-Subsidization of Loans marketplace. However, it is important to design the deliv- ery of credit and savings products in a way that minimizes Some MFIs choose to subsidize certain products with transaction costs for both the client and the MFI. revenue generated by other more profitable products. When interest is calculated on the declining balance This is usually time-bound in some way until the new and there are no additional financial costs to a loan, the products have reached sustainability. Cross-subsidization effective interest rate is the same as the nominal interest of loans is not very common among MFIs, because many rate. Many MFIs, however, calculate the interest on a flat have only one or two loan products. However, there are basis, charge fees as well as interest, or require borrowers three significant examples, as shown in box 5.2. to maintain savings or contribute to group funds (trust or insurance funds). The cost to the borrower is, there- fore, not simply the nominal interest charged on the loan Calculating Effective Rates but includes other costs. Consideration must also be given to the opportunity cost of not being able to invest MFIs often speak about the "effective interest rate" on their loans. However, there are many ways in which effective rates are calculated, making it very difficult to Box 5.2 Cross-Subsidization of Loans compare institutions' rates. The effective rate of interest I N G RAMEEN B ANK 8 PERCENT INTEREST RATES ON is a concept useful for determining whether the condi- housing loans are subsidized by a 20 percent rate for gen- tions of a loan make it more or less expensive for the bor- eral loans. A housing loan has eligibility criteria: a mem- rower than another loan and whether changes in pricing ber must have received at least two general loans, she must policies have any effect. Because of the different loan have an excellent repayment record, and she must have variables and different interpretations of effective rates, a utilized her loans for the purposes specified on her appli- standard method of calculating the effective rate on a cations. To protect women and ensure accountability for loan (considering all variables) is necessary to determine the loan, Grameen Bank requires that women have title to the true cost of borrowing for clients and the potential the homestead land in her name. revenue (yield) earned by the MFI. The Unit Desa system in Bank Rakyat Indonesia gen- The effective rate of interest refers to the inclusion of all erates a significantly higher return on average assets than the bank as a whole. This profitability makes the Unit direct financial costs of a loan in one interest rate Desa system extremely important for the bank. Indirectly, Effective interest rates differ from nominal rates of loans made to poorer clients subsidize loans made to interest by incorporating interest, fees, the interest calcula- wealthier clients at lower interest rates. tion method, and other loan requirements into the finan- The Association for the Development of Microenterprises cial cost of the loan. The effective rate should also include provides microloans and small business loans. Although the cost of forced savings or group fund contributions by the interest rate and fees charged on microloans are higher the borrower, because these are financial costs. We do not than those charged on smaller loans, some cross-subsidiza- consider transaction costs (the financial and nonfinancial tion still exists whereby the larger loans generate sufficient costs incurred by the borrower to access the loan, such as revenue to subsidize any losses resulting from the opening a bank account, transportation, child-care costs, microloan portfolio. or opportunity costs) in the calculation of the effective Source: Author's findings. rate, because these can vary significantly depending on the 144 MICROFINANCE HANDBOOK the money that the borrower must pay back in regular calculating the effective rate if no financial calculator or installments (the time value of money). spreadsheets are available. Variables of microloans that influence the effective rate include: Estimating the Effective Rate I Nominal interest rate I Method of interest calculation: declining balance or If you do not have access to a financial calculator or a flat rate computer spreadsheet, you can compute an estimation of I Payment of interest at the beginning of the loan (as a the effective rate (see Waterfield 1995). The estimation deduction of the amount of principal disbursed to the method considers the amount the borrower pays in inter- borrower) or over the term of the loan est and fees over the loan term. The estimation method I Service fees either up front or over the term of the can be used to determine the effect of the interest rate loan calculation method, the loan term, and the loan fee. An I Contribution to guarantee, insurance, or group fund estimation of the effective rate is calculated as follows: I Compulsory savings or compensating balances and the corresponding interest paid to the borrower either Amount paid in interest and fees Effective cost = by the MFI or another institution (bank, credit Average principal amount outstanding union) Note: I Payment frequency (Sum of principal amounts outstanding) Average principal = I Loan term amount outstanding number of payments I Loan amount. When all variables are expressed as a percentage of the To calculate the effective cost per period, simply loan amount, a change in the amount of the loan will not divide the resulting figure by the number of periods. change the effective rate. A fee that is based in currency As previously illustrated, the amount of interest rev- (such as $25 per loan application) will change the effec- enue is largely affected by whether interest is calculated on tive rate if the loan amount is changed; that is, smaller a flat or declining balance basis. With all other variables loan amounts with the same fee (in currency) result in a the same, the effective rate for a loan with interest calculat- higher effective rate. ed on a declining balance basis will be lower than the effec- Calculation of the effective rate demonstrates how dif- tive rate for a loan with interest calculated on a flat basis. ferent loan product variables affect the overall costs and Using an example similar to that in tables 5.2 and revenues of the loan. Two methods of calculating the 5.3, the effective rate is estimated for a 1,000 loan with effective rate of interest are an estimation method, which interest of 20 percent and a 3 percent fee, first with inter- uses a formula that does not require a financial calcula- est calculated on the declining balance and then with tor, and the internal rate of return method. interest calculated on the flat basis. Note that the estimation method does not directly Calculating the interest on the declining balance take into account the time value of money and the fre- results in an estimated annual effective rate of 25 percent quency of payments, which are considered in the internal or 2.1 percent per month (table 5.4). rate of return method. Although the difference may be Calculating the interest on a flat basis results in an minimal, the greater the length of the loan term and the estimated annual effective rate of 42 percent or 3.5 per- less frequent the loan payments, the more substantial the cent per month (table 5.5). difference will be. This is because the longer the loan is With all other factors the same, the effective rate outstanding and the less frequent the payments, the increases from 25 percent (2.1 percent per month) to 42 greater the effect on the cost will be and hence the differ- percent (3.5 percent per month) when the method of ence between the estimated effective cost and the internal calculation is changed from declining balance to flat. rate of return calculation. In addition, the estimation The effective rate also increases when the loan term is method does not take into account compulsory savings shortened if a fee is charged. This is because fees are calculat- or contributions to other funds, such as trust or insur- ed on the initial loan amount regardless of the length of the ance funds. It is presented here simply as a method for loan term. If the loan term is shortened, the same amount of DESIGNING LENDING PRODUCTS 145 Table 5.4 Effective Rate Estimate, Declining Balance Loan amount: 1,000; 12-month loan term; monthly loan payments: 92.63; interest rate: 20 percent; fee: 3 percent (30) Outstanding Month Payments Principal Interest balance 0 -- -- -- 1,000.00 1 92.63 75.96 16.67 924.04 2 92.63 77.23 15.40 846.79 3 92.63 78.52 14.21 768.29 4 92.63 79.83 12.81 688.46 5 92.63 81.16 11.48 607.30 6 92.63 82.51 10.12 524.79 7 92.63 83.88 8.75 440.91 8 92.63 85.28 7.35 355.63 9 92.63 86.70 5.93 268.93 10 92.63 88.15 4.49 180.78 11 92.63 89.62 3.02 91.16 12 92.63 91.16 1.53 0.00 Total 1,111.56a 1,000.00 111.76a 6,697.08 Effective rate = (111.76 + 30)/558.09* = 25% (2.1% per month) * 6,697.08 / 12 months = 558.09 a. Difference of 0.2 is due to rounding. Source: Ledgerwood 1996. Table 5.5 Effective Rate Estimate, Flat Method Loan amount: 1,000; 12-month loan term; monthly loan payments: 100; interest rate: 20 percent; fee: 3 percent (30) Outstanding Month Payments Principal Interest balance 0 -- -- -- 1,000.00 1 100 83.33 16.67 916.67 2 100 83.33 16.67 833.34 3 100 83.33 16.67 750.01 4 100 83.33 16.67 666.68 5 100 83.33 16.67 583.35 6 100 83.33 16.67 500.02 7 100 83.33 16.67 416.69 8 100 83.33 16.67 333.36 9 100 83.33 16.67 250.03 10 100 83.33 16.67 166.70 11 100 83.33 16.67 83.37 12 100 83.33 16.67 0.00 Total 1,200 1,000.00 200.00 6,500.22 Effective rate = (200 + 30)/541.69* = 42% (3.5% per month) *6,500.22 / 12 months = 541.69 Source: Ledgerwood 1996. 146 MICROFINANCE HANDBOOK money needs to be paid in a shorter amount of time, thus worth more than 1,000 a year from now. In other increasing the effective rate. (This difference is greatest when words, 1,000 that is to be received a year from now has a fee is charged on a loan with interest calculated on the a present value of less than 1,000. How much less declining balance. This is because the shorter loan term depends on how much can be earned if the funds are increases the relative percentage of the fee to total costs.) invested. This is the time value of money principle. The The effective rate can be estimated for a number of present value of 1,000 to be received any number of loan variables, including an increase in the loan fee and a years in the future can be calculated as follows: decrease in the loan term. Table 5.6 illustrates the effect PV = 1 / (1 + i )n that a change in the loan fee and a change in the loan term have on the effective rate (the examples are calculat- where i = the interest rate and n = the number of periods ed on both a flat and declining balance basis). until the expected receipt of funds. To incorporate pre- Note that the effect of an increase in the fee by 5 per- sent value into the effective rate calculation, a financial cent (to 8 percent) has the same effect (an increase of 0.8 calculator is required. percent per month in effective rate) whether the loan is Note that for the purposes of calculating the effec- calculated on a declining basis or flat method. This is tive rate, the assumption is that there are no late pay- because the fee is calculated on the initial loan amount. ments or defaulted loans (that is, that there is no credit risk--a situation that is rarely the case, but we are try- Calculating the Effective Interest Rate with Compulsory ing to determine the full cost to the borrower and the Savings or Other Loan Variables yield to the lender before taking risk into account). It is also assumed that the present value for the series of To determine the effective rate of interest considering all loan payments is equal to the rate of return that would financial costs of a loan, including the time value of be earned by the borrower if she reinvested the loan money, compulsory savings, and contributions of other proceeds and paid it back all at the end of the maturity funds, the internal rate of return method is used. As period rather than in regular installments. In doing so, explained above, compulsory savings should be consid- we take into account the opportunity cost associated ered a component of the loan rather than a separate sav- with not having the full amount of the loan for the full ings product. To calculate the true cost of borrowing all loan term. cost components must be considered. Calculating the internal rate of return for each alter- The internal rate of return is defined as the specific inter- native involves three steps: (adapted from Rosenberg est rate by which the sequence of installments must be discount- 1996) determining the actual cash flows; entering the ed to obtain an amount equal to the initial credit amount. cash flows into the calculator to determine the effective To calculate the internal rate of return, an under- rate for the period; and multiplying or compounding standing of present value is required. The concept of pre- the internal rate of return by the number of periods to sent value is based on the notion that 1,000 today is determine the annual rate (this will be discussed further Table 5.6 Change in Loan Fee and Loan Term Effect Calculation 20% Service fee Effective cost per Change Fee/term annual rate (percent) Loan term month (percent) (percent) 3% fee; 12-month term Flat 3 12 months 3.5 Raise fee to 8% Flat 8 12 months 4.3 0.8 Lower term to 3 months Flat 3 3 months 4.0 0.5 3% fee; 12-month term Declining balance 3 12 months 2.1 Raise fee to 8% Declining balance 8 12 months 2.9 0.8 Lower term to 3 months Declining balance 3 3 months 3.2 1.1 Source: Ledgerwood 1996. DESIGNING LENDING PRODUCTS 147 in the section below dealing with the difference between Table 5.7 summarizes the six variables above and effective rate and effective yield.) the corresponding effective rates for borrowing 1,000 To determine the effective rate by calculating the for four months (calculations provided in appendix 2). internal rate of return, the loan variables are entered into The effect of the loan term is illustrated through a financial calculator or computer spreadsheet (such as calculating each variable with a six-month loan term Lotus 1-2-3 or Excel). This analysis assumes that all cash rather than a four-month loan term (calculations not flows are constant (that is, loan payments are the same shown). for the entire loan term). As table 5.7 shows, the effective rate (assuming the same nominal rate) is increased by all of the following vari- PV = Present value, or the net amount of cash ables: disbursed to the borrower at the beginning I Flat interest calculation instead of declining balance of the loan. calculation i = Interest rate, which must be expressed I Up-front interest payments in same time units as n below. I Loan (or service) fees n = Loan term, which must equal the number I More frequent payments of payments to be made. I Compulsory savings or group fund contributions. When the loan term is extended, the increase in the PMT = Payment made each period. effective rate from the base case is greater if interest is cal- FV = Future value, or the amount remaining culated on a flat basis (since the borrower pays 3 percent after the repayment schedule is completed a month in interest for an additional two months), inter- (zero except for loans with compulsory est is paid up front, or compulsory savings or group fund savings that are returned to the borrower). contributions are required. To demonstrate the internal rate of return method of When the loan term is extended, the increase in the calculating effective rates, the rate calculation for a sam- effective rate from the base case is lower if there are ser- ple loan with six alternatives is presented to illustrate the vice fees or payments are made more frequently. effect of different loan variables, including: I Flat interest calculation Calculating the Effective Interest Rate with Varying I All interest paid up front Cash Flows I Loan fee (service fee) I Change in payment frequency It is also possible to calculate the effective rate using the I Compulsory savings with interest internal rate of return method and taking into account I Contributions to group funds (no interest). cash flows that vary during the loan term. This is because Each of these variables is then calculated if the loan the internal rate of return calculation can consider each term increases. cash flow rather than a constant stream. This would be Table 5.7 Variables Summary Effective rate, Effective rate, Loan variables 4­month term (percent) 6­month term (percent) Base case, declining balance 36.0 36.0 Flat interest 53.3 59.3 Up-front interest payment 38.9 40.2 Loan service fee 51.4 47.1 Payment frequency 45.6 43.4 Compulsory savings 39.1 42.0 Group fund contribution 41.2 44.7 Source: Adapted from Rosenberg 1996. 148 MICROFINANCE HANDBOOK useful for loans that have a grace period during the loan rather than simply multiplied by the number of peri- term or if a lump sum payment is made at the end of the ods, because we assume that the reinvestment rate of loan term. For a discussion of calculating the internal rate the borrower is equal to the internal rate of return. In of return on a loan with varying cash flows, see appendix 3. other words, the internal rate of return per period is the amount the borrower forgoes by repaying the loan How does the Effective Cost for the Borrower Differ in installments (that is, the principal amount available from the Effective Yield to the Lender? declines with each installment). Therefore the return earned per period if the borrower could reinvest Yield refers to the revenue earned by the lender on the port- would compound over time. folio outstanding, including interest revenue and fees. I For the yield to the lender, the per period rate is multi- Calculating the effective yield earned on the portfolio plied, because we assume that the revenue generated is allows an MFI to determine if enough revenue will be used to cover expenses and is not reinvested (only the generated to cover all costs, leading to financial self- principal is reinvested). Therefore, the average port- sufficiency. Projecting the effective yield is also useful for folio outstanding does not increase (unless revenue forecasting revenues and determining the effect of exceeds expenses). changes in pricing policies on revenue. To compound the rate, the periodic internal rate of The projected effective yield is also useful for MFIs to return must be stated as a decimal amount rather than as compare to the actual yield on their portfolios. a percentage amount. To do this, the periodic internal The effective yield to the lender differs from the effec- rate of return is divided by 100. To annualize the period- tive cost to the borrower for two reasons: ic rate, the following formula is used: I If there are components of the loan pricing, such as a annual internal rate of return = (1 + IRRp) ­1 compulsory savings or fund contributions, that are not held by and hence do not result in revenue to the where IRRp equals the internal rate of return per period lender (for example, if savings are put into a commer- divided by 100 and a equals the number of periods in a cial bank or fees are charged to the borrower by another year (a = 12 if p is one month; a = 32 if p is one week). institution, such as training fees), they are not included N For example, a 1,000 loan, repaid in four equal in the yield calculation but are included in the effective monthly payments of principal and interest, with an interest rate (for the borrower) calculation. interest rate of 36 percent per year, calculated on a flat I The expected yield to the lender differs because the basis results in the following internal rate of return: cost to the borrower is determined on the initial loan PV = ,1000; PMT = ­280; n = 4 amount, whereas the return to the lender must be based on the average portfolio outstanding. This Solving for i yields an internal rate of return of 4.69 results in differing calculations for annualizing period- percent. ic interest rates--periodic rates are either multiplied The borrower's effective rate is 73.3 percent and by the number of periods or compounded. the yield to the lender is closer to 56.3 percent, a dif- If all elements of the loan are both costs to the bor- ference of 17 percent (example 5.3). rower and revenue to the lender, then the only difference The effective yield on the portfolio of an MFI is between the effective cost and the effective yield is the reduced by: method used to annualize the rate. I The amount of delinquent (or non-revenue-generating) In the examples above, for simplicity's sake we multi- loans plied the periodic interest rate to achieve an annual rate. I Late payments However, the decision to multiply or compound the I Low loan turnover (idle funds) periodic rate is based on whether or not the cost to the I Fraud borrower or the yield to the lender is being determined. I Reporting errors or failures that lead to delayed collec- I For the cost to the borrower, the internal rate of return tion of loans is compounded by the number of payment periods in a I Prepayments if interest is calculated on the declining year. The internal rate of return must be compounded balance and funds remain idle. DESIGNING LENDING PRODUCTS 149 The loan loss rate (LL) represents the annual loss due Example 5.3 to defaulted loans. Past loan loss experience is an impor- Annual rate, compounded (1.0469)12 tant indicator of this rate. The loan loss rate may be con- = 1.733 siderably lower than the delinquency rate: the former = (1.733) ­ 1 x 100 reflects loans that must actually be written off, while the = 73.3% latter reflects loans that are not paid on time--many of Annual rate, multiplied 4.69 x 12 periods which will eventually be recovered. Microfinance organi- = 56.3% zations with loan loss rates greater than 5 percent tend not to be viable. Many good organizations run at about 1 Appendix 1. How Can an MFI Set a to 2 percent of average outstanding portfolio. Sustainable Rate on Its Loans? The cost of funds (CF) rate takes into account the actual cost of funds of the organization when it funds its MFIs can determine the rate necessary to charge on loans portfolio with savings and commercial debt. When an based on their cost structure. The following is one MFI also benefits from concessional funding, the calcu- method of approximating the effective interest rate that lation must include an estimation of the funding costs if an MFI will need to charge on its loans to cover all of its it were replaced with commercial debt and equity. Prior costs and thus be sustainable (adapted from Rosenberg to determining the cost of funds rate, an estimation of 1996). the financial assets (excluding fixed assets) and the pro- Note that this method assumes a mature MFI with portion of debt and equity to fund these assets needs to relatively stable costs, that is, start-up costs have already be determined, based on future funding policies. (Note been amortized and the MFI is operating at full capacity. that the proportion of total assets that is nonfinancial-- It is understood that an MFI should not expect to break or nonproductive--will have a large impact on the even at every point along its average cost curve. A certain financial margin required to become sustainable.) Once scale of operations is required to break even and hence to this is done, two methods are suggested to determine the make this method applicable. cost of funds: The annualized effective yield (R) charged on loans is a I The estimation method: multiply the financial assets function of five elements, each expressed as a percentage of by the higher of the rate that local banks charge medi- average outstanding loan portfolio (LP): administrative um-quality commercial borrowers or the inflation rate expenses (AE), the cost of funds (CF), loan losses (LL), the projected for the planning period. desired capitalization rate (K), and investment income (II). I The weighted average cost of capital method: based on the sources used to fund the financial assets, including AE + CF + LL + K R = ­ II loans to the organization, deposits if licensed to collect, 1 ­ LL and equity. (This does not consider the funding of Each variable should be expressed as a decimal fraction fixed assets and cash holdings, that is, nonproductive (percentage of average loan portfolio). For example, oper- assets. Depending on the proportion of nonproductive ating expenses of 200,000 on an average loan portfolio of assets, the impact on the financial margin required 800,000 would yield a value of 0.25 for the AE rate. may be substantial.) This method estimates the Included as administrative expenses (AE) are all annu- absolute amount of the annual cost of financing. For al recurrent costs except the cost of funds and loan losses, loans to the organization, the commercial bank lending including salaries, benefits, rent, utilities, and deprecia- rate to medium-quality borrowers should be used. For tion. Also included are the value of any donated com- deposits mobilized by the MFI, the average local rate modities or services, such as training or technical paid on deposits, including an adjustment for legal assistance that the microfinance organization will have to reserve requirements, should be used. For equity, the pay for in the future as it grows independent of donor projected inflation rate should be used, because infla- subsidies. Administrative expenses of efficient mature tion represents a real annual reduction in the purchas- organizations tend to range between 10 and 25 percent ing power of the organization's net worth. (Some of their average loan portfolio. experts argue in fact that a market rate for equity 150 MICROFINANCE HANDBOOK should be used, because in the marketplace, equity is effective yield). Because microentrepreneurs generally generally more costly than debt because of the greater have difficulty accessing credit, economic theory predicts risk inherent with equity. Using the inflation rate may that on average they will be able to use each additional understate the true cost of the equity.) unit of capital more profitably than richer households or Calculate the total cost by adding together the firms. Based on rates charged by moneylenders in most costs for each class of funding. It is important in this developing countries, relatively high interest rates do not analysis to express the cost of funds as a percentage of the seem to deter microclients. The most important point to average loan portfolio, not simply as the estimated cost of bear in mind when determining the rate to charge is the debt and equity. This is because every variable in the efficiency of the organization. The purpose is to provide formula above is expressed as a percentage of the average clients with long-term continued access to the organiza- loan portfolio to provide costs on a consistent measure. tion, which can only be achieved if all costs are covered. The capitalization rate (K) represents the net real profit If an organization is inefficient in delivering credit and that the organization would like to achieve, expressed as a consequently needs to charge substantially higher rates percentage of the average loan portfolio. Accumulating of interest, clients may not find value in the credit based profit is important because the amount of outside funding on its price (interest and fees), and the organization will an MFI can safely borrow is a function of the amount of not survive. its equity (leverage). Once that limit is reached, further growth requires increases in its equity base. The rate of real profit targets depends on how aggressive a microfi- Appendix 2. Calculating an Effective nance organization is and its desired rate of growth. To Interest Rate Using the Internal Rate support long-term growth, a capitalization rate of 5 to 15 of Return Method percent of the average loan portfolio is suggested. The investment income rate (II) is the income expect- To demonstrate the internal rate of return method of cal- ed to be generated by the organizations' financial assets, culating effective rates, the rate calculation for a sample excluding the loan portfolio. Some of these, such as cash, loan with six alternatives is presented (adapted from checking deposits, and legal reserves, will yield little or Rosenberg 1996). The effective rate is calculated first for no interest; others, such as certificates of deposits and a "base case" (interest calculated on the declining bal- investments, may produce significant income. This ance; no fees or compulsory savings). Then the effective income, expressed as a percentage of the loan portfolio, is rate is calculated to illustrate the effect of different loan entered as a deduction in the pricing equation. variables, including: For example, if an organization has operating expens- I Flat interest calculation es of 25 percent of its average loan portfolio, 23.75 per- I Up-front interest payments cent cost of funds (stated as a percentage of its average I Loan (or service) fee loan portfolio), loan losses of 2 percent, a desired capital- I Change in payment frequency ization rate of 15 percent, and investment income of 1.5 I Compulsory savings with interest percent, the resulting effective annual interest rate is: I Contributions to group funds (no interest). AE + CF + LL + K Calculating the internal rate of return for each alter- R = ­ II native involves three steps: determining the actual cash 1 ­ LL flows, entering the cash flows into the calculator to .25 + .2375 + .02 + .15 determine the effective rate for the period, and multiply- R = ­ .015 1 ­ LL ing or compounding the internal rate of return by the number of periods to determine the annual rate. R = 65.6 While this rate may seem high, the issue of what Base Case: Declining Balance Interest effective rate to charge depends on what the market will bear. (Also, the effective rate can be a combination of Loan amount 1,000; repaid in four equal monthly interest and fees, as discussed in the section above on payments of principal and interest. Nominal interest DESIGNING LENDING PRODUCTS 151 rate is 36 percent a year, or 3 percent a month, calcu- Alternative 4: Weekly Payments lated on the declining balance. The effective interest rate in the base case is the same as the stated nominal Same as base case, except that four month's worth of rate. payments are paid in 16 weekly installments. To compute monthly payment (note that on most Effective interest rate: financial calculators, present value and payment must be PV = 1,000; PMT = ­67.26; n = 16 entered with opposite signs): Solving for i yields an effective weekly rate of 0.88 PV = ­1,000; n = 4; i = 36/12 = 3 percent, which is multiplied by 52 for an annual percent- Solving the equation for PMT yields a monthly pay- age rate of 45.6 percent. ment of $269.03. Alternative 5: Compulsory Savings with Interest Alternative 1: Flat Interest Same as the base case, except that as a condition of the Same as base case, except that interest is calculated on the loan the client is required to make a savings deposit of entire loan amount (flat basis) rather than on the declin- 50 along with each month's payment. The savings ing balance and is prorated over the four monthly pay- account yields of 1 percent per month, uncompounded, ments. is available to the client for withdrawal at any time after Effective interest rate: the end of the loan. (Alternatives 5 and 6 assume that the MFI receives and PV = 1,000; PMT = ­280; n = 4 holds the compulsory savings and group contributions. In Solving for i yields an effective monthly rate of 4.69 per- such a case, the yield to the organization and the cost to cent, which is multiplied by 12 for an annual percentage the client are the same. If compulsory savings are held by rate of 56.3 percent. someone other than the MFI, such as a bank, then the amounts deposited should not enter into the computation Alternative 2: Up-front Interest Payments of yield to the microfinance organization but should be included in calculating the cost to the borrower.) Same as base case (interest calculated on declining bal- Effective interest rate: ances) but all interest is charged at the beginning of the PV = 1,000; PMT = ­319.3; n = 4; FV = 203 loan. Effective interest rate: Solving for i yields an effective monthly rate of 3.26 per- cent, which is multiplied by 12 for an annual percentage PV = 923.88; PMT = ­250; n = 4 rate of 39.1 percent. Solving for i yields an effective monthly rate of 3.24 per- cent, which is multiplied by 12 for an annual percentage Alternative 6: Group Fund Contribution rate of 38.9 percent. Same as the base case, except that as a condition of the loan Alternative 3: Loan or Service Fee the client is required to make a contribution to a group fund of 50 along with each month's payment. No interest Same as base case except that a 3 percent loan fee is is paid on the group fund, although it is available to the charged up front. client for withdrawal at any time after the end of the loan. Effective interest rate: Effective interest rate: PV = 970; PMT = ­269.03; n = 4 PV = 1,000; PMT = ­319.3; n = 4; FV = 200 Solving for i yields an effective monthly rate of 4.29 per- Solving for i yields an effective monthly rate of 3.43 per- cent, which is multiplied by 12 for an annual percentage cent, which is multiplied by 12 for an annual percentage rate of 51.4 percent. rate of 41.2 percent. 152 MICROFINANCE HANDBOOK Appendix 3. Calculating the Effective Rate To calculate this example using a financial calculator with Varying Cash Flows the following variables are entered: CFo = 970; CFj = 269.03; Nj = 2; CFj = 0; Nj = 2; CFj = 269.03; Nj = 2; i = 3 It is also possible to calculate the effective rate by using the internal rate of return method to take into account Solve for internal rate of return (IRR). cash flows that vary during the loan term. This is This results in a much lower effective rate than when because the internal rate of return calculation can con- there was no grace period. The effective rate in this sider each cash flow rather than a constant stream. This example is 36.81 percent, while a loan with a four- would be useful for loans that have a grace period at month loan term and no grace period would have an some point during the loan term or if a lump sum pay- effective rate of 51.4. This is because the client had use ment is made at the end of the loan term. of the funds for six months rather than four and paid N For example, a 1,000 loan with interest at 36 percent the same amount of interest and fees as with a four- per year calculated on the declining balance, a fee of month loan. 3 percent, and a loan term of six months, with a two- To calculate the effective rate for a loan that is month grace period during the middle of the loan repaid in a lump sum at the end of the loan term, the (resulting in four monthly payments), results in the interest rate for the loan term is considered the period internal rate of return shown in table A5.3.1. rate and, therefore, the effective rate. However, with Table A5.3.1 Internal Rate of Return with Varying Cash Flows (Grace Period) Nominal annual interest rate: 36 percent; calculation method: declining balance; service fee: 3 percent of loan amount; payment fre- quency: monthly; loan term: six months; loan amount: 1,000 Period Inflow Outflow Net flow 0 1,000 ­30 970 1 -- ­269 ­269 2 -- ­269 ­269 3 -- 0 0 4 -- 0 0 5 -- ­269 ­269 6 -- ­269 ­269 Total 1,000 1,106 106 Internal rate of return = 3.0678 (or 36.8% annual). Table A5.3.2 Internal Rate of Return with Varying Cash Flows (Lump Sum) Nominal interest rate: 3 percent per month; calculation method: declining balance; service fee: 3 percent of loan amount; payment frequency: monthly (interest only); loan term: four months; loan amount: 1,000 Period Inflow Outflow Net flow 0 1,000 ­30 970 1 -- ­19 ­19 2 -- ­19 ­19 3 -- ­19 ­19 4 -- ­1,019 ­1,019 Total 1,000 1,106 106 Internal rate of return = 1.747 (or 21.0% annual) DESIGNING LENDING PRODUCTS 153 loans that require the interest to be paid in install- Sources and Further Reading ments and the principal to be paid at the end of the loan term, the internal rate of return calculation is Benjamin, McDonald, and Joanna Ledgerwood. 1998. "The used. Association for the Development of Microenterprises N For example, a 1,000 loan with interest at 3 percent (ADEMI): `Democratising Credit' in the Dominican per month calculated on the declining balance, a fee Republic." World Bank, Sustainable Banking with the of 3 percent, and a loan term of four months, with Poor, Washington, D.C. Ledgerwood, Joanna. 1996. Financial Management Training for interest paid monthly and the principal paid at the Microfinance Organizations: Finance Study Guide. New end of the loan term, results in the internal rate of York: PACT Publications (for Calmeadow). return shown in table A5.3.2. Rosenberg, Rich. 1996. "Microcredit Interest Rates." CGAP To calculate the above example using a financial cal- Occasional Paper 1. World Bank, Consultative Group to culator the following variables are entered: Assist the Poorest (CGAP), Washington, D.C. CFo = 970; CFj = 19; Nj = 3; CFj = 1,019; i = 3 Von Pischke, J.D. 1991. Finance at the Frontier: Debt Capacity and the Role of Credit in the Private Economy. Washington, The substantially lower effective rate is a result of D.C.: World Bank, Economic Development Institute. the borrower having use of the full amount of the Waterfield, Charles. 1995. "Financial Viability Model-- principal (1,000) for the entire loan term of four Facilitator's Notes." SEEP Network. New York. months. For the MFI to earn an effective yield equal Women's World Banking. 1994. "Principles and Practices of to a loan that has principal and interest paid in install- Financial Management for Microenterprise Lenders and ments, the nominal rate would have to be increased Other Service Organizations." In Women's World Banking considerably. Best Practice Workbook. New York. CHAPTER SIX Designing Savings Products S avings services are often not available to MFI for on-lending from donors at lower rates than they clients for two main reasons. First, there is a would have to pay depositors. Furthermore, accessing mistaken belief that the poor cannot and do not donor funding is often less costly than developing the save, meaning that their demand for savings products infrastructure required to mobilize deposits. goes unheard and unnoticed. Second, due to the regu- If an MFI is to mobilize savings effectively, there latory constraints of most MFIs, many are not legally must be suitable economic and political environments allowed to mobilize deposits. As these reasons rein- in the country in which it is working. Reasonable lev- force and perpetuate each other, deposits from the poor els of macroeconomic management and political sta- are formally mobilized only infrequently. Furthermore, bility are required because they affect the rate of as MFIs expand their operations, few are able to inflation, which influences the ability of an MFI to increase their outreach to a significant number of offer savings services in a sustainable manner. In addi- clients unless they increase their funding sources to tion, an appropriate and enabling regulatory environ- include voluntary deposits. MFIs that depend on exter- ment is necessary. nal sources of donated funding often concentrate on Finally, the institution itself must have established the demands of donors rather than on the demands of a good record of sound financial management and potential clients, especially potential savings clients internal controls, which assures depositors that their (GTZ 1997a). funds will be safely held by the MFI. "Savings mobilization is considered now to be a This chapter highlights the demand for savings ser- crucial factor in the development of sound vices and the environment in which the MFI operates, financial markets. There are a growing number including the legal requirements for providing savings of successful savings mobilization programs in services. This activity implies a necessary development developing countries and governments and of the institutional capacity of the MFI, including its international agencies have recently become human resources, infrastructure, security, manage- much more interested in such activities. Rural ment information system, and risk management. The and non-wealthy households in particular have discussion in this chapter includes an overview of the become the focus of policies to promote savings, types and pricing of savings products that MFIs pro- as the myths that the poor have no margin over vide and the costs associated with delivering savings consumption for saving and do not respond to services. Its focus is on the design of voluntary savings economic incentives are increasingly being ques- products, which are a separate and distinct product tioned." (FAO 1995, 14) from loans and are generally accessible whether clients Subsidized credit funds also contribute to the lim- are borrowers or not. Voluntary savings differ from ited mobilization of savings deposits. MFIs receiving compulsory savings (discussed in the previous chap- subsidized funding have little or no incentive to ter) in that they are not a mandatory requirement for mobilize deposits, because they have funds available receiving a loan. 155 156 MICROFINANCE HANDBOOK This chapter will be of primary interest to readers that low-income clients can save substantial sums, tradi- working with MFIs that are legally able to accept volun- tional banks are not set up to collect small amounts and tary savings (or are considering creating an institutional may not find it cost efficient to do so. structure that allows for the mobilization of voluntary Microentrepreneurs, like other business people, save savings). Few MFIs are currently authorized to accept for at least five reasons (Robinson 1994): voluntary deposits under current regulations, which I Consumption and for consumer durables means that there is limited experience to draw from. I Investment However, as the field matures, more donors and practi- I Social and religious purposes tioners are realizing that there is a demand for savings I Retirement, ill health, or disability services and best practices will be further developed. This I Seasonal variations in cash flow. chapter addresses the issues that an MFI must consider The people who make up the target market of most if when introducing voluntary savings. Practitioners, con- not all MFIs need access to savings services that ensure sultants, and donors involved with these organizations their funds will be safe, liquid, and divisible. Savings may not be fully aware of the complexities involved in clients are interested in three major benefits: adding voluntary savings services, and the purpose of this I Convenience. Clients want access to savings services chapter is to raise awareness of these complexities. without taking too much time away from their busi- nesses. I Liquidity. Clients want access to their savings when Demand for Savings Services needed. As discussed at the beginning of chapter 5, the cash flow patterns of microentrepreneurs vary. During times of Box 6.1 Deposit Collectors in India excess cash flow clients need a safe and convenient way to IN KALANNAGAR, INDIA, LOW-INCOME CLIENTS USE THE save. When cash flow is limited these same clients need services of a "deposit collector." Each saver has a crudely to be able to access their savings (in other words, they printed "card" with 220 cells on it arranged in 20 columns need a liquid form of savings). and 11 rows. Every second day a woman collector comes Without access to savings services, clients often save to the saver's door to collect a set amount, which then fills by keeping cash in the household or investing in grain, up one cell. The saver can fill the cells up as quickly or as livestock, gold, land, or other nondivisible assets. Saving slowly as she likes. For example, she can deal with several in cash is often unsafe, particularly for women, because cells in a day and then stop her savings for days on end. the male head of the household may demand that any Standard "per cell" deposits are 50 paise and 1, 2, 5, and excess money be spent. In many communities household 10 rupees. As soon as all 220 cells are filled she receives the income is shared, and if family members know that one value of 200 of them. Assuming she fills in a cell every day, paying one rupee per cell, she deposits 220 rupees over member has money they may demand that it be given 220 days (an average investment of 110 rupees). She thus (or lent) to them. Furthermore, cash can be stolen or pays 20 rupees to have this service. This results in a nega- destroyed by fires or floods. tive interest rate on her savings of approximately 30 per- Saving in hard assets such as grain or animals does not cent (or simply, a fee of 10 percent). provide safety (the animal may die), liquidity (market These savers know very well that they are paying to values may change or a market may not exist at the time save. However, they point out that despite the small of the sale), or divisibility (they cannot sell half a goat if expense the system is still a benefit to them, particularly if they need only a portion of their savings). In addition, they have school fees or some other expense coming up. maintaining livestock, land, or grain may incur costs that Saving money at home is difficult and saving a few rupees add to cash flow difficulties when cash is unavailable. a day in a bank is inefficient because the saver must visit Low-income clients are often unable to access savings the branch often. Savers feel it is well worth the fee to services from traditional banks due to a limited branch have the service come to their door. network or the reluctance of banks to deal with small Source: Rutherford 1996. amounts of money. Although it has been demonstrated DESIGNING SAVINGS PRODUCTS 157 I Security. Clients want to be sure that their savings are Bank Rakyat Indonesia is one of the best known cases safe and that the institution that collects them is stable. of a state bank providing savings services to low-income For the most part, savings clients have not considered clients in a sustainable manner. Bank Rakyat Indonesia the interest earned on savings to be a priority; however, it has shown that microentrepreneurs utilize savings ser- does become a priority when resources are scarce and there vices more than credit services (box 6.2). are many profitable investment opportunities. Legal Requirements for Offering Voluntary Savings Services Is There an Enabling Environment? There are two major legal requirements that generally To mobilize savings effectively, an MFI needs to be oper- need to be fulfilled before an MFI can offer voluntary ating in a country in which the financial sector has been savings services: licensing and reserve requirements. liberalized. This includes abolishing interest rate ceilings and foreign exchange controls, admitting new entrants L ICENSING . For the most part, an MFI needs to be into the market, as well as establishing reasonable capital licensed to collect savings. This usually means that it requirements. Furthermore, the government must not becomes subject to some form of regulation. allow unqualified institutions to mobilize public savings. When discussing voluntary savings services, it is Nor must is allow more institutions to mobilize savings important to differentiate between savings services that than the supervisory body is able to supervise effectively. are provided to borrowers or members and those avail- Savings clients cannot be expected to have enough able to the general public. Many informal or semiformal information to evaluate the soundness of the institution MFIs collect savings from their members and either on- with which they save. The government must supervise lend these savings or deposit them in a formal financial institutions that mobilize public deposits, either directly institution. Although some are licensed (as NGOs), they or through an effectively managed body that it approves. are not necessarily licensed as financial intermediaries This usually entails a willingness on the part of the gov- (that is, to accept deposits) and are generally not super- ernment to modify its banking supervision so that MFIs vised or regulated in their deposit activities. are appropriately regulated and supervised (see chapter 1). To accept deposits from the general public, an organi- "Well-designed and well-delivered deposit services zation must have a license. MFIs that are licensed as can simultaneously benefit households, enterprises, deposit-taking institutions are generally subject to some groups, the participating financial institutions, and form of regulation and supervision by the country's the government. Good savings programs can con- superintendent of banks, central bank, or other govern- tribute to local, regional, and national economic ment department or entity. Adhering to regulatory and development and can help improve equity." supervisory requirements usually imposes additional costs (Robinson 1994, 35) on the MFI (such as reserve requirements). External regulations should be based on international Ultimately, to be licensed to accept deposits an MFI banking standards and more specifically on international must have the financial strength and institutional capaci- accounting principles, minimum capital requirements, ty to do so. Determining the capacity of MFIs is the techniques to reduce and diversify the risk exposure of responsibility of the licensing body. assets, provision policies, and performance criteria (GTZ 1997a). RESERVE REQUIREMENTS. Reserve requirements refer to MFIs that are not operating in enabling environments a percentage of deposits accepted by an institution must lobby the appropriate authorities to make the nec- that must be held in the central bank or in a similar essary changes. To do this they must familiarize them- safe and liquid form. They are imposed by the local selves with international and local experience and superintendent of banks on formalized MFIs accept- evidence to support their arguments. They should also ing deposits to ensure that depositors' funds are both conduct market research on the demand for savings safe and accessible. By mandating a certain percentage instruments in their own markets (CGAP 1997). of deposit funds as a reserve, governments restrict 158 MICROFINANCE HANDBOOK Box 6.2 Savings Mobilization at Bank Rakyat Indonesia IN JUNE 1983 THE INDONESIAN GOVERNMENT DEREGULATED urban, for a liquid savings instrument in which both finan- the financial sector to allow government banks to set their cial stocks and savings from income flows could be safely own interest rates on most loans and deposits. At that time deposited. Also, many lower-income people wanted to con- TABANAS, the government's national savings program, had vert some of their nonfinancial savings into financial savings, been offered in Bank Rakyat Indonesia's more than 3,600 if these could be deposited in liquid accounts. There was also local-level banks for about a decade but had mobilized only demand for fixed deposit accounts. $17.6 million. State banks had been required to lend at 12 As of December 31, 1995, Bank Rakyat Indonesia's pri- percent and to pay 15 percent on most deposits, providing a mary savings product, SIMPEDES, and its urban counter- negative incentive for deposit mobilization. In addition, part, SIMASKOT, accounted for 77 percent of the total TABANAS, the only instrument available, limited with- deposits in the bank's local banking system. The banking drawals to twice a month, a provision that was unacceptable system that mobilized $17.6 million in its first decade mobi- to most villagers, who wanted to be sure that their savings lized $2.7 billion in its second decade--from the same cus- would be accessible whenever they were needed. tomers in the same banks! This spectacular change occurred But the low savings total of the local banking system was primarily because after 1983 the bank had an incentive to widely attributed not to its real causes--primarily the inter- mobilize savings and began to learn about its local markets. est rate regulations and the poorly designed TABANAS Deposit instruments were then designed specifically to meet instrument--but to untested assumptions that villagers did different types of local demand. not save and that even if there were some who did, they With 14.5 million savings accounts as of December 31, would not save in banks. Both assumptions were wrong. 1995, Bank Rakyat Indonesia's local banking system provides Extensive field research in the 1980s showed--much to the savings services to about 30 percent of Indonesia's households. surprise of most government and Bank Rakyat Indonesia Households and enterprises also benefit from the expanded officials--that there was vast unmet demand, both rural and volume of institutional credit financed by the savings program. Source: Robinson 1995. MFIs with respect to the total amount of funds avail- Canada, or any of its subsidiaries (Morris, Power, and able to them for on-lending. Varma 1986). Reserve requirements result in increased costs to Reserve requirements in Indonesia are 5 percent, in MFIs, because the amount held as reserve cannot be lent the Philippines 8 percent, and in Colombia 10 percent. out and thus does not earn the effective portfolio yield or Some government banks, such as those in Thailand, are investment return rate. exempt from reserve requirements, which gives them a Reserve requirements are often separated between pri- considerable advantage over other institutions. mary reserves and secondary reserves. For example, in Canada all chartered banks are required to maintain a Deposit Insurance reserve with the central bank equal to the aggregate of 10 percent of Canadian currency demand deposits, 2 per- The best way to ensure the safety of deposits is to pre- cent of Canadian currency notice deposits, 1 percent of vent the failure of the MFIs providing the deposit ser- Canadian currency notice deposits in excess of vices. This is best done by ensuring that MFIs have CDN$500 million, and 3 percent of foreign currency sound management and prudent lending policies and deposits held in Canada by Canadian residents. This pri- are supported by adequate regulation and supervision. mary reserve is not interest bearing to the chartered bank. However, most industrial countries (and many develop- Secondary reserves are bank reserves, which must be ing countries) have established deposit insurance plans maintained in prescribed interest-bearing public debt or as a safety net to protect depositors should an institution day loans to investment dealers in Canada, or cash. not be prudently managed. (Informal financial institu- These reserves may not be greater than 12 percent of tions collecting deposits are generally unable to partici- deposits of Canadian residents of the bank, its offices in pate in deposit insurance schemes.) Deposit insurance DESIGNING SAVINGS PRODUCTS 159 guarantees a maximum nominal value of deposits to each saver (based on the amount they have deposited) if Box 6.3 Deposit Insurance in India the institution collecting deposits fails. This insurance is DEPOSIT INSURANCE IN INDIA IS IMPLEMENTED BY A PUBLIC generally government controlled and provided by an sector agency called the Deposit Insurance and Credit apex institution, to which each regulated deposit-taking Guarantee Corporation, which was set up in 1962 in the institution pays a fee. In some countries the government wake of the collapse of a private bank. The agency offers also contributes funds. In all cases it is important that deposit and credit insurance to the entire banking sector in the government provide the necessary backup support to India, including cooperatives and rural banks. While credit make the system effective (box 6.3). insurance is voluntary, deposit insurance is compulsory for Deposit insurance is established for three main rea- the banks. The Deposit Insurance and Credit Guarantee sons (FAO 1995): Corporation charges a premium of 5 paise per year for 100 I It strengthens confidence in the banking system and rupees on a half-yearly basis (0.0005 percent). Each depos- itor is eligible for a maximum level of protection of 1 lakh helps to promote deposit mobilization. (approximately US$2,565), which means practically all of I It provides the government with a formal mechanism the small savers are protected in India. for dealing with failing banks. The scheme seems to have created a significant level of I It ensures that small depositors are protected in the confidence among small savers, which has enabled banks event of a bank failure. to mobilize savings fairly easily, even in remote areas. Two of the main disadvantages of deposit insurance While banks frequently submit claims on credit insur- plans are that savers have less incentive to select stable ance, very rarely do they invoke deposit insurance, given MFIs and MFIs have less incentive to manage their orga- the general financial stability in India. nizations prudently. However, the need to protect those Source: Shylendra 1998. who deposit in MFIs is particularly important, due to the imperfect information about the MFI available to low- income clients and the lack of interest of governments in Although deposit insurance plans should include all intervening and compensating savers when an MFI fails, deposit-taking institutions, including cooperatives and since in most cases its failure is not likely to be seen as a semiformal MFIs, it may not be appropriate to provide threat to the financial stability of a country. the same insurance for all institutions. Having separate Box 6.4 Security of Deposits in the Bank for Agriculture and Agricultural Cooperatives, Thailand T HE B ANK FOR A GRICULTURE AND A GRICULTURAL Thailand. Unfortunately, this implicit guarantee approach Cooperatives (BAAC) has operated in the rural areas of may not be very conducive to depositor peace of mind, and Thailand for 30 years. Originally conceived of as a vehicle for it may create perverse managerial incentives for bankers if delivering low-cost credit to Thai farmers, in recent years the implicit "safety net" encourages them to pursue overly BAAC has come to be recognized as one of the few sector-spe- risky lending operations. As a result, BAAC portfolio quali- cialized government-owned rural financial institutions. ty is poor. One of the important aspects of this success is the bank's Further financial market distortions are created when the impressive record in the past decade in mobilizing financial public believes that such implicit government guarantees are resources. It has done this through vigorous efforts to attract operated in an uneven manner between state-owned and pri- savings in rural areas, which finance its rapidly expanding vately owned financial institutions. The appropriate explicit agricultural lending portfolio. deposit guarantee framework is a risk-based deposit insurance There is no formal deposit insurance program in the program in which each participating financial institution Thai banking sector. However, on a case by case basis, the (whether state-owned or privately owned) contributes an government has saved depositors from losing their funds by amount to the deposit insurance fund based on an indepen- mounting rescue operations overseen by the Bank of dent assessment of that institution's portfolio risk. Source: GTZ 1997b. 160 MICROFINANCE HANDBOOK deposit insurance funds for each major type of financial it programs exist and government intervention in pricing intermediary diversifies risk. For example, banks and and customer selection prevails. MFIs must be allowed credit union insurance funds were not affected by the to operate free of political interference. They must not savings and loan crisis in the United States, and thus be discouraged from mobilizing savings by frequent insurance premiums had to be raised only for savings injections of inexpensive public funds. Furthermore, if and loans associations (FAO 1995). external regulation and supervision is less stringent for Even without the availability of deposit insurance, public than private MFIs, insufficient risk management what seems to be most important to the saver is the rep- may put depositor funds at risk (box 6.5). utation of the bank. Potential savers look for an estab- Privately owned MFIs need to ensure that they both lished bank with a good reputation, appropriate are and are perceived to be sound financial intermedi- instruments, and a convenient local outlet. aries. Relationships with well-respected individuals, fam- ilies, or institutions such as a religious organization or larger holding company can help strengthen the confi- Does the MFI Have the Necessary dence and trust of their depositors. Furthermore, MFIs Institutional Capacity to Mobilize Savings? that were traditionally focused only on providing credit services need to ensure that their identity is reestablished Prior to offering voluntary savings services, MFIs must as a safe place for deposits. This will depend to a large ensure that they have the institutional structure that extent on the quality of the credit services that they pro- allows them to mobilize savings legally and adequate vide and their reputation as serious lenders who do not institutional capacity or the ability to develop it. 1 accept clients who default on their loans. Institutional capacity requires that adequate governance, The governance structure of MFIs is crucial for ensur- management, staff, and operational structures are in ing appropriate financial intermediation services between place to provide savings services. Furthermore, additional savers and borrowers. MFIs that mobilize savings are (and sometimes substantial) reporting requirements to likely to have a more professional governance structure, the superintendent or other regulatory body imply with greater representation from the private financial sec- increased personnel costs and expanded management tor or the membership than those whose sole business is information systems. disbursing loans. Ownership, governance, and organizational structures have a strong impact on client perceptions of MFIs' Organizational Structure mobilization of savings. Moreover, the introduction of voluntary savings services implies the addition of many Most MFIs that mobilize savings have organizational new customers, which in turn requires increased capacity structures that are both extensive and decentralized. A of staff and staff training programs, management, mar- location close to deposit customers reduces transaction keting systems, infrastructure, security and internal con- costs for both the MFI and its clients and is an impor- trols, management information systems, and risk tant part of establishing a permanent relationship built management. on mutual trust, which is a key to successful savings mobilization. Ownership and Governance Successful MFIs organize their branches or field offices as profit centers and employ a method of transfer pricing Public ownership can make an MFI seem reliable and that ensures full-cost coverage throughout the branch net- secure, particularly if strong political support exists and work. Transfer pricing refers to the pricing of services political interference is minimal. Depositors know that provided by the head office to the branches on a cost- the government will protect them in the case of a severe recovery basis. For example, costs incurred in the head liquidity or solvency crisis. However, public ownership office to manage the overall organization are prorated to can also impose limitations on savings if subsidized cred- the branches based on a percentage of assets (loans) or lia- 1. This section draws substantially from Robinson (1995) and GTZ (1997a). DESIGNING SAVINGS PRODUCTS 161 Box 6.5 Ownership and Governance at the Bank for Agriculture and Agricultural Cooperatives, Thailand T HE B ANK FOR A GRICULTURE AND A GRICULTURAL donor agencies or on behalf of the government (as in the Cooperatives (BAAC) is governed by a board of directors case of mandatory commercial bank deposits). The situation appointed by the council of ministers. The board controls has since changed; BAAC is now responsible for the steward- the policies and business operations of the bank. Its enabling ship and safekeeping of a large volume of resources from its legislation requires that board membership also include rep- new depositor-clientele. resentatives from the prime minister's office, the Ministry of Despite the radical change in BAAC's financial resource Finance, the Ministry of Agriculture and Cooperatives, and base, there has so far not been an adaptation of the member- other government departments. Thus the board is mainly ship of the board of directors to reflect the substantial shift composed of public officials who reflect the interests of the in the array of the institution's stakeholders. Nor have the government but lack financial enterprise acumen and experi- opportunities for ownership of the institution been modified ence. The composition of the board reflects the 99.7 percent or broadened significantly to respond to these new interests. government ownership of the bank. A change in the governance structure of the bank could This governance structure was inherited from earlier make an important contribution to addressing some of the years when the institution acted primarily as a dispenser of stakeholders issues that fall outside the original state-owned financial resources received from the government and foreign enterprise design concept. Source: GTZ 1997b. bilities (deposits) held at the branch. Funding costs are institution and its human resources are considerable. apportioned as well. A branch that disburses a larger vol- Managing a financial intermediary is far more complex ume of loans than the deposits it collects needs to receive than managing a credit organization, especially since the funding from the head office (or another branch) to fund size of the organization often increases rapidly. This has those loans. The head office (or a regional office) will act serious consequences if the MFI is like many that have as a central funding facility to ensure that any excess one or two key individuals who head the organization deposits in one branch are "sold" to another branch to and are personally involved in almost every aspect of fund their loans. The branch that has excess deposits operations. In particular, MFIs mobilizing deposits will receives payment (interest revenue) for those funds, while be subject to much greater scrutiny and supervision by the branch receiving them pays a fee (interest expense). If government regulators and others. The level of effective the head office determines that there is no excess funding management must be greater than that of an MFI that is within the system, it will then access external funding and providing credit only. on-lend it to its branches for a set price. Staff recruitment should focus on selecting staff from The transfer price charged to branches (or paid to the local region who are familiar with customs and cul- branches for excess deposits) is set either close to the ture and, if applicable, the dialect spoken in the region interbank lending rate or at a rate somewhat higher than (box 6.6). Local staff tend to instill confidence in clients the average cost of funds for the MFI. This provides and make it easier for them to communicate with the incentives for branches to mobilize savings locally rather MFI. Studies have found that a strong customer orienta- than relying on the excess liquidity of other branches tion and service culture attitudes are crucial ingredients within the network. It can also result in some additional for successful savings mobilization (GTZ 1997a). revenue at the head office level to cover overheads. Significant training of both management and staff Transfer pricing ensures transparency and instills becomes imperative when introducing savings. Man- accountability and responsibility in the branches. agers and staff need to learn how local markets operate, how to locate potential savers, and how to design instru- Human Resources ments and services for that market. They also need to understand basic finance and the importance of an ade- Once savings are introduced, the MFI becomes a true quate spread between lending and deposit services. financial intermediary, and the consequences for the Often MFI staff have a social development background 162 MICROFINANCE HANDBOOK ies with each of its savings products. Each month a draw- Box 6.6 Using Local Human Resources in Caisses ing is made from the names of deposit account holders Villageoises d'Epargne et de Crédit Autogérées, Mali and the winners receive a gift such as a television or bicy- ONE OF THE STRENGTHS OF MICROFINANCE INSTITUTIONS cle. Lotteries have proven to be an effective means of that involve local managers chosen by the communities marketing savings services, and they have been adopted themselves is that these managers know the clients well by BancoSol in Bolivia as well. and are trusted by them. This is the case in the caisses vil- Another marketing method employed by Bank lageoises d'épargne et de crédit autogérées (self-managed vil- Rakyat Indonesia is the displaying of posters showing lage savings and credit banks) in Pays Dogon, a remote pictures of the local subbranch, its larger supervising Sahel region of Mali, West Africa. All of the village bank branch, and the bank's impressive looking head office. managers (tellers and internal controllers) are from the vil- This approach is based on the fact that savers tend to lage itself. Initially they were trained in the course of a be concerned about placing their savings in what donor project (KFW), which was fully staffed with local professionals who were all fluent in at least two local appears to be a small local bank. When the bank has a dialects, were well respected by the villagers, and had a good reputation, emphasis on the relationship between genuine interest in making the program a success to help the sub-branch and the larger bank is reassuring to out their own region. clients. Source: Contributed by Cecile Fruman, Sustainable Banking with MFIs can also establish "open door" days when clients the Poor Project, World Bank, 1998. can come in and meet staff as well as physically see the safe where the money is kept. If savers are unfamiliar with banks in general, they may derive comfort from and may not appreciate the fact that low-income clients observing a safe environment. can and do save. While successful microlending relies on The design of special trademarks for savings products the loan officer personally understanding each client, the has also attracted depositors. For example, Thailand's provision of savings services can be quite different, Bank for Agriculture and Agricultural cooperatives has a depending on how those savings are collected and with- savings product called "BAAC's Save to Increase your drawn. The training needs of all management and staff Chances," and Banco Caja Social in Colombia markets a should be assessed and delivered both initially and peri- "Grow Every Day" savings account. While special trade- odically as the MFI grows. marks make it easier for clients to understand the partic- ular design of each savings product, they also help to Marketing distinguish the products from those offered by a com- peting MFI. MFIs providing credit services must select borrowers Ultimately however, word of mouth is the best form whom they trust to repay the loans. When collecting sav- of advertising. To ensure positive word of mouth, the ings, however, it is the customers who must trust the service provided to depositors must be timely, consider- MFI. Deposit clients must be convinced that the MFI ate, and honest. Once again, client demand must ulti- staff is competent and honest. MFIs must actively seek mately be met through appropriate savings product out potential depositors, keeping in mind their target design. A woman living in a rural area of an eastern market. Most important, the MFI must first design its Indonesian island who was asked what she thought savings products appropriately and then publicize its ser- about Bank Rakyat Indonesia opening a bank office vices in locally appropriate ways. there said, "If [Bank Rakyat Indonesia] opens here and During the market research phase, MFIs need to learn provides good service, it will be wonderful for us. We what savings services are demanded by potential clients need a bank." Then she said, "There is little to do here and then incorporate this information into both product in the evenings except talk. If the bank does something design and marketing. For example, to encourage savings good here, you can be sure that we will talk about it for mobilization Bank Rakyat Indonesia determined that years to come!" "Of course," she added, "if the bank savers would like the opportunity to receive gifts when does something bad here, we will talk about that for they open a savings account. The bank now offers lotter- years to come" (Robinson 1995, 14). DESIGNING SAVINGS PRODUCTS 163 Infrastructure separation of duties, audit trails, and the use of fireproof vaults and drop boxes (Calvin 1995). Experts from the As previously mentioned, introducing voluntary savings formal financial sector should be contracted to advise on products greatly alters the organization and management these issues. of an MFI. Depending on the collection method employed, some MFIs may need to develop branches Management Information Systems close to their clients. However, this may not be neces- sary until the MFI reaches a certain volume of business. An appropriate and soundly operating management Alternatively, MFIs can provide "mobile banking" ser- information system is the core of an efficient internal vices, with the MFI staff traveling to the client to collect control and monitoring system. The management infor- savings rather than having the client visit the branch. mation system should be simple, transparent, and objec- Bank Rakyat Indonesia savings officers go weekly to tive. An effective management information system each village and collect savings and provide withdrawals. becomes fundamentally important when savings services Mobile banking works well so long as the MFI is consis- are introduced both for internal management and exter- tent in its scheduled visits, so that clients are assured of nal reporting. their ability to deposit or withdraw savings on specific When designing management information systems to days of the week. manage savings products, it is important to consider accu- Other MFIs enhance their branch structures to make racy, reliability, and speed. There are three major goals for them more convenient for their clients and more user an information system (adapted from Calvin 1995): friendly. For example, BancoSol operates primarily in I Transaction processing urban areas (including secondary cities) where branches I Customer service are relatively accessible to their clients. They modified I Management information. their branches to include teller windows and a comfort- To make strategic decisions about the savings prod- able and inviting environment for savers. BancoSol also ucts, the MFI needs the following information: number hired and trained staff specifically to manage and main- of accounts, value of accounts, and average balances; tain deposit accounts. number of transactions per account; the distribution of Furthermore, convenient and flexible hours of opera- the balances of the accounts; and the daily turnover as a tion that meet the needs of clients helps to make savings percentage of balances (which may vary by season). services more attractive. Over time it is important to maintain a client database that allows the institution to analyze the relationship Security and Internal Controls between credit and savings products and other market details. As an MFI begins to offer voluntary savings services, it It is increasingly important for management informa- needs to consider the issue of security. At a minimum the tion systems to provide the reporting and monitoring MFI needs to purchase a safe to keep cash available for requirements demanded by the supervisory body potential withdrawals and to minimize the risk of rob- responsible for regulating the MFI. Systems must be bery. If the MFI collects savings through mobile bank- designed to ensure that this information is produced ing, it should have two officers working together to regularly and that it also provides information that is minimize the risk of both fraud and robbery. Bank beneficial to the MFI itself. (See chapter 7 for a more Rakyat Indonesia, which uses this system, has found that detailed discussion of management information the community itself ensures that bank officers are not systems.) robbed, knowing that it is their money that is at risk. MFIs need to develop internal controls to address Risk Management and Treasury security risk when providing voluntary savings products. There is extensive knowledge in the formal financial sec- The introduction of deposit-taking greatly complicates tor that addresses risk, including such checks and bal- the management of assets and liabilities. In addition ances as dual custody of cash and combinations, to meeting the reserve requirements stipulated by the 164 MICROFINANCE HANDBOOK regulatory authorities, it is also important to have the Sequencing the Introduction of Savings right amount of cash in the right branches at the right Services times. Poor liquidity management can seriously ham- per the operational capacity of each branch. Most MFIs planning to introduce voluntary savings services MFIs generating numerous small deposits have found must consider the order in which to carry out the vari- that very small savings accounts, even those with an ous steps. Box 6.7 identifies six steps that an MFI should unlimited number of withdrawals, generally experi- follow. ence little turnover and therefore represent a stable liquidity cushion. Publicly owned MFIs often rely on large savings deposits from state-owned institutions or Types of Savings Products for legally enforced deposits from commercial banks, Microentrepreneurs which implies heightened liquidity risk if government regulations change and they lose their large depositors. Deposit instruments must be designed appropriately to MFIs that create a central funding facility (discussed meet local demand. It is crucial for an MFI to conduct above under transfer pricing) may be better able to market research prior to introducing savings products effectively manage their liquidity risk. (Delinquency, and to offer a good mix of products with various levels of liquidity, and asset and liability management are dis- liquidity that respond to the characteristics and financial cussed in chapter 10.) needs of various market segments. Box 6.7 The Sequencing of Voluntary Savings Mobilization 1. Enhance the knowledge of the MFI's board and managers nearby the bank offices; this is known as "the easy money." about the experience of other local and international Market penetration of the wider service area, however, MFIs. By understanding the savings services that exist requires other methods. These include the development of locally, MFI managers can build on existing systems. a systematic approach to identification of potential deposi- 2. Carry out market research and train staff selected for the tors; the implementation of a staff incentive system based pilot stage. Examine the regulatory environment and the on the value of deposits collected so that the staff will seek institutional capacity of the MFI. out the potential depositors rather than wait for them to 3. Conduct and evaluate a pilot project. This is a crucial step, come to the bank; the development of effective methods because until the extent of demand and the costs of the for intrabank communication; more extensive market different products, including labor, are known, only tem- research; a major overhaul of public relations; and massive porary interest rates can be set. staff training. 4. When necessary, carry out and evaluate a second set of The sequencing outlined here may appear lengthy and pilot projects testing products and prices that were revised cumbersome, but it is necessary for building the long-term as a result of the first pilot. At the same time, hold wider viability of the MFI. There are often temptations to go too staff training. During this period pay attention to planning fast, but instituting voluntary savings mobilization too rapidly the logistics and management information systems that is a prime example of "haste makes waste." For example, an will be required for the expansion of the program of MFI is asking for trouble if it tries to implement its new sav- deposit mobilization. ings program before its management information system is 5. When the instruments, pricing, logistics, information sys- ready and the staff is well trained in its use. A financial institu- tems, and staff training are completed, gradually expand tion that gets its sequencing wrong can lose its clients' trust, savings mobilization throughout all branches. which in turn can lead to the loss of its reputation and eventu- 6. When expansion of saving services to all branches is ally of its viability. achieved, switch attention from the logistics of expansion Getting the "when" and "how" of introducing voluntary to the techniques of market penetration. When well-run savings mobilization right enables MFIs to meet local demand MFIs offer appropriate deposit facilities and services, they for savings services and provide a larger volume of microcredit, can quickly gain the accounts of people living or working thereby increasing both outreach and profitability. Source: Robinson 1995. DESIGNING SAVINGS PRODUCTS 165 When Bank Rakyat Indonesia decided to offer volun- ture are offered, including at least one liquid savings tary savings products, it conducted an extensive study of product with unlimited withdrawals. local demand for financial services and systematically I Attractive rates of interest are offered. For private identified potential savers. The bank also looked closely MFIs this may mean rates of interest that are higher at existing savings services in the informal sector and than market rates as a sort of risk premium to attract designed its products to overcome limitations and to deposits, particularly if they are operating in a com- replicate the strengths of existing services (box 6.8). petitive environment. Public MFIs may be able to pay For the most part, individual voluntary savings have lower rates of interest based on the perceived safety of been found to be more successful than group savings. public institutions. Most savings products for microclients include the fol- I Interest rates should increase with the size of the sav- lowing features (GTZ 1997a): ings account to provide a financial incentive for savers I The required minimum opening balances are general- to increase deposits and refrain from withdrawing. ly low. I Conversely, savings account balances below a speci- I A mix of liquid savings products, semiliquid savings fied minimum are exempt from interest payments to products, and time deposits with a fixed-term struc- partially compensate for the relatively higher adminis- trative costs of small accounts. I Some MFIs may charge fees and commissions to Box 6.8 Bank Rakyat Indonesia Savings Instruments depositors for opening and closing accounts and for specific services related to account management, such AFTER A SERIES OF PILOT PROJECTS THAT BEGAN IN 1984, Bank Rakyat Indonesia introduced four savings instru- as issuing a passbook. ments with different ratios of liquidity and returns There are three broad deposit groups based on the nationwide in 1986. SIMPEDES, a savings instrument degree of liquidity: highly liquid current accounts, semi- that permits an unlimited number of withdrawals, liquid savings accounts, and fixed-term deposits. became the "flagship" of the new savings program. The SIMPEDES instrument incorporates lotteries for the Liquid Accounts depositors that were modeled loosely on Bank Dagang Bali's highly successful depositor lotteries. SIMPEDES Highly liquid deposits provide the greatest flexibility and was aimed at households, firms, and organizations that liquidity and the lowest returns. Current accounts--or demand liquidity in combination with positive real demand deposits--are deposits that allow funds to be returns. TABANAS, which provides a higher interest rate deposited and withdrawn at any time. Frequently no inter- than SIMPEDES, was continued, with the withdrawal est is paid. Highly liquid accounts are difficult to manage, rules later liberalized. TABANAS was aimed at depositors who wanted middle levels of both liquidity and returns. because they require substantial bookkeeping and are not Deposito Berjangka, a fixed deposit instrument previ- as stable a source of funding as term deposits. Only a lim- ously available from Bank Rakyat Indonesia only through ited portion of demand deposits can be used to provide its branches, was now offered as well through the bank unit loans (based on reserve requirements), because the MFI in its local banking system. Deposito Berjangka is used by must be able to service withdrawal requests at all times. wealthier villagers and firms hoping to realize higher returns and also by those saving for long-term goals, such as build- Semiliquid Accounts ing construction, land purchase, and children's education. Many Deposito Berjangka account holders also hold SIM- Semiliquid accounts provide some liquidity and some PEDES accounts. The fourth instrument, Giro, a type of returns. Some savings accounts are semiliquid, which current account, is primarily for use by institutions that means that the borrower can usually withdraw funds a must meet special government requirements. With the exception of Giro, these deposit instruments have generally limited number of times per month and deposit funds at provided positive real interest rates. any time. Unlike current accounts, savings accounts gen- erally pay a nominal rate of interest, which is sometimes Source: Robinson 1995. based on the minimum balance in the account over a given period (monthly, yearly). They begin to act like 166 MICROFINANCE HANDBOOK term deposits if interest is paid only when a minimum balance is held in the account. This encourages the client Box 6.9 The Choice of Savings Products in Pays Dogon, to hold a certain amount of money in the account, which Mali makes for easier management by the MFI. IN PAYS DOGON, A NETWORK OF 55 SELF-RELIANT AND self-managed village banks has been established. The rural Fixed-Term Deposits communities have themselves determined the financial products that they wish their banks to offer. Both demand Term deposits are savings accounts that are locked in for a and term deposits are offered. The results are that term specified amount of time. They provide the lowest liquid- deposits are much more popular (constituting about 85 percent of total deposits) due to liquidity management, ity and the highest returns. Term deposits are generally a which dictates that only term deposits are used for lending stable source of funding for the MFI and pay a higher rate (out of community solidarity those who are wealthy of return to the saver. Generally, the interest rate is based enough to save prefer to save long-term so that other peo- on the length of the term of the deposit and the expected ple from the village can access loans); interest, which on movement in market rates. Term deposits range from a term deposits is high (15 to 20 percent annually); and the one-month term to several years; they allow the MFI to tradition of Dogon farmers of long-term savings in kind fund loans for a period of time just short of the deposit (before the severe draughts hit the Sahel in the 1980s, term. This makes liquidity and gap management easier they were all cattle owners, but with the loss of a majority for the MFI, because the funds are available for a set peri- of their herds during dry periods, these farmers are now od of time, which is offset by reduced liquidity and happy to have alternative monetary savings products that increased interest-rate risk for the saver. (Gap manage- offer much higher security). ment and interest-rate risk are discussed in chapter 10.) Source: Contributed by Cecile Fruman, Sustainable Banking with The choice of products to offer must be based on the the Poor Project, World Bank. needs of the clients in the area served by the MFI. Many first-time savers will choose a highly liquid account. As they become more experienced with managing their material for the initial launch; safes; computer systems, savings, they will likely open more than one deposit including hardware and software; as well as existing or new account, using a semiliquid or fixed-term deposit staff costs. These costs must all be estimated and a plan account for longer-term savings (box 6.9). devised prior to developing the savings products. Direct costs are those costs incurred specifically in the provision of savings services. Direct costs can be variable Costs of Mobilizing Voluntary Savings or fixed. Variable costs are those that are incurred per account or per transaction. They include time and mate- The cost of savings mobilization depends not only on rials used in opening, maintaining, and closing an internal factors such as operational efficiency, but also on account. These costs can be estimated per account based external factors such as minimum reserve requirements on average activity and the time taken to process the (discussed above), tax rates, and general market condi- transactions in a given time period, such as a month. tions. Determining both internal and external costs helps Fixed direct costs are those incurred to deliver savings to establish what rate to pay on different savings products products that do not vary with the number of accounts (taking into account the market rate for deposits). (units) opened or maintained. These include salaries and Costs include (adapted from Christen 1997): ongoing training for savings officers, additional manage- I Setup costs ment, additional accounting personnel, branch infrastruc- I Direct costs ture, and any special promotions for savings products. I Indirect costs Indirect costs are costs that do not relate directly to the I Cost of funds (paid to depositors). provision of savings services but a portion of which Setup costs include research and development, which should be charged to the savings operation by virtue of may include hiring outside consultants and savings being part of the overall services provided by the organi- experts; the printing of passbooks and other marketing zation. These include overhead (such as management), DESIGNING SAVINGS PRODUCTS 167 premises, general operating costs, as well as other head office and branch costs. Indirect costs are generally pro- Box 6.10 Administrative Costs for Banco Caja Social, rated based on the amount of business activity generated Colombia by savings products as opposed to other business activi- BANCO CAJA SOCIAL RECENTLY BEGAN COST ACCOUNTING ties (such as loans, investments, training, and so forth). by savings product and found that the administrative costs These costs can often be estimated by examining for an actively used traditional passbook savings account existing costs as well as in reference to other organiza- below US$10 was roughly 18 percent of the balance (with- tions providing savings services. These will vary depend- out considering interest payments during the year). ing on the delivery method chosen for providing savings However, it also found that for the average savings account in 1996, administrative costs were below 1 percent of the services. balance. This demonstrates that while the administration of The cost of funds refers to the interest rate paid to very small savings is costly, financial institutions can mini- depositors. The rates vary based on the liquidity of the mize those costs by offering a good mix of savings products account and the amount of time a deposit is held. and implementing cost-reducing procedures. All of these cost factors must be considered when Source: GTZ 1997c. offering savings products. However, it is difficult to esti- mate accurately the operational costs and demand for each product and the interest rates necessary both to make the instrument attractive and the spread profitable. tions, the rate of inflation, and market supply and Pilot projects are thus imperative to estimate the costs demand. Risk factors such as liquidity risk and interest- accurately, to set appropriate interest rates, to establish rate risk must also be considered based on the time peri- suitable spreads, and to ensure that the product is od of deposits. Finally, the costs of providing voluntary designed to meet client needs. Furthermore, the incen- savings also influences deposit pricing policies. tive system of MFIs can be designed to increase staff pro- MFIs incur greater operational costs to administer ductivity and in turn reduce administrative costs. (For a highly liquid accounts and thus the interest rates are more complete treatment of the costs of savings lower. Fixed-term deposits are priced at a higher rate, accounts, see Christen 1997, 201­11.) because the funds are locked in and are not available to Furthermore, transaction costs incurred by deposit the depositor. Thus they are of lower risk to the MFI, customers must be considered. Locating branches in and as a result term deposits are a more stable source of densely populated areas or close to business centers where funding than demand deposits (ignoring for the moment a significant number of people meet to carry out econom- asset-liability management and term matching issues for ic transactions reduces transaction costs for customers. the MFI and liquidity risk and interest-rate risk for the Also, MFIs should minimize the waiting time for clients depositor, which are discussed in chapter 10). to make deposits by ensuring that enough tellers are pro- Savings products need to be priced so that the MFI vided, based on the number of clients and transaction vol- earns a spread between its savings and lending services umes. Clients who are only able to visit the MFI branch that enables it to become profitable. Labor and other during evenings or weekends should be accommodated. nonfinancial costs must be carefully considered when set- While these measures reduce transaction costs for clients, ting deposit rates. However, there are a number of the administrative costs for the MFI increase. A cost-ben- unknowns at the beginning of savings mobilization. For efit analysis should be carried out to determine the best example, a highly liquid savings account that is in high way to both lower transaction costs for clients and mini- demand can be quite labor intensive and thus costly, mize the administrative costs for the MFI (box 6.10). especially if there is a large number of very small accounts. Experience to date indicates that most savers who select a liquid account are not highly sensitive to Pricing Savings Products interest rates and prefer better service to higher interest. A study conducted by Ostry and Reinhart (1995) finds The interest rate paid on deposits is based on the prevail- that the lower the income group, the less sensitive they ing deposit rates of similar products in similar institu- are to interest rates paid on savings: 168 MICROFINANCE HANDBOOK Colombia." CGAP (Consultative Group to Assist the "A one percentage point rise in the real rate of Poorest) Working Group, Washington, D.C. interest should elicit a rise in the saving rate of ------. 1997d. "Comparative Analysis of Savings Mobilization only about two-tenths of one percentage point for Strategies--Case Study for Bank Rakyat Indonesia (BRI), the 10 poorest countries in the sample. For the Indonesia." CGAP (Consultative Group to Assist the wealthiest countries, by contrast, the rise in the Poorest) Working Group, Washington, D.C. saving rate in response to a similar change in the ------. 1997e. "Comparative Analysis of Savings Mobilization real interest rate is about two-thirds of a percentage Strategies--Case Study for Rural Bank of Panabo (RBP), point." (Ostry and Reinhart 1995, 18). Philippines." CGAP (Consultative Group to Assist the Poorest) Working Group, Washington, D.C. Morris, D.D., A. Power, and D.K. Varma, eds. 1986. 1986 Sources and Further Reading Guide to Financial Reporting for Canadian Banks. Toronto, Canada: Touche Ross. Calvin, Barbara. 1995. "Operational Implications of Introducing Ostry, Jonathan D., and Carmen M. Reinhart. 1995. "Saving Savings Services." Paper delivered at the MicroFinance and the Real Interest Rate in Developing Countries." Network Conference, November 8, Cavite, Philippines. Finance & Development 32 (December): 16­18. Christen, Robert Peck. 1997. Banking Services for the Poor: Patten, H. Richard, and Jay K. Rosengard. 1991. Progress with Managing for Financial Success. Washington, D.C.: Profits. The Development of Rural Banking in Indonesia. San ACCION International. Francisco: ICS Press. CGAP (Consultative Group to Assist the Poorest). 1997. Robinson, Marguerite. 1994. "Savings Mobilization and "Introducing Savings in Microcredit Institutions: When Microenterprise Finance: The Indonesian Experience." In and How?" Focus Note 8. World Bank, Washington, D.C. Otero and Rhyne, eds., New World of Microenterprise FAO (Food and Agriculture Organization). 1995. Finance. West Hartford, Conn.: Kumarian Press. "Safeguarding Deposits: Learning from Experience." FAO ------. 1995. "Introducing Savings Mobilization in Agricultural Services Bulletin 116. Rome. Microfinance Programs: When and How?" Paper delivered Fruman, Cecile. 1998. "Pays Dogon, Mali" Case Study for the at the annual meeting of the MicroFinance Network, Sustainable Banking with the Poor Project, World Bank, November 8, Cavite, Philippines. Washington, D.C. Rutherford, Stuart. 1996. "Comment on DevFinance GTZ (Deutsche Gesellschaft für Technische Zusammenarbeit). Network, September 6, 1996." Internet discussion net- 1997a. "Comparative Analysis of Savings Mobilization work, Ohio State University. Devfinance@lists.acs.ohio- Strategies." CGAP (Consultative Group to Assist the state.edu Poorest) Working Group, Washington, D.C. Shylendra, H.S. 1998. "Comment on DevFinance Network, ------. 1997b. "Comparative Analysis of Savings Mobilization March 12." Ohio State University. Devfinance@ Strategies--Case Study for Bank of Agriculture and lists.acs.ohio-state.edu Agricultural Cooperatives (BAAC), Thailand." CGAP Wisniwski, Sylvia. 1997. "Savings in the Context of (Consultative Group to Assist the Poorest) Working Group, Microfinance: Comparative Analysis of Asian Case Studies." Washington, D.C. Paper delivered at the Fourth Asia-Pacific Regional Work- ------. 1997c. "Comparative Analysis of Savings Mobilization shop on Sustainable Banking with the Poor, November Strategies--Case Study for Banco Caja Social (BCS), 3­7, Bangkok. CHAPTER SEVEN Management Information Systems T he management information system of an insti- cluding the reworking of staff responsibilities (some- tution includes all the systems used for generat- times even of some staff qualifications), the redesign of ing the information that guides management in work processes and information flows, the revision and its decisions and actions.1 The management informa- rationalization of financial policies, and significant tion system can be seen as a map of the activities that investment in computer technology. However, the are carried out by the MFI. It monitors the operations benefits associated with these costs are substantial. of the institution and provides reports that reflect the Good information is essential for the MFI to perform information that management considers the most signif- in an efficient and effective manner. The better the icant to track. Staff, management, board, funding orga- information, the better the MFI can manage its nizations, regulators, and others rely on reports resources. produced by the management information system to Good information systems can: give an accurate portrait of what is occurring in the I Improve the work of field staff, enabling them to institution. better monitor their portfolios and provide better With the current trend in the microfinance commu- services to an increasing number of clients nity toward the significant scale-up of activities, man- I Enable supervisors to better monitor their areas of agers of many MFIs are becoming increasingly aware of responsibility, pinpointing priority areas that most the acute need to improve their information systems. require attention Methodological issues, staff development, and even I Help senior management to better orchestrate the financing are frequently not proving to be the critical work of the entire organization and make well- constraints to growth. Rather, an institution's ability to informed operational and strategic decisions by track the status of its portfolio in a timely and accurate regularly monitoring the health of the institution manner is often the most pressing need. The reliability through a set of well-chosen reports and indica- of the systems tracking this information is in many cases tors. the difference between success and failure of the lending This chapter will interest practitioners who are in the and savings operations and, therefore, of the institution. process of setting up, evaluating, or improving their More accurate, timely, and comprehensive information management information systems. In particular, MFIs on operations, especially on the loan portfolio, will that are transforming their institutional structure or strengthen the management's capacity to enhance finan- adding new products will find this chapter of interest. cial performance and expand client outreach. Donors may also wish to understand the characteristics Setting up good information systems may necessi- of a good management information system for the MFIs tate a significant restructuring of the institution, in- that they are supporting. 1. This chapter was written by Tony Sheldon. It draws substantially on Waterfield and Ramsing (1998) and Waterfield and Sheldon (1997). 169 170 MICROFINANCE HANDBOOK An Overview of Issues Related to Management Information Systems Box 7.1 Determining the Information Needs of an MFI At the most fundamental level, it is important to note IN ORDER TO DETERMINE THE INFORMATION NEEDS OF AN the distinction between "data" and "information." MFI, it is necessary to identify the users of the informa- Relevant data is generated by an institution in many tion and evaluate the needs of each user group. I What key information do the users need? forms: checks made payable to employees, suppliers, and I What key indicators or ratios do the users need to customers; client loan repayment records; and bank withdrawals, deposits, and transfers. It is the task of a monitor to perform their job well? I What additional information should the users have to management information system to transform this "raw be knowledgeable about the organization's perfor- data" (or input) into meaningful information (or output) mance and broader goal achievement? that can be used by the MFI in its decisionmaking. I How might the users' information needs change in the Some MFI management information systems are man- future and how might those changes affect the design ual, particularly those of MFIs just starting out. However, of the management information system ? as MFIs expand, many of them develop computer-based Source: Author's compilation. management information systems requiring software pro- grams that are designed to capture and report on the rele- vant information. Many of the problems in developing ing information needs is to evaluate the needs of the and implementing computer-based information systems users of that information. An integral part of this process arise from the complexity of the information to be cap- is a consideration of how information is to be presented tured and reported. The most prevalent cause of poorly and what frequency and timeliness are required. In the performing management information systems is a lack of system design stage (discussed more fully below), a focus clarity on the part of users and systems designers about on the specific needs of each user group, including a precisely what needs to be tracked and reported. careful mapping of the flow of information through the Many MFIs have spent substantial time and money institution, is essential to avoid the problem of poor developing information systems that turn out to be dis- identification of information needs. appointing or unsatisfactory due to the following three Even when information needs are clearly understood factors: and defined by users, MFI managers typically do not I Poor identification of information needs speak with the same vocabulary as information systems I Poor communication between management and sys- designers. The language of MFI management (for exam- tems personnel ple, cost centers, portfolio aging, flat amount or declining I Unrealistic expectations about information technology. balance interest rate calculations) is not necessarily under- Frequently, managers, field staff, board members, and stood in depth by systems designers. Likewise, the lan- information systems staff do not have a good sense of the guage of systems designers (databases, operating systems, complete information needs of their institution. They network platforms) is not readily understood by nontech- may be aware of many of the required reports (such as an nical managers. It is crucial that the various users of the income statement, balance sheet, or portfolio aging) and management information system not only identify their several key indicators that need to be tracked, but the information needs but also clearly articulate them to the precise data inputs and information outputs may be systems designers. The translation of user needs into the insufficiently defined. If the specifications for a manage- kind of specifications required to develop or maintain a ment information system are poorly conceived, the sys- management information system is comparable to a rigor- tem will never succeed in transforming data inputs into ous and precise translation of a text from one language to useful management reports. another. The starting point in the development of any man- Finally, unrealistic expectations about information agement information system is to determine what infor- technology by potential users also causes frustration and mation the institution needs to make appropriate disappointment. A computerized management informa- decisions and perform well (box 7.1). The key to defin- tion system can serve many essential functions, but it will MANAGEMENT INFORMATION SYSTEMS 171 not function any better than the policies and procedures underlying logic of the accounting side of management that it tracks and reports. Introducing a management information systems. Frequently, these standards are information system into an MFI will not solve problems developed by a central authority (such as the Financial that lie in the underlying structure and work flow of the Accounting Standards Board in the United States) and institution. As mentioned earlier, good information are applied to institutions that fall within a particular comes at a price; an MFI must be willing to pay that area of the tax code (such as nonprofit or nongovern- price (for ongoing data entry and maintenance as well as mental organizations [NGOs]). General ledger software up-front development and purchase) if it intends to reap programs incorporate and reflect these accounting stan- the benefits of a good management information system. dards and conventions. Therefore, finding an existing With good information provided in a useful form and accounting program that performs at least the basic on a timely basis, all the stakeholders in the institution-- functions required of a microfinance institution and staff at all levels, clients, board members, donors, provides the essential accounting reports (income state- investors, regulators--can be provided with the informa- ment, balance sheet, and cash flow) is a relatively easy tion they need to participate productively in the activities task. Management should clearly define the kinds of of the MFI. variations on these basic reports that it will need to over- see operations effectively, such as income statements by branches or balance sheets by funding organizations. Three Areas of Management Information The general ledger can then be designed to capture the Systems relevant data and generate reports at the appropriate level of detail. MFI management information systems generally fall into three main areas: C HART OF ACCOUNTS . The core of an institution's I An accounting system, with the general ledger at its accounting system is its general ledger. The skeleton of core the general ledger is, in turn, the chart of accounts. The I A credit and savings monitoring system, which cap- design of the chart of accounts reflects a number of fun- tures information and provides reports about the per- damental decisions by the institution. The structure and formance of each loan disbursed, often with a savings level of detail established will determine the type of infor- system that monitors all transactions related to client mation that management will be able to access and ana- savings lyze in the future. Management must be clear about its I A system designed to gather data on client impact. information needs and be able to reach a balance between Not all institutions have a management information two contrasting considerations. On the one hand, if the system that covers all of these areas. The first two (general chart of accounts captures information at too general a ledger and loan tracking) almost always exist in some level (for example, by not separating interest income from form; however, only MFIs that accept deposits need a sys- fees), the system will not provide the kind of detailed tem to monitor savings. Client impact data is generally information that management needs to make informed gathered only in an informal manner, and the "informa- decisions. On the other hand, if the chart of accounts is tion system" that keeps track of this information is often designed to capture too great a level of detail, the system similarly informal. In this discussion the focus is on the will track needless amounts of data and generate informa- first two types of management information systems, par- tion that is so disaggregated that management cannot ticularly loan-tracking software, which is the most prob- identify and interpret trends properly. In addition, the lematic component of an MFI's information systems. greater the level of detail, the longer and more costly it will be to gather the data and process the information. Accounting Systems Nearly all MFIs make use of key financial indicators to monitor the status of their operations (see chapter 9). Although the standards guiding accounting and auditing These financial indicators are based on information that procedures vary greatly from country to country, there is extracted at least in part from the chart of accounts. are almost always basic principles that determine the Therefore, management should determine the indicators 172 MICROFINANCE HANDBOOK they intend to use and make sure that the chart of widely accepted guidelines for the loan-tracking side of a accounts is structured to support the generation of those management information system. (Here, "loan-tracking indicators. software" and "portfolio system" are used synonymously The charts of accounts should be designed to meet the with "credit and savings monitoring system," including needs of management, providing the information with a the capacity to track savings-related information.) As a degree of detail that is meaningful for managers at all lev- result, each software program designed for loan tracking els. Donor, regulatory, and auditing requirements are has its own approach to the information that is tracked, usually less detailed and, therefore, can be met if manage- the kinds of reports that are generated, and, most impor- ment needs are met. (There are some cases, however, in tant, the kinds of features that are included. Some of the which regulatory bodies will require the use of a specific key features that vary widely among loan-tracking software chart of accounts for institutions under their jurisdiction, programs include the types of lending models supported especially for formal financial intermediaries.) (such as group, individual, or village banking), the meth- In addition to the general ledger, low-cost software is ods of calculating interest and fees, the frequency and generally available for other accounting operations, such composition of loan payments, and the format of reports. as accounts payable, accounts receivable, and payroll. Each MFI tends to have its own idiosyncratic way of struc- Depending on the scale and particular needs of the MFI, turing its credit operations, and so an MFI's loan-tracking those additional functions can be automated in the man- software attempts to reflect the operational procedures and agement information system. work flow of the institution. The following features should be sought in general Because there are no agreed standards for loan-tracking ledger software packages (Women's World Banking systems and the information to be tracked and reported is 1994): relatively complex, institutions face a number of chal- I A system that requires a single input of data to gener- lenges when considering how to improve the loan man- ate various financial reports (for example, those need- agement component of their management information ed by management, auditors, and donors) system. A portfolio system should be designed to work I Software that incorporates rigorous accounting stan- with all major types of financial products that are current- dards (for example, does not accept entries that do ly offered (and are likely to be offered in the future). All not balance; supports accrual accounting) MFIs offer loans, which are the most complex product for I A flexible chart of accounts structure that allows the the system to track. Some also offer savings accounts, time organization to track income and expenses by pro- deposits, checking accounts, shares, credit cards, insurance gram, branch, funding source, and so on policies, or other products. The portfolio system will need I A flexible report writer that allows the organization to to accommodate each of these distinct products. generate reports by program, branch, funding source, The system should be designed to establish distinct and so on sets of "rules" for the different kinds of products within I The capacity to maintain and report on historical and each major type of financial product (loans, savings, and budget, as well as current, financial information--a so forth). For example, an institution may offer working "user-friendly" design that includes clear menu lay- capital loans, fixed asset loans, small business loans, and outs, good documentation, and the capacity to sup- solidarity group loans. Each of these types of loans will port networked operations, if relevant have distinctly different characteristics or sets of rules. I Reasonably priced local support, either by phone or in Interest rates, interest calculation methods, maximum person (particularly during the transition and training allowable amounts and terms, definition of overdue pay- periods) ments, eligible collateral, and many other factors will I Relatively modest hard-disk utilization requirements. vary among the different loan products. Again, contrasting considerations must be balanced. Credit and Savings Monitoring Systems The more financial products the institution offers, the more complex the portfolio system becomes. The more While well-established accounting practices are reflected in complex the system, the higher the cost of purchasing general ledger software, there are currently no standards or (or developing) the software, the higher the level of MANAGEMENT INFORMATION SYSTEMS 173 expertise needed to use and support the system, and the systems can easily exceed US$100,000, the increased higher the risk of programming and data entry errors. A quality of information and management control, as well loan-tracking system should be full-featured enough to as the automation of many key tasks, can lead to signifi- handle the range of existing and anticipated financial cant efficiencies in operations of this scale, which justify products, but not so complex as to become too difficult the high level of investment required. or expensive to develop, use, or maintain. To decide among the alternatives, the MFI should assess whether an existing software package meets a ASSESSING LOAN-TRACKING SOFTWARE. There are three majority of its portfolio management needs. An initial alternative approaches to acquiring loan-tracking software: assessment of software programs should include a careful I Purchase an "off-the-shelf" software package from review of all documentation available and, ideally, a either a local developer or an international supplier review of any demonstration or trial versions of the soft- I Modify an existing system, incorporating key features ware. This initial assessment should focus on major issues required for the specific MFI of compatibility among the types of financial products I Develop a fully customized system, designed and pro- supported, basic interest methods supported, and so on, grammed specifically for the institution. rather than the more technical details, such as procedures A decision on which route to take is based largely on for calculating penalties, because these are sometimes dif- the scale of operations of the institution. MFIs can be ficult to determine from basic documentation. (Appendix categorized usefully, if simplistically, into three sizes: 1 provides a summary of the most well-known software small, medium, and large. packages available.) Small MFIs are those with fewer than 3,000 clients When areas of incompatibility are identified, they that have no plans for significant expansion. These MFIs should be carefully noted for discussion with the software do not expect to convert into formal financial institu- firm. Often incompatibilities can be addressed through tions or to offer a broad range of financial products. undocumented features of the software and can be solved Consequently, their management information system through relatively minor alterations. At other times needs are fairly basic and do not demand a rigorous and seemingly minor incompatibilities can completely rule versatile portfolio system. These institutions need a rela- out the system if the necessary modifications affect its tively simple system that monitors and reports on the core and even extensive programming cannot resolve the quality of the portfolio and can generate the information issue. This level of incompatibility may lead large MFIs required for key management indicators. to decide to develop a brand new system (box 7.2). Medium MFIs are institutions that have from 3,000 to If a particular software package passes an initial assess- 20,000 clients, are expanding, and offer an array of credit ment (or if a decision is made to design a new system), and savings products. These MFIs require a much more the following framework for loan-tracking software can rigorous management information system with solid secu- serve as a guide to a fuller evaluation of a portfolio man- rity features and a thorough audit trail, that handles a large agement system. (Appendix. 2 amplifies this framework volume of transactions, includes the capacity to monitor into several key considerations for each category.) savings accounts, and will stand up to the scrutiny of There are six aspects of loan-tracking software that banking regulators. These MFIs cannot generally afford should be evaluated: the development of a complete, customized loan-tracking I Ease of use system and must, therefore, sacrifice some features and I Features flexibility to make use of an existing package, which may I Reports be modified to meet some specific information needs. I Security Large MFIs exceed or intend to exceed 20,000 clients. I Software and hardware issues These institutions are large enough generally to justify the I Technical support. substantial modification of an existing management infor- mation system--or the development of a new system--to EASE OF USE. A system's value is based in part on how more specifically meet their information needs. Although easy it is for staff to use. Several issues should be consid- the costs of developing such management information ered, including the "user-friendliness" of screen design 174 MICROFINANCE HANDBOOK Box 7.2 Microfinance Institutions with Developed "In-House" Systems THE ALEXANDRIA BUSINESS ASSOCIATION IN EGYPT PROVIDES has been upgraded into a local area network (LAN) version. individual loans to microentrepreneurs. In 1997 it had approx- The association has purchased a new Windows-based system imately 13,000 clients. The association employs the same basic to replace the current DOS-based system very soon. Both software systems that it used when it started its operations the accounting and the loan-tracking systems are manual at seven years ago. It uses a modified off-the-shelf accounting the branch level. However, each of the five regional offices is package and a homemade portfolio system, which are not inte- equipped with a computer that processes accounting and grated. Each month the data between the two are reconciled. portfolio data. In addition, each regional office has a printer The Alexandria Business Association describes its accounting and a backup power source. The information is transferred system as partially manual because the organization is required by diskette to the head office. by Egyptian law to maintain two accounting books. PRODEM is a Bolivian NGO that focuses on serving The association is in the process of decentralizing its oper- rural microentrepreneurs with financial services. In 1997 it ations. Currently, 3 of its 10 branches are decentralized, had 28,000 clients. In 1992 PRODEM established a new which means that there is a management information system data systems department. A systems manager was hired to employee in the branch who captures applications and work with four other department staff members at the receipts. For the other seven branches this function is fulfilled national office. Together the department staff has developed at the head office. Information between decentralized branch- a system completely in-house. es and the head office travels on diskette by courier; however, The system was designed in the environments of DOS the Alexandria Business Association is currently accepting 6.22 and LAN 3.11. The loan portfolio and accounting bids to establish a dedicated leased line. Each branch has one modules are fully integrated at the national level only. computer, one printer, and one uninterupted power supply Information from these modules is exported into Excel to (UPS) backup. produce monthly performance indicators. Association de Crédit et d'Epargne pour la Production Under the current system the information flows between (ACEP) in Senegal provides both individual and solidarity the national office and branches using a variety of systems, group loans. In 1997 it had about 4,000 clients. It has used depending on the location. The most common method is for the same software system since it began operations. For the the loan portfolio to be sent weekly by each branch to the accounting, payroll, and fixed assets management system it regional office. Typically, rural branches do not have access employs a modified off-the-shelf accounting software pack- to management information systems. The information is age called SAARI, which is widely used in France and in consolidated at the regional office with information from all French-speaking countries. For its portfolio-tracking system branches in the region. The regional office sends consolidat- the association employs a modified version of the U.S. ed information (about loan portfolio, accounting, and DataEase software. The accounting and the loan-tracking administration) to the head office monthly via either fax systems are not integrated. For the headquarters, DataEase modem or an internet link. Source: Waterfield 1997. and data entry, the quality of program documentation the software design at the beginning. If an otherwise and of any tutorials, the handling of errors, the availabili- appealing software package is not available in the lan- ty of help screens, and so forth. guage required, the institution should be sure that the design of the software allows for a means of adding more FEATURES. This category includes some of the most languages, rather than necessitating a major revision of important aspects of the software system: user lan- the program. guages, setup options, loan product definitions, lending A robust system provides setup options, which require methodology issues, management of branch office the user to input parameters that serve as the underlying information, overall accounting issues, reports, and structure for the system, including the widths of various security issues. data fields, the structure of client and loan account num- In the development of any software package, support bers, and so on. Many of these items, which may appear for multiple-user languages should be incorporated into straightforward, can have far-ranging implications for MANAGEMENT INFORMATION SYSTEMS 175 the use of the system. For example, changing something REPORTS. When selecting a portfolio system, it is crucial to as seemingly simple as the number of decimal places in review the precise definitions of how information is pre- the currency field may mean that every data screen and sented to determine if the system is suitable for the institu- report format needs to be redesigned. tion's needs. For example, it is not sufficient to know that As mentioned previously, the method of lending is an the system produces an "aging report" without knowing area of extreme importance. There are many different the specifics of how the aging report is structured. lending methods in practice, and a management infor- The type of reports the MFI requires, as well as the mation system may not appropriately handle the format and content of those reports, depends on several approach used in a specific MFI. For example, to know factors, including the lending methodologies used, the that a system handles "solidarity group lending" does not services offered, the staffing structure, and the key indica- mean that it will work for every institution employing tors chosen by management. Users of reports within the solidarity group lending. In some MFIs each group loan MFI can be grouped into three broad categories: field is considered a single loan, whereas in other institutions staff needing day-to-day, detailed operational informa- each client has an "individual" loan while mutually guar- tion; supervisors needing reports that facilitate oversight anteeing the loans of other group members. The chosen of field staff; and senior management and board mem- management information system must support the lend- bers looking for more strategic information. There are ing approaches used by the institution. also external users of reports, including auditors, regula- Most MFIs define the basic parameters of their loan tors, donors, investors, and clients (box 7.3). product, which determine interest rate methods, penalty Key issues in report design include: calculation methods, minimum and maximum allowable I Categorization and level of detail. Similar information loan sizes, and so on. The more parameters available, the may need to be presented at varying levels of aggre- more flexible the program will be in terms of supporting gation (for example, portfolio quality by individual multiple lending and savings products. However, the loan officer, by branch office, by region, by overall greater complexity may imply a need for more advanced institution). systems maintenance. I Frequency and timeliness. Because the purpose of provid- If an MFI has branch offices, each will need its own ing information is to enable staff to take constructive management information system database. The head office actions, appropriate frequency and timeliness is critical. then needs a means by which to consolidate that data to I Performance indicators. Key performance indicators generate reports for the institution as a whole. The man- that management monitors need to be generated, agement information system needs to be designed with preferably including trend information. these decentralization and consolidation requirements in mind. SECURITY. Security features protect data, restrict user A good loan portfolio package will allow for easy access, and limit possible fraud. Security features cannot transfer and reconciliation of relevant data to the guarantee that fraud or data loss will not occur, but they accounting system, including summary information on can minimize the likelihood and negative effects of such disbursements, repayments, interest and fee income, sav- occurrences. ings transactions, and so forth. The data for these trans- In general, information on computer systems should actions on the loan-tracking side of the management be protected against two threats: information system must tally with the comparable data I Purposeful attempts to gain unauthorized access to on its accounting side. Some software packages feature data or to certain user functions fully integrated loan-tracking and accounting capabili- I Data corruption from hardware, software, or power ties that transfer relevant data automatically; however, failures. such systems are expensive and unnecessary for most These threats imply two different types of security. MFIs. Most institutions need a loan-tracking system The first requires securing data and user routines from that produces a transaction report that summarizes the unauthorized use. For example, user routines such as relevant information. This information is then recon- loan disbursement and loan repayment should not be ciled with the accounting transactions on a timely basis. accessible by all staff. This type of security is referred to 176 MICROFINANCE HANDBOOK Box 7.3 Improving Reporting Formats: Experience of the Workers Bank of Jamaica T HE W ORKERS B ANK OF J AMAICA GREW OUT OF THE Although the system provided adequate capabilities for Government Savings Bank, established in 1870. The inputting information and making financial calculations, the Workers Bank inherited the Post Office Banking Network, bank found its standardized reports insufficient. Management with almost 250 post office banking windows where small was not receiving the accurate and timely information it savers maintained accounts. By 1995 the Post Office needed about portfolio activity by loan officer, post office, Division had more than 95,000 small savers with more than and region. And neither loan officers nor managers were get- US$10 million in deposits. That same year the bank decided ting reports that would enable them to properly manage a to set up a microbanking unit to expand microfinance ser- microloan portfolio with weekly payment cycles. To help vices by offering microloans through the Post Office develop report formats that would better meet the bank's Division. needs in managing its new microloan portfolio, the bank At that time the bank was in the early stages of develop- hired an independent consultant with many years of experi- ing a comprehensive management information system. It ence in computer programming and managing MFIs to determined that it needed to purchase a system that would advise on design. manage its small loans and the small savings accounts in the The new reporting formats meet the needs of loan offi- Post Office Division until the larger management informa- cers and managers for controlling delinquency and managing tion system was fully operational. After analyzing several sys- a growing microloan portfolio. They provide timely and tems, the bank chose an internationally available system to accurate information, are easy to use, and customize infor- manage its postal banking operation. mation to fit the needs of users. Source: Waterfield and Ramsing 1998. as "system security" and includes such features as user I Multiple computers in a server-based local area net- passwords. The second type of threat requires frequent work (LAN) backup of data and a method to re-index lost or corrupt- I Wide-area networks that connect computers in dis- ed data. This type of security is called "data security." tant locations (WAN). An MFI should have as many strong security fea- Networks offer many advantages but they also add tures as possible in the software. These features should additional complexity. Many MFIs might be concerned extend beyond basic passwords entered by staff, which about installing a network, considering them to be both grant them access to different levels of activity, to thor- expensive and heavily dependent on technical support. ough audit trails that determine who can input and However, computer networking technology now sup- change specific information in the database. In addi- ports simple networks requiring fairly minimal technical tion, a secure system should have means to prohibit support. tampering with the databases from outside the system. Benefits gained from a simple peer-to-peer network are associated with information efficiency. In general, many SOFTWARE AND HARDWARE ISSUES. In evaluating loan- people need access to information at the same time. In a tracking software, a number of key technical issues must networked system, critical information is stored on one be explored. (For a more complete treatment of these soft- user's computer but is available to all staff who need ware and hardware issues, see Waterfield and Ramsing access to it. Networking requires more sophisticated pro- 1997.) These include should the MFI run its system on a gramming techniques to ensure data integrity, so that network, and if so, what kind of network is most appropri- when more than one person is accessing the same data, ate, and which operating system should the software run only one is able to modify the data at a time. While a on (DOS, Windows, Windows 98, or Windows NT). peer-to-peer approach is not recommended for a growth- There are essentially four applications of personal oriented MFI, it could be implemented in smaller MFIs computer architecture for MFIs: that want the benefits of networking at minimal cost. I Stand-alone computers Growth-oriented and larger MFIs could make effi- I Multiple computers in a peer-to-peer network ciency gains by employing a "server-based" local area MANAGEMENT INFORMATION SYSTEMS 177 Box 7.4 Management Information Systems at the Association for the Development of Microenterprises, Dominican Republic T HE A SSOCIATIO N FO R T H E D EVE L OPME NT O F more advanced UNIX operating system instead of the more Microenterprises (ADEMI), one of the early prominent common PC-based systems. The system consists of fully microcredit programs in Latin America, was also one of the integrated modules for loan portfolio, savings data (only 170 first to automate its information systems. In 1984 it devel- time deposits), accounting, client data, payroll, and fixed oped a database system that was advanced enough even to assets. automate the printing of loan contracts and client re- The system allows all the Santo Domingo offices and the payment vouchers and let senior managers monitor the bal- regional offices to be connected on line. Other branch ance of their bank accounts directly from their desktop offices, however, are not connected into the management computers. information system. They send in transaction receipts every With technological changes and growth to more than few days to the regional office, where information is input 18,000 loans, ADEMI revamped its management informa- and reports are generated. These reports are then sent to the tion system in 1994. Nearly all the work was done in-house branch offices, where they are compared with independent by an experienced management information system depart- systems used by the branches to verify the accuracy and com- ment. As with the earlier system, ADEMI chose to use the pleteness of data entry. Source: Waterfield 1997. network (LAN). A server is a personal computer with a DOS-based systems often have a problem accessing suf- faster processor, more memory, and more hard disk ficient memory to run efficiently. Also, the operating space than a typical stand-alone computer. The data to speed of systems depends heavily on the programming be broadly accessible is stored on the network. Network language used, the level of hardware, the amount of cards and cables connect the server to individual com- memory, and, even more important, the skill level of puters, enabling each user to access the information the programmer. A system may appear to operate well stored on the server. Security and data integrity issues initially, but after reaching a certain number of clients are handled by the network software and by the data- or accumulating a year's worth of information it may base system. slow down unacceptably. In selecting a system, it is crit- LANs consist of computers in the same building con- ical to carefully assess the future scale of activity and the nected by cables. Wide area networks (WANs) are com- capacity of each system under consideration. puters that are connected via phone lines, so that remote offices can access and transfer data as if they were in the SUPPORT. The role of the software provider does not end same place. However, WANs are beyond the technical when the system is installed. There is always a need for and cost capabilities of most MFIs. For MFIs with continuing support to fix bugs, train staff, and modify the remote branch offices, the best alternative for transfer- program. This need for ongoing support is the primary ring data is by modem at the end of the day or by reason that many MFIs develop their own customized sys- diskette on a weekly or monthly basis. tems. They then have the in-house capacity to provide Most loan-tracking systems operate in DOS, which these support services, albeit at substantial cost. However, has been adequate for MFIs of all scales. There is no real many MFIs cannot afford to have a systems designer on restriction on the number of clients or transactions that staff and must depend on outside support. If it is available a DOS-based system can process. Practically speaking, locally, there is a good chance that the support will be reli- however, systems designed for use a few years from now able. However, many systems currently provide remote will need to be redesigned to work under one of the support, which is quite risky as well as potentially expen- Windows operating systems. Making a substantial sive. Many problems cannot be communicated and investment in a DOS-based system should be done only resolved quickly by e-mail, phone, or courier. Because an after careful thought. MFI cannot afford to have its management information Because DOS was developed before the memory system not functioning for an extended period of time, the capacity of personal computers was as large as it is now, viability of remote support should be explored carefully. 178 MICROFINANCE HANDBOOK Other considerations to make in choosing loan- The following framework presents a series of over- tracking systems include: lapping steps involved in implementing a management I Whether the system has been used internationally information system. I Whether the system has been in use for a sufficient amount of time to be well tested and to have reached Institutional Assessment a high level of reliability I Whether the company supporting the system can make Before deciding on a management information system the a strong case that it is able to support the MFI's opera- MFI must assess the "fit" between the MFI's practices-- tions (that is, provide training and technical assistance) both present and future--and the software's capabilities. I Whether the system has already been used for As stated previously, management information systems microfinance (as opposed to commercial banking). differ not only in philosophy and approach but also in features and capabilities. Any selection process needs to Client Impact Tracking Systems begin with absolute clarity about the expected functionali- ty of the system. Most MFIs want to track socioeconomic and impact data (that is, the number of jobs created, the number of Configuration clients moving out of poverty). However, client impact tracking is even less standardized than portfolio manage- This stage follows closely on the work done in the insti- ment and presents major challenges to MFIs looking for tutional assessment. Configuration consists primarily of: appropriate software packages. Virtually no off-the-shelf I The general ledger, the basis of which is the chart of system for monitoring impact is currently available. accounts. Sometimes modification of the existing Institutions that choose to monitor impact have normal- chart of accounts is required to meet the redefined ly selected information-gathering approaches that com- plement their other information collection processes. For example, they may use data already gathered in the loan application process and then keep track of this Box 7.5 Framework for a Management Information System at Freedom From Hunger information on a spreadsheet. One approach that is often used to track impact-related FREEDOM FROM HUNGER, A SUCCESSFUL INTERNATIONAL NGO operating out of Davis, California, has developed a information is to incorporate the relevant data into the "Framework for Designing a Management Information loan-tracking system. Special screens and reports can be System." The framework consists of 17 steps for designing designed in a customized system. However, existing soft- and implementing an accounting and management infor- ware packages frequently include several "user-defined" mation system for MFIs. Each step includes a specific list fields in client and loan records. These can be set up to of questions and issues pertaining to each step. track and report on significant impact-related data. The framework was developed to strengthen and streamline the accounting system of the Reseau des Caisses Populaires, Freedom From Hunger's partner in Installing a Management Information System Burkina Faso. The framework may be used as an imple- mentation or management tool to guide the user through Many managers do not fully comprehend the substantial key issues in assessing the MFI's information needs, the process that needs to be followed when selecting and system alternatives available, and system implementation. The tool provides the user with a logical structure to installing a management information system. The task is examine the management information system needs both often much greater than typically envisioned. The inter- in the field and at headquarters. (For detailed informa- relationship between the management information sys- tion on the assessment framework see SEEP Network tem and the entire structure and operating procedures of 1996.) the institution means that in some cases the installation Source: SEEP Network 1996. of a system must be done concurrently with the restruc- turing of the entire organization. MANAGEMENT INFORMATION SYSTEMS 179 needs of the institution as well as any requirements Testing or limitations of the software package. I Loan-tracking systems, which define the various finan- Once the software design is finalized, it is essential to run cial products, each with its own specific characteris- a rigorous test of the system with actual data. This testing tics, such as minimum and maximum amounts, phase serves two purposes: it allows the behavior of the interest rate methods, the various "linkages" that the system to be carefully studied, identifying any bugs or MFI uses between different accounts (such as block- problem areas needing changes; and it allows the devel- ing savings when a client has a loan), and treatment of opment of a strategy for data conversion or input of the delinquency. initial data to the system. I The structure of branch relationships, which determines For the general ledger, testing consists of processing how information will be shared and consolidated sample transactions, including receipts of income, pay- among branches. ment of checks, and journal entries. Trial balance and If the initial assessment has been doing well, the con- financial statements should be generated, with the results figuration of the system can be handled fairly straight- confirmed and presented in the formats desired by the forwardly, provided the MFI's needs are fully institution. compatible with the software's capabilities. If changes to For loan tracking, the testing phase is more complex the software are needed to accommodate the require- and includes verifying that crucial procedures are han- ments of the MFI, then the process of software modifi- dled correctly. Key questions include: cation begins. I Are repayment schedules, interest charges, penalties, and delinquencies being properly calculated? Software Modifications I Does the system crash inexplicably? I Does the network function adequately? This is potentially the most costly and time-consuming I Does the system allow for corrections of data entered component of installing a management information in error? system. If the software package has not been used by I Are there user-friendliness issues that need to be other MFIs or similar operating environments (such as addressed? different types of MFIs in different countries) or if the software has not previously been adapted to a wide Data Transfer variety of operating methodologies (for example, indi- vidual lending or group lending), there may be a need This is another time-consuming and costly process. The for major modifications. Extensive programming may volume of start-up information that must be entered can still be required if the MFI is unwilling to compromise be substantial. For the portfolio system, inputting names on its operating methods and accounting conventions, and other data on clients as well as all outstanding balances even if the software package is nearly exhaustive in for loans and savings accounts is extremely time consum- functionality. ing. At a minimum for the general ledger, opening The modification process is costly and time-consum- amounts for all balance sheet accounts must be entered. ing for several reasons. The source code (or underlying A frequently encountered problem is incompatibility logic and commands of the software) for a complex sys- between the way the previous system handled transactions tem can only be modified by a handful of people--ide- and the way the new management information system ally, the original programmers. Any change, no matter treats the same transactions. For example, if delinquency how minor, needs to be carefully tested and debugged, is calculated even slightly differently by the two systems, because a change in one area of the program can affect the new system may calculate penalties on outstanding what seem to be wholly unrelated areas of the program. loans in a way not agreed to with the client. Program modification should only be undertaken by It is often difficult to predict precisely how long or highly qualified programmers and only if the MFI can- difficult this data transfer process will be, even if there not perform its essential work by using a packaged loan has been a careful initial assessment. One way to avoid tracking system. these problems is to continue to use the existing portfolio 180 MICROFINANCE HANDBOOK system until all outstanding loans are repaid, entering only new loans into the new system. Box 7.6 Running Parallel Operations in the Fédération des Caisses d'Epargne et de Crédit Agricole Mutuel, Training Benin IN THE LOCAL CREDIT UNIONS OF THE FÉDÉRATION DES Because a full-featured management information system Caisses d'Epargne et de Crédit Agricole Mutuel, account- is complex, implementation may require major shifts in ing was done manually until computers started slowly to the operational procedures of an institution, requiring be introduced and tested in 1993. For the first credit extensive training at all levels of staff. unions to be computerized, it was required that account- ing be recorded both manually and by computer for one full year, so that the staff could become fully acquainted Parallel Operations with the new software and procedures and so that the information systems technicians could make necessary After data transfer and staff training, the new manage- adjustments to the software. As experience is gained by ment information system is ready to be used. However, it both parties, managers of the federation believe the is important to run the new system in parallel with the required period for running accounting in parallel could old system. This is essential to ensure that the new system be brought down to two months. runs reliably and generates reports that can be trusted. Source: Contributed by Cecile Fruman, Sustainable Banking with During the period of parallel operation, data should the Poor Project, World Bank. be entered into both systems and the outputs carefully compared. Any discrepancies should be evaluated and accounted for. Any errors or bugs in the new system system that offers every ideal feature. The more that should be carefully documented and corrected. is demanded of a system , the more complex that sys- The duration of parallel operations can vary, but it tem becomes and the less likely it is to work correctly should generally last at least two months to ensure that and trouble free. nearly every client has come to make at least one pay- I An MFI should consider adapting some of its proce- ment and that the system has gone through two dures to the standards of the loan-tracking system. Of month-end closing processes. Once management is sat- course, this would ideally function the other way isfied that the new system is performing well, the old around: a management information system would be system can be discontinued, but all printouts and data capable of implementing every aspect of an institu- files should be carefully stored for future reference (box tion's policies and procedures. However, acceptance 7.6). of existing features that do not fundamentally alter the institution's goals and objectives may save thou- Ongoing Support and Maintenance sands of dollars and months of customizing time. As discussed previously, the institution's need for man- agement information system support continues once the Appendix 1. Overview of Commercial system is installed. The cost of support will depend on Management Information System Software how stable and reliable the system is and will normally Packages decline as the MFI grows more experienced with it and thus more capable of resolving problems. There are no set prices for commercial management There are two important lessons that can be gleaned information software packages.1 Pricing depends on from the experience of MFIs in installing management specific conditions in each installation, such as the information systems: scale of the MFI, the extent of customization, and the I An institution should be content with the essentials. availability of local technicians who the software firm An MFI should not seek a management information can call on. None of the commercial firms referenced 1. This appendix is excerpted from Waterfield (1997). MANAGEMENT INFORMATION SYSTEMS 181 in this overview provided a set price; however, price The SRTE 2.0 version of the system covers loans, sav- ranges and typical contract prices have been included ings accounts, time deposits, current accounts, customer where possible. information, and general ledger in one integrated pack- age. EXTE, the extended version of the system, allows the introduction of substantial modifications and cus- IPC Banking System tomization through access to parts of the source code. Institutions with procedures that differ substantially from The IPC banking system was developed over seven years standard systems may need to acquire this version of the by IPC Consulting for use in MFIs participating in long- software. term technical assistance contracts and has never been The purchase of the software includes a warranty installed independent of a technical assistance contract. support from an authorized support provider by phone, The system is currently used in Albania, Bolivia, Brazil, fax, or email for three months. FAO recommends that Colombia, Costa Rica, El Salvador, Paraguay, Russia, an authorized support provider also be contracted to Uganda, and Ukraine. help with installation and customization and that a con- The system runs in DOS on stand-alone PCs and tract for longer-term assistance be established with one Novell or Windows networks. It contains modules for of the support providers when the warranty period has loan monitoring, savings, fixed deposit, and accounting. It expired. provides extensive support of the entire lending process, The current price (1997) of the SRTE 2.0 is US$800 starting with registration of application, support for loan for each installation. This includes an unlimited number application analysis, integrated generation of loan agree- of users per installation, but no customization can be ment and payment plan, wide range of possible payment made, nor is any installation support available for this plans, and various interest rate calculation procedures. The price. The access fee of the customizable EXTE version is system supports a full range of teller-based operations, US$8,000. Customization can only be done for addi- including monitoring cash and check payments, monitor- tional fees by an authorized support provider. The cost of ing various cashiers, and a money-changing function. a contract for assistance with minor customization and IPC personnel visit on site for installation, configura- configuration assistance, installation, and basic training, tion, and training. Maintenance contracts are available, typically runs between US$40,000 and US$70,000, which provide for software modifications and updates. depending on the size of the MFI and the extent of the The system price depends on the type of contract, the modifications. extent of modifications required, and the number of installations involved, but the minimum contract amount considered is US$50,000. The Reliance Credit Union Management System The FAO Microbanking System This system is owned by CUSA Technologies, Inc. (CTI), and has been marketed primarily to credit unions FAO's Microbanker was developed over the past 10 years in the United States, although the company is also work- and is currently running in 900 installations in more ing at developing an international market. than 20 countries in Asia, Africa, Latin America and the It is currently installed in over 100 locations in the Caribbean, Eastern Europe, and some former Soviet United States, and installation procedures are currently Union republics. under way in Puerto Rico, Central America, and Australia The system was developed as a low-cost computer (26 credit unions with more than 150 branches). The software alternative for the automation of banking opera- system is currently available in English and Spanish. A tions for small and medium-size financial intermediaries. French version is in development. It runs on basic PC equipment and is flexible enough to The Reliance software incorporates modules for on- work as a single teller, a stand-alone installation, or as a line teller transactions, loan processing, mortgage multiteller installation. lending, integrated general ledger, ATM transaction pro- 182 MICROFINANCE HANDBOOK cessing, credit card processing, electronic payroll process- client. Maintenance is contracted for a fee negotiated ing, share draft processing, and signature and picture ver- with the client. ification. Technical assistance is available 12 hours per day, Monday to Saturday, by phone and Internet. CTI is cur- Solace for Workgroups rently establishing a Spanish support center (a distribu- tor) in Puerto Rico. This system is owned by Solace, an Australian software System pricing is by module, by asset size, and by the firm, but its distribution has been licensed to De- number of system users. CTI was reluctant to provide centralized Banking Solutions (DBS), a South African any price estimates, but indicated that the system is gen- firm actively marketing the software to the microfinance erally priced for larger institutions. community. The system is currently available in English and is installed in 13 institutions in Australia, New Zealand, South Africa, and Zimbabwe. The SiBanque System The system runs on Windows 95, Windows NT, UNIX, and AS400 operating systems. The teller-based sys- The SiBanque system is supported by Centre International tem supports savings, loan, checking, and equity accounts. de Credit Mutuel (CICM) in France. The system is cur- It provides for printing of passbooks and receipts and sup- rently available in French and English, but may be trans- ports audit inquiries. It can be used in a wide range of lated to other languages. SiBanque is currently installed in institutional settings, including networks of more than approximately 100 offices in 8 institutions in Burundi, 200 users. Technical support is provided via the Internet, Cameroon, Congo, Guinea, Mali, and Senegal. with potential local support through a joint venture with a The system is set up for "front office" and "back local partner when business volume warrants. office" transactions, is oriented toward credit union Price is based on a nominal solace software purchase structures, and provides some configuration through the fee of US$3,000 per user terminal (for example, an instal- use of parameters. It is a dBase-compatible system that lation requiring access to 20 computers would cost operates on DOS. US$60,000) plus run-time costs and an annual support Since the software is not supported by a commercial fee dependent upon local requirements and conditions. software firm, the pricing situation is nonstandard. Annual support is often based on the number of terminals The software itself does not have a cost, but CICM as well. A typical installation, therefore, can cost in excess requires MFI staff to be trained as a requirement prior of US$100,000, and at least one installation costs to installation. The cost of this training is billed to the US$370,000, all expenses included. MANAGEMENT INFORMATION SYSTEMS 183 Appendix 2. Criteria For Evaluating Loan Tracking Software Ease of use Hardware/software issues Documentation Programming language Tutorials Data storage format Error handling Network support Help screens Operating system Interface Access speed Features Support Languages Customization available? Setup options Training Methodology issues Cost issues Loan product definition multiple loan products Reports principal repayment methods Existing reports monitoring methods New reports fund accounting of portfolio Print preview data disaggregation Printers supported interest calculations Width of reports fee calculations Security savings Passwords and levels of authorization Branch office management and consolidation Data modification Linkages between accounting and portfolio Backup procedures Audit trails Source: Developed by Waterfield and Sheldon (1997) for Women's World Banking. Sources and Further Reading Waterfield, Charles, and Nick Ramsing. 1998. "Management Information Systems for Microfinance Institutions: A SEEP (Small Enterprise Education and Promotion) Network. Handbook." Prepared for CGAP (Consultative Group to 1995. Financial Ratio Analysis of Micro-Finance Institutions. Assist the Poorest) by Deloitte Touche Tohmatsu Interna- New York: PACT Publications. tional, in association with MEDA Trade & Consulting and ------. 1996. Moving Forward: Tools for Sustainability and Shorebank Advisory Services, Washington, D.C. Expansion. New York: PACT Publications. Waterfield, Charles, and Tony Sheldon. 1997. "Assessing and Waterfield, Charles. 1997. "Budgeting for New Information Selecting Loan Tracking Software Systems." Women's Systems for Mature Microfinance Organizations." World Banking, New York. Calmeadow. Toronto, Canada. Women's World Banking. 1994. Principles and Practices of Financial Management. New York. Part III­ Measuring Performance and Managing Viability CHAPTER EIGHT Adjusting Financial Statements T o analyze the financial performance of MFIs prop- ments. This decision would be made by each MFI or erly, it is necessary to ensure that their financial reader individually. However, the following adjustments statements are consistent with generally accepted should be made (either on the financial statements or in accounting principles. Due to the structure of many micro- an external spreadsheet) to reflect the unique nature of finance institutions and their initial reliance on donor fund- MFIs and to account for the benefit of donor financing: ing, adjustments to balance sheets and income statements I Accounting for subsidies. Donor funding either for are often needed before the financial performance can be loan capital to on-lend or to cover operating expenses analyzed (see chapter 9). There are generally two types of is generally provided as either grants or concessional adjustments required: those that are necessary to adhere to loans (loans with lower than market rates of interest) proper accounting standards, which are often neglected in and therefore contains a subsidy to the MFI. MFIs, and those that restate financial results to reflect more I Accounting for inflation. The cost of inflation results accurately the full financial position of the MFI. in a loss in the real value of equity and other balance The first type requires accounting entries that adjust sheet accounts. an MFI's financial statements. The amount and type of These adjustments are needed to assess the MFI in an adjustment required will vary with each MFI, depending accurate and meaningful manner and to allow for com- on its adherence to standard accounting principles.1 parisons among institutions to some extent. These adjustments include: Finally, MFIs may wish to restate their financial state- I Accounting for loan losses and loan loss provisions. MFIs ments in "constant currency terms," particularly if they that fail to properly consider potential or existing loan are operating in a highly inflationary environment. This losses, understate total expenses, and overstate the means expressing the financial results in real rather than value of the loan portfolio. nominal terms by converting financial data from previ- I Accounting for depreciation of fixed assets. Failing to ous year nominal values (in local currency) to constant, depreciate or inaccurately depreciating fixed assets present-year values. This permits year-over-year compar- understates the true expense of operations. isons and performance trend analysis. I Accounting for accrued interest and accrued interest This chapter presents the reasons that adjustments expense. Accruing interest revenue on delinquent should be made, ways to determine the appropriate loans results in higher stated revenue than actually amount of adjustments, and ways to make them. received, thereby overstating profits, and failure to Examples of adjusted financial statements for subsidies accrue interest expense on liabilities understates and inflation are provided in the appendixes. This chap- expenses. ter will be of interest primarily to practitioners or consul- The second type of adjustments do not necessarily tants evaluating MFIs. To make the most use of this have to be formally recorded in the MFI's financial state- chapter, it is necessary to have some understanding of 1. Portions of this chapter are based on material originally published in Ledgerwood 1996. 187 188 MICROFINANCE HANDBOOK accounting and financial analysis. Readers who are not as possible and do not want to admit (to donors or oth- familiar with financial statements may wish to provide ers) that some of the loans they have made are not per- this chapter to the financial manager of an MFI or a forming. Further, many MFIs choose to maintain bad financial consultant who is working with the MFI. (For loans on the books because they feel that if they were to basic instruction in accounting and the creation of finan- write off a loan, all efforts to collect it would stop. cial statements, see Ledgerwood and Moloney 1996.) Initially, MFIs may avoid accounting for loans losses, but as MFIs mature and their exponential growth slows down, the problem becomes more significant. Carrying Accounting Adjustments loans on the books that have little or no chance of being repaid overstates the assets on the balance sheet and Among the adjustments that are often needed before the results in a lower-than-expected yield on assets (this is financial performance of an MFI can be analyzed are reflected in lower revenue than expected based on the accounting entries that adjust an MFI's financial state- effective yield calculation detailed in chapter 5). ments so that they adhere to proper accounting stan- Furthermore, accurately accounting for loan losses on a dards. These adjustments include accounting for loan periodic basis saves the MFI from taking a large loss (and losses, accounting for depreciation of fixed assets, and a consequent increase in expenses) all at one time for accounting for accrued interest and accrued interest loans that were made over a period of years (Stearns expense. 1991). Loan losses should be recorded close to the period in which the loan was made. Accounting for Loan Losses In order to accurately account for loan losses, the first step is to determine the quality of the loan portfolio. This Accounting for loan losses is an important element of is done by referring to the "portfolio report" (table 8.1). MFI financial management and one of the most poorly managed. To accurately reflect the financial performance T HE PORTFOLIO REPORT . One of the most important of an MFI, it is necessary to determine how much of the monitoring tools of an MFI is its portfolio report, which portfolio is generating revenue and how much is likely to provides information on the quality of the loan portfolio be unrecoverable. This is done by examining the quality and the size of the lending activities. Since the loan port- of the loan portfolio, creating a loan loss reserve, and folio is most often the MFI's largest asset and, therefore, periodically writing off loans. its main revenue-generating asset, ensuring accurate and Many MFIs do not like to make loan loss reserves timely reporting of the portfolio is crucial to the financial because they are anxious to report their expenses as low management of an MFI. Table 8.1 Sample Portfolio Report Portfolio data 1995 1994 1993 Total value of loans disbursed during the period 160,000 130,000 88,000 Total number of loans disbursed during the period 1,600 1,300 1,100 Number of active borrowers (end of period) 1,800 1,550 1,320 Portfolio outstanding (end of period) 84,000 70,000 52,000 Average outstanding balance of loans 75,000 61,000 45,000 Value of payments in arrears (end of period) 7,000 9,000 10,000 Value of outstanding balance of loans in arrears 18,000 20,000 20,000 Value of loans written off during the period 500 3,000 0 Average initial loan size 100 100 80 Average loan term (months) 12 12 12 Average number of credit officers during the period 6 6 4 Source: SEEP Network and Calmeadow 1995 ADJUSTING FINANCIAL STATEMENTS 189 Portfolio outstanding refers to the principal amount of off (discussed below). Defaulted loans are often called loans outstanding. Projected interest is not usually consid- such if there is little hope of ever receiving payment. ered part of the portfolio. Principal outstanding is an Some MFIs determine that a loan is delinquent if no asset for an MFI; interest contributes to the income of an payment was received on the due date. However, many MFI and is recorded as revenue. loans are paid within a few days of the due date, and Readers should note that some MFIs capitalize other MFIs allow a short period for late payments and do (record as an asset on the balance sheet) the interest rev- not consider a loan delinquent until one or two weeks or enue expected when the loan is disbursed (particularly if payment periods have passed. they are operating with British accounting standards) or Some MFIs fail to consider the amount of time a loan when the loan term has ended and an outstanding bal- is in arrears and base the likelihood of default (loan loss) ance is still remaining. For the purposes of this hand- more on the term of the loan--in other words, a loan is book, we assume that interest is not recorded until reported as overdue when the loan term ends, rather than received. However, those readers working with MFIs that after a certain period of time in which no payments have do capitalize interest revenue should be aware that this been received. This handbook does not take into account results in higher total assets than if the MFI did not capi- whether or not the loan term has ended. Any loan that talize interest payments and affects the way in which loan has an amount in arrears is considered delinquent, based payments are recorded for accounting purposes. on the number of days it has been in arrears. When a loan payment is received, if interest has not MFIs must also determine whether or not they consid- been capitalized, interest revenue is recorded on the er the entire outstanding balance of loans that have an income statement as an increase in revenue, and the prin- amount in arrears (portfolio at risk) to be delinquent or cipal repaid is recorded on the balance sheet as a reduc- just the amount that has become due and has not been tion in loans outstanding (asset). For MFIs that capitalize received. Generally, MFIs should consider the entire out- interest, the entire loan payment amount is recorded as a standing amount as delinquent, because this is the amount reduction in the loan portfolio, since interest revenue was that is actually at risk (that is, if the borrower has missed a recorded (on the income statement) when the loan was payment, chances are he or she may miss more payments made (see the section on accounting for accrued interest or will simply stop repaying the loan all together). and accrued interest expenses below). A delinquent loan becomes a defaulted loan when the Both the "value of payments in arrears" and the chance of recovery becomes minimal. Defaulted loans "value of outstanding balance of loans in arrears" refer to result in loan write-offs. Write-offs should be considered loans that have an amount that has come due and has after a certain period of time has passed and the loan has not been received. The "value of payments in arrears" not been repaid. The decision to write off a loan and the includes only the amount that has come due and not timing of doing so are based on the policies of the MFI. been received, while the "value of outstanding balance of Some MFIs choose to write loans off relatively quickly so loans in arrears" includes the total amount of loans out- that their balance sheets do not reflect basically worthless standing, including the amount that has not yet become assets. Others choose not to write off loans as long as due, of loans that have an amount in arrears. This is there is a remote possibility of collecting the loan. commonly referred to as the "portfolio at risk." Regardless of when write-offs take place, it is important Note that the terms "arrears," "late payments," "over- that an MFI have an adequate loan loss reserve (discussed due," "past due," "delinquent," and "default" are often below) so that the net value of loans stated on the bal- used interchangeably. Generally, loans that are "in ance sheet accurately reflects the amount of revenue-gen- arrears, past-due, and overdue" have become due and erating assets. have not been paid. These are also referred to as delin- Based on an analysis of the quality of the loan portfo- quent loans (that is, "this loan is delinquent" means it lio, managers can determine what percentage of the loans has an amount that is owing--is in arrears, past due, is likely to default. To reflect this risk and the corre- overdue--and has not been paid). To say that loans are sponding true value of the loan portfolio accurately, it is in default is also referring to delinquent loans, but most necessary to create a balance sheet account called the often to loans that have been or are about to be written "loan loss reserve." 190 MICROFINANCE HANDBOOK LOAN LOSS RESERVE. A loan loss reserve is an account at risk reports should be compared over time with pre- that represents the amount of outstanding principal vious reports to determine if the portfolio at risk is that is not expected to be recovered by an MFI. It is the increasing or decreasing and to determine if new poli- amount "reserved" to cover losses of the loan portfolio. cies implemented to control delinquencies are working. The loan loss reserve is recorded as a negative (or con- This is particularly important if the portfolio is growing tra) asset on the balance sheet. The loan loss reserve rapidly, because recording a large number of new loans reduces the net portfolio outstanding. (Some MFIs can hide a delinquency problem (that is, the overall record the loan loss reserve as a liability. The net effect loan quality can appear good because a large portion of is the same.) loans may not have become due). The amount of the loan loss reserve should be based Once an aging analysis has been completed, the loan on historical information regarding loan defaults and loss reserve is established based on the likelihood of the amount of time loans have been delinquent. Past loan recovery for each aging category. (The loan loss performance of delinquent loans is the most important reserve as a percentage of loans outstanding may be reg- indicator for predicting future performance. ulated in some countries, which would be applicable if To determine an adequate loan loss reserve, an aging the MFI is regulated as a formal financial institution.) analysis should be performed periodically (monthly or For example, if approximately 10 percent of loans that quarterly). Aging analysis refers simply to the classifica- are past due less than 30 days have historically default- tion of delinquent loans into the periods of time that ed, a provision of 10 percent should be created. If 50 they have been in arrears. Loan portfolios are aged percent of loans that are past due greater than 30 days using categories such as 1­30 days, 31­60 days, 61­90 but less than 90 days have historically defaulted, a 50 days, 91­120 days, and more than 120 days. The choice percent provision should be created (table 8.2). of aging category depends on the frequency of payments Once the loan loss reserve is calculated, it is then and the loan terms of the MFI. Generally, the categories necessary to compare it to the existing loan loss reserve should be based on one or two payment periods (week- (if there is one) on the balance sheet and determine if it ly, biweekly, monthly, and so on). needs to be increased. Increases in loan loss reserves are In the aging analysis, the entire outstanding balance made by creating a loan loss provision. of the loan that is in arrears (portfolio at risk) is considered. LOAN LOSS PROVISION. The loan loss provision is the The purpose of calculating an aged portfolio at risk amount expensed in a period to increase the loan loss is that it allows MFIs to estimate the required loan loss reserve to an adequate level to cover expected defaults reserve. The longer a loan has been delinquent, the of the loan portfolio. It is based on the difference greater the likelihood of default and, therefore, the between the required loan loss reserve and the current greater the need for a loan loss reserve. Aged portfolio outstanding loan loss reserve. Table 8.2 Sample Loan Loss Reserve Calculation, December 31, 1995 (A) (B) (C) (D) Number of loans Portfolio at risk Loan loss reserve Loan loss reserve ($) past due (outstanding balance) (%) (B) x (C) 1­30 days past due 200 8,750 10 875 31­60 days past due 75 5,000 50 2,500 61­90 days past due 60 2,500 75 1,875 90­120 days past due 15 1,100 100 1,100 > 120 days past due 10 650 100 650 Total 360 18,000 -- 7,000 Source: SEEP Network and Calmeadow 1995. ADJUSTING FINANCIAL STATEMENTS 191 To make the adjustment, the first time a loan loss affect the income statement when they occur. Write- reserve is created, a loan loss provision is recorded on the offs reduce the "loan loss reserve" and the "outstanding income statement as an expense (debit) in an amount loan portfolio" by equal amounts. (If the loan loss equal to the required loan loss reserve. The loan loss reserve has been recorded as a liability, it is reduced reserve is then recorded on the balance sheet as a negative when loans are written off.) The resulting effect is to asset (credit). This reduces the net outstanding loan port- leave the net portfolio on the balance sheet unchanged folio. Subsequent loan loss provisions are recorded on the because the reserve has already been made (and ex- income statement in the amount necessary to increase pensed as a loan loss provision). When making a write- the loan loss reserve to its required level. Once debited off, a reduction in the loan loss reserve (which is a on the income statement, the amount is added (credited) negative asset) is recorded as a debit. The write-off is to the existing loan loss reserve on the balance sheet. offset by a reduction (credit) of the outstanding loan Note that the loan loss provision is a noncash expense and portfolio. does not affect the cash flow of an MFI. If the loan loss reserve is too low relative to the value of loans to be written off, then the loan loss reserve needs RECORDING LOAN WRITE-OFFS. Once it has been deter- to be increased by making an additional loan loss provi- mined that the likelihood of a particular loan being sion on the income statement. repaid is remote (the borrower has died, left the area, or If a loan that was previously written off is recovered simply will not pay), a write-off occurs. Write-offs are (in other words, the borrower has repaid the loan) then only an accounting entry; they do not mean that loan the full amount recovered is recorded as revenue (credit) recovery should not continue to be pursued if it makes on the income statement. This is because the principal economic sense. Writing off a loan does not mean the amount written off was recorded as an expense organization has relinquished its legal claim to recover (through the loan loss provision that created the loan the loan. loss reserve), and, therefore, if recovered it is recorded To make the adjustment, actual loan losses or write- as revenue and not as a decrease in outstanding loan offs are reflected on the balance sheet only and do not portfolio (assets). The outstanding loan portfolio is not Box 8.1 The Impact of Failure to Write Off Bad Debt MFIS OFTEN FAIL TO WRITE OFF BAD LOANS BECAUSE OF "old" bad debt that occurred in the past and, therefore, legal restrictions or the desire to show exaggerated profits should have been written off. Instead, it is often measured or a high portfolio amount. When bad debts are not writ- as loan collection over disbursed (the amount falling due ten off, reported loan recovery performance may be severe- plus the cumulative total of previous years' bad debt). ly distorted. The calculation below provides an example of Assume that the MFI lends 100 units of currency every the type of distortion that can arise when a bad debt is not year of which 10 are never collected and never written off. The written off but rather carried over as part of the outstand- denominator (disbursed) consequently increases by 10 units ing loan portfolio. Loan-collection performance should be annually, resulting in a misleading deterioration of the recov- measured as follows: loan collection over the amount ery ratio while actual performance has remained constant--at falling due during the reported year net of accumulated a 90 percent annual collection ratio. Ratio Year 1 Year 2 Year 3 ... Year 10 Repaid 90 = 90% 90 = 82% 90 = 75% ... 90 = 47% Disbursed 100 10 + 100 (10 x 2) + 100 (10 x 9) + 100 This illustration proves the futility of overreliance on the mance. The absence of write-offs makes this financial ratio repayment rate as an adequate test for collection perfor- inaccurate in portraying the actual collection performance. Source: Yaron 1994. 192 MICROFINANCE HANDBOOK affected because the loan amount was already removed When a fixed asset such as a piece of machinery or a from the balance sheet when the write-off occurred. motorcycle is purchased, there is a limited time that it The offsetting entry (debit) is an increase in assets, usu- will be useful (that is, able to contribute to the genera- ally cash. tion of revenue). Depreciation expense is the account- To determine if a write-off has taken place, it is neces- ing term used to allocate and charge the cost of this sary to have the previous year's closing balance sheet, usefulness to the accounting periods that benefit from this year's closing balance sheet, and this year's income the use of the asset. Depreciation is an annual expense statement. The first step is to determine the difference that is determined by estimating the useful life of each between the amount of the loan loss reserve in the pre- asset. Like the loan loss provision, depreciation is a non- vious year and the amount of the loan loss reserve this cash expense and does not affect the cash flow of the year. This amount should equal the amount of the MFI. loan loss provision for this year. If it does not, the dif- To make the adjustment, when a fixed asset is first ference is the amount of write-off taken in the current purchased, it is recorded on the balance sheet at the year. current value or price. When a depreciation expense N For example, the financial statements (not shown) for (debit) is recorded on the income statement, it is off- the sample MFI portfolio report in table 8.1 showed set by a negative asset (credit) on the balance sheet a loan loss reserve of 5,000 on the balance sheet at called accumulated depreciation. This offsets the gross the end of 1994, a loan loss reserve of 7,000 on the property and equipment reducing the net fixed assets. balance sheet at the end of 1995, and a loan loss pro- Accumulated depreciation represents a decrease in the vision on the income statement in 1995 of 2,500. value to property and equipment that is used up dur- Adding the 1995 loan loss provision (2,500) to the ing each accounting period. (Accumulated deprecia- 1994 loan loss reserve (5,000) should result in a loan tion is similar to the loan loss reserve in that both loss reserve for 1995 of 7,500. Since it is not 7,500 serve to reduce the value of specific assets on the bal- but only 7,000, a write-off of 500 must have ance sheet.) occurred. This can be verified by looking at the sam- N For example, when an MFI purchases a motorcycle, it ple portfolio report in table 8.1. estimates its useful life for a number of years. Note that write-offs reduce the amount of portfolio Recording depreciation on the motorcycle is a process at risk. MFIs that wish to report a healthy portfolio (or of allocating the period cost to the accounting periods minimal portfolio at risk) may choose to write off loans that benefit from its use. If the useful life is estimated more frequently than they should. The decision on at, say, five years, then depreciation would be record- when to write off a loan should be based on a sound ed for each of the five years. This results in a reduced policy established and agreed to by the board members book value of the motorcycle on the MFI's balance of an MFI. sheet. There are two primary methods of recording depreci- Accounting for Depreciation of Fixed Assets ation: the straight-line method and the declining bal- ance method. The straight-line method allocates an equal Many MFIs depreciate fixed assets according to gener- share of an asset's total depreciation to each accounting ally accepted accounting principles. However, some do period. This is calculated by taking the cost of the asset not. If in an analysis of an MFI's financial statements and dividing it by the estimated number of accounting there does not appear to be an operating account periods in the asset's useful life. The result is the amount (expense) called depreciation on the income state- of depreciation to be recorded each period. ment, or if the account seems too large or too small N In the example above, if the motorcycle costs relative to the amount of fixed assets on the balance 5,500, has an estimated useful life of five years, and sheet, it is necessary to adjust the financial statements has an estimated value of 500 at the end of the five and depreciate each capital asset by the appropriate years (salvage value), its depreciation per year cal- amount for the number of years that the MFI has culated by the straight-line method is 1,000, as owned it. follows: ADJUSTING FINANCIAL STATEMENTS 193 Cost ­ salvage value recorded as revenue and as an asset under accrued interest Service life in years or interest receivable. (Had the interest revenue been received, it would be recorded as revenue [credit] and as 5,500 ­ 500 cash [debit]. With accrued interest, the debit entry 5 [increase in assets] is to interest receivable rather than to = 1,000 depreciation per year cash.) MFIs vary substantially in their treatment of accrued The declining balance method refers to depreciating a interest revenue. Some only accrue interest revenue fixed percentage of the cost of the asset each year. The when the loan term has ended with a portion of the loan percentage value is calculated on the remaining undepre- still outstanding. Others accrue interest revenue during ciated cost at the beginning of each year. the term of the loan when payments become due but are N In the example, if the motorcycle is depreciated on a not received. Still others record the entire loan principal declining balance basis at 20 percent a year, the first and interest revenue at the time of loan disbursement as year's depreciation would be 1,100 (5,500 x 20%). In an asset (referred to as capitalizing interest as mentioned the second year the depreciation would be based on above) either as part of the loan portfolio or in a separate the reduced amount of 4,400. Twenty percent of asset account. Many MFIs do not accrue interest rev- 4,400 is 880, which is the depreciation expense for enue at all. The decision of whether or not to accrue year 2. This continues until the asset is either fully interest revenue and, if so, when, must be determined by depreciated or sold: each MFI based on its accounting standards. Two adjustments may be required for accrued interest Beginning Depreciation (at Ending revenue. If an MFI does not accrue interest revenue at all Year value 20 percent a year) value and makes loans that have relatively infrequent interest 1 5,500 1,100 4,400 payments (quarterly, biannually, or as a lump sum at the 2 4,400 880 3,520 end of the loan term), it should accrue interest revenue at 3 3,520 704 2,816 the time financial statements are produced. (MFIs that 4 2,816 563 2,253 have weekly or biweekly payments need not necessarily 5 2,253 451 1,802 accrue interest revenue; it is more conservative not to and If a depreciated asset is sold for an amount greater than may not be material enough to consider.) Interest rev- its recorded book value (that is, its cost net of deprecia- enue is accrued by determining how much interest rev- tion), then the book value is reduced to zero and the dif- enue has been earned but not yet become due (the ference is recorded as revenue on the income statement. number of days since the last interest payment that the Many countries set standards for depreciation by clas- loans have been outstanding times the daily interest rate) sifying assets into different categories and setting the rate and recording this amount as revenue (credit) and debit- and method for depreciating each class. This is referred ing the asset account accrued interest. to as the capital cost allowance. The second adjustment for accrued interest revenue is to reflect the amount of revenue that an MFI is unlikely Accounting for Accrued Interest and Accrued Interest to receive due to delinquent loans. MFIs that accrue Expense interest on delinquent loans overstate both the value of their assets and their revenue and should make an adjust- The last adjustment that may be required to adjust the ment to remove the accrued interest. financial statements of MFIs is an adjustment for accrued If an MFI has accrued interest on loans that have little interest revenue and accrued interest expense on liabilities. chance of being repaid, it is necessary to make an adjust- ment to the financial statements. To make this adjust- ADJUSTING FOR ACCRUED INTEREST REVENUE. Recording ment, the amount of accrued interest on delinquent interest that has not been received is referred to as accru- loans is deducted on the balance sheet by crediting the ing interest revenue. Based on the assumption that the asset where it was recorded initially (outstanding loan interest will be received at a later date, accrued interest is portfolio or interest receivable) and reversed on the 194 MICROFINANCE HANDBOOK position of the MFI as of that date. When MFIs borrow Box 8.2 Cash and Accrual Accounting for Microfinance from external sources (concessional or commercial loans Institutions or client savings or both) to fund their assets, they incur a SOME MFIS ACCRUE INTEREST REVENUE WHEREAS OTHERS liability. Liabilities carry financing charges that are paid account for it on a cash basis. MFIs that accrue interest periodically. At year-end it is likely that the MFI will owe revenue account for it as earned whether or not it has interest on borrowed funds for the period from the last come due and been paid. This is standard practice in the loan or interest payment to the day the year ends. The banking industry; however, banks have to stop accruing financial statements will need to be adjusted to reflect this interest once a loan is past due more than 90 days financing expense that has been incurred but not yet paid. (nonperforming loans). This is referred to as accrued interest expense. If interest Most MFIs account for interest revenue more conserv- owed is not accrued, the MFI's financing costs at the end atively, on a cash basis. They treat interest as income only of the year are understated, its profitability is overstated, after it is actually received. In institutions that do not and its liabilities are understated. grow, the difference between the two systems does not entail a great difference in the final results presented at To accrue interest expense, the amount of interest year end. If payment frequency is high (for example, owing as of the date the balance sheet is created is debit- weekly) then there is little difference between the two sys- ed on the income statement as a financing cost and cred- tems: in the case of weekly payments, the closing balance ited as a liability on the balance sheet in the accrued done on a cash basis would fail to reflect only a week's interest expense account. worth of interest income. Likewise, the final results under N For example, an MFI borrows 10,000 on January 1 at the two systems are not widely different in institutions an annual interest rate of 5 percent with interest due that are not growing: under a cash accounting system in a once a year on January 1 and the principal due at the no-growth program, the omission of earned but unpaid end of five years. If its year-end is December 31, it interest is approximately balanced by the inclusion of cash will have had the loan for one year less a day when it interest income earned in the prior period but received in closes its balance sheet. However, it will not have paid the present period. In MFIs where growth is strong and payments are less any interest yet, as interest is not due until January 1. frequent, the results produced by cash and accrual To properly reflect this interest expense, the MFI accounting can be substantially different. MFIs should should debit its income statement to include 500 in eventually move toward the banking industry practice of interest expense (financing costs) and credit its accruing interest. However, many MFIs do not have the accrued interest expense account (liability) 500. computing power to do accrual accounting, and they should not necessarily acquire it only for this purpose. Source: Christen 1997. Adjusting for Subsidies and Inflation Unlike the three adjustments described above, which income statement by decreasing interest revenue (debit). reflect proper accounting procedures, the following adjust- This adjustment correctly states the true financial posi- ments may or may not be required to adhere to proper tion of the MFI. accounting procedures, depending on which country the N For example, if an MFI has accrued interest of 500 in MFI is operating in. However, adjustments for both subsi- the interest receivable account on a loan that has been dies and inflation should be made to accurately determine overdue for more than a year, it should decrease inter- the true financial viability of an MFI. Some readers may est revenue by 500 and decrease interest receivable by wish to make these adjustments separately from the adjust- 500 to correctly state its revenue and assets. ed financial statements described above. The best practice for MFIs is not to accrue interest The following describes why adjustments for subsi- revenue on delinquent loans at all. dies and inflation should be made and how to make actual accounting entries to the financial statements. ADJUSTING FOR ACCRUED INTEREST EXPENSE. When a fiscal (Adjusted financial statements are provided in appen- year ends, the balance sheet should reflect the financial dixes 1 and 2.) ADJUSTING FINANCIAL STATEMENTS 195 Accounting for Subsidies To make subsidy adjustments, it is necessary first to determine the value of the subsidy and then record Unlike traditional financial intermediaries that fund their accounting entries. For concessional loans and donated loans with voluntary savings and other debt, many MFIs equity, it is also necessary to determine the appropriate fund their loan portfolios (assets) primarily with donated "cost of funds" to apply to the subsidies. equity or concessional loans (liabilities). Concessional Adjustments for subsidies result in a change in the net loans, donated funds for operations (including indirect income reported on the income statement equal to the subsidies--such as reserve requirement exemptions or value of the subsidies; that is, subsidy adjustments result government assumption of loan losses or foreign in increased expenses, which lower the net income. exchange losses--or "in-kind" donations--such as free Subsidy adjustments do not, however, ultimately change office space, equipment, or training provided by govern- the totals on the balance sheet, because a new capital ments or donors), and donated equity are all considered account is created to reflect the increase in equity neces- subsidies for MFIs. sary to compensate for the effect of the subsidies, which Yaron (1992) shows that standard accounting mea- offsets the lower retained earnings. Audited financial sures of profitability are not valid for analyzing the per- statements may reflect the effect of direct donations formance of institutions receiving subsidies. Accounting (grants to cover operating expenses) to the MFI but profits are simply the sum of profits (or losses) and subsi- would not likely be adjusted for implicit subsidies (in- dies received. For example, if an MFI receives a grant to kind donations or concessional loans). cover operating shortfalls and records it as revenue, its net income will be overstated. In this case a highly subsi- DONATIONS TO COVER OPERATIONAL COSTS AND DONATIONS dized MFI appears more profitable than a better- IN KIND. Funds donated to cover operational costs are a performing, subsidy-free MFI. direct subsidy to the MFI. The value of the subsidy is, It is, therefore, necessary to separate out and adjust the therefore, equal to the amount donated to cover expenses financial statements for any subsidies to determine the incurred in the period reported. Some donations are pro- financial performance of an MFI as if it were operating vided to cover operating shortfalls over a period greater with market debt and equity rather than donor funds. In than one year. Only the amount "spent" in the year is addition, as an MFI matures, it will likely find it necessary recorded on the income statement as revenue. Any to replace grants and concessional loans with market rate amount still to be used in subsequent years remains as a debt (or equity) and would thus need to determine its liability on the balance sheet (referred to as deferred rev- financial viability if it were to borrow commercial funds. enue). This occurs because, theoretically, if an MFI To some extent, these adjustments also allow for com- stopped operations in the middle of a multiyear operat- parison among MFIs, because it puts all MFIs on equal ing grant, they would have to return the unused dona- ground analytically, as if they were all operating with tion to the donor. Therefore, the unused amount is commercially available third-party funds (Christen 1997). considered a liability. It is important to note that while the adjustments for Donated funds for operations should be reported on subsidies may ultimately be entered on the MFI's finan- the income statement separately from revenue generated cial statements, they do not represent actual cash out- by lending and investment activities to accurately report flows. Subsidy adjustments are calculated for analysis the earned revenue of the MFI. These funds should be purposes only to accurately assess the institution, its sub- deducted from revenue or net income prior to any finan- sidy dependence, and the risk that it would face if subsi- cial performance analysis, because they do not represent dies were eliminated. revenue earned from operations. (Note that any costs There are three types of subsidies typically received by incurred to obtain donor funds--"fundraising costs"-- MFIs: should also be separated from other operating expenses, I Funds donated to cover operational costs and dona- because the benefit of receiving the funds is not included.) tions in-kind In-kind donations such as free office space or staff I Concessional loans training constitute a subsidy to the MFI and should be I Donated equity. recorded as an expense on the income statement. This 196 MICROFINANCE HANDBOOK accurately reflects the true level of expenses that would incurred to collect deposits and the added cost of be incurred if the MFI were to operate without in-kind reserve requirements on deposits). donations. I Inflation rate plus 3 to 5 percentage points per year To make the adjustment, if an MFI has recorded dona- (particularly if the prime rate is negative in relation to tions to cover operational costs on the income statement as inflation). revenue earned, the amount should be reported below the MFIs may wish to select the same market rate for all net income line, resulting in a reduction in operational rev- types of concessional debt, or they may choose different enue and, therefore, a reduction in the amount transferred rates, depending on what form of funding might replace to the balance sheet as current year net surplus (deficit). An each subsidy. This must be decided by each MFI based offsetting credit entry is made to the balance sheet in the on the existing forms of debt and equity available and the accumulated capital--subsidies account (equity). ability to access different types of debt and equity. Similarly, the market value of donations in kind is Adjustments must be made to both the balance sheet debited on the income statement (an increase in expens- and the income statement. The amount of the subsidy is es) and offset in the same equity account on the balance entered as an increase to equity (credit) under the accu- sheet: accumulated capital--subsidies. mulated capital--subsidies account and as an increase in N For example, in the sample financial statements in financial costs (debit). appendix 1, a donation of 950 was received for opera- N For example, in the sample financial statements in tions (assume no donations in kind were received). appendix 1, concessional funds outstanding at the end The income statement is adjusted to reflect a reduc- of the year were 30,000. If we assume that the interest tion in profits of 950 (debit) and a credit of 950 to paid for the year on the 30,000 concessional loan was the accumulated capital account on the balance sheet 900 and the market rate was 10 percent, then the sub- is made. sidy would equal 2,100 [(30,000 * 10%) ­ 900]. The subsidy can also be determined by calculat- C ONCESSIONAL LOANS . Concessional loans are loans ing the average rate of interest paid on the concession- received by the MFI with lower than market rates of al loans. For example, if we assume that the average interest. Concessional loans result in a subsidy equal to interest rate paid on those funds was 3 percent and the value of the concessional loans times the commercial the market rate for commercial loans is 10 percent, or market rate, less the amount of interest paid. the annual subsidy is equal to (30,000 x 7%) or 2,100. Subsidy = [(concessional loans x market rate) ­ amount of interest paid] To make this adjustment on the balance sheet, 2,100 is entered in the capital accumulation account The actual financing costs of concessional loans are (credit), which increases equity. To offset this entry, netted out in this calculation because they are captured an adjustment has to be made to the income state- in the recorded financing costs on the income statement ment. The amount of the subsidy is entered as an (as an actual cash outflow or accrued interest payable). increase in financial costs (debit) of 2,100, which in To determine the appropriate market rate (cost of funds) turn reduces the profit of the MFI. When this to apply, it is best to choose the form of funding that the reduced profit is transferred to the balance sheet, the MFI would most reasonably be able to obtain if it were net equity changes result in no change to the totals on to replace donor funds with market funds. Suggested the balance sheet. market rates include: I Local prime rate for commercial loans (adjusted F UNDS DONATED FOR LOAN CAPITAL ( EQUITY ). Funds upward for a potential risk premium due to the per- donated for loan capital are often treated as equity by ceived risk of MFIs) MFIs and, therefore, are not always considered when I 90-day certificate of deposit rate (adjusted for risk) adjusting for subsidies, since these funds are not usually I Interbank lending rate (adjusted for risk) included on the income statement as revenue. (Note that I Average deposit rate at commercial banks (adjusted some MFIs do record donations for loan fund capital on upward to account for the additional operating costs the income statement, which then "flow through" to the ADJUSTING FINANCIAL STATEMENTS 197 balance sheet. If this is the case, an adjustment similar to The resulting figure is then recorded as an expense the one made for funds donated for operations should be (financing costs) on the income statement and an made.) increase is made to equity (credit) under the accumulated Funds donated for loan capital are reported on the capital--subsidies account. balance sheet as an increase in equity (sometimes referred For the purposes of this handbook, funds donated for to as loan fund capital) and an increase in assets (either as loan fund capital are treated as equity and are adjusted cash, loan portfolio outstanding, or investments, depend- for inflation only (therefore, no example is provided in ing on how the MFI chooses to use the funds). the appendix for a subsidy adjustment to funds donated Donations for loan fund capital differ from donations for loan capital). for operations in that the total amount received is meant to be used to fund assets rather than to cover expenses Accounting for Inflation incurred. Some analysts would suggest that, therefore, no subsidy adjustment needs to be made for funds donated Inflation is defined as a substantial rise in prices and vol- for loan capital. Furthermore, donors are not normally ume of money resulting in a decrease in the value of looking for any return on their funds (as a normal equity money. Goldschmidt and Yaron (1991) state that con- investor would), nor are they expecting to receive the ventional financial statements are based on the assump- funds back. What they are interested in is the ability of tion that the monetary unit is stable. Under inflationary the MFI to maintain the real value of the donated funds, conditions, however, the purchasing power of money relative to inflation, so that it can continue to provide declines, causing some figures of conventional financial access to credit for the target group. Therefore, the argu- statements to be distorted. Inflation affects the nonfinan- ment goes that donations for loan fund capital need only cial assets and the equity of an organization. Most liabili- be adjusted for inflation, not for subsidies. ties are not affected, because they are repaid in a devalued On the other hand, if an MFI did not receive dona- currency (which is usually factored into the interest rate tions for loan capital, it would have to either borrow set by the creditor). (However, variable-rate liabilities are (debt) or receive investor funds (equity). Both debt and nonmonetary and must be adjusted in the same way as equity have associated costs. The cost of debt is reflected fixed assets. This will not be covered here because it as financing costs. The cost of equity is normally the becomes quite complicated; for more information see return required by investors (either through dividends or Goldschmidt and Yaron 1991.) capital gains). Equity in formal financial institutions is This section describes how to account for inflation, often more "expensive" than debt, because the risk of loss considering its effect on both equity and the nonfinancial is higher and, therefore, the return demanded is higher. It assets of an MFI. (If donations for loan fund capital are can be argued, therefore, that MFIs receiving donations included in the subsidy adjustment, it is double-counting for loan fund capital should make a subsidy adjustment to to calculate an inflation adjustment for these funds as reflect the market cost of equity (or debt, depending on well. Adjusting other equity and nonfinancial assets for which one the MFI might use to replace donor funds). inflation may, however, still be required.) Hence, adjusting for the effects of funds donated for To adjust for inflation, two accounts must be consid- loan fund capital poses two possibilities: ered: I Adjust for the rate of inflation (inflation adjustment) I The revaluation of nonfinancial assets I Adjust using the market rate for commercial debt or I The cost of inflation on the real value of equity. equity (subsidy adjustment). Nonfinancial assets include fixed assets such as land, To adjust donations for loan fund capital for the buildings, and equipment. Fixed assets, particularly land effect of inflation, an adjustment is made to equity as dis- and buildings, are assumed to increase with inflation. cussed below. However, their increase is not usually recorded on an To reflect the subsidy on donations for loan fund capi- MFIs' financial statement. Thus their true value may be tal based on market rates, a similar calculation to the understated. adjustment for concessional loans is made. A market rate Since most MFIs fund their assets primarily with is chosen and simply multiplied by the amount donated. equity, equity must increase at a rate at least equal to the 198 MICROFINANCE HANDBOOK rate of inflation if the MFI is to continue funding its tion. As a guideline, International Accounting Standard 29 portfolio. (If MFIs acted as true financial intermediaries, suggests that adjustments should be made if the accumu- funding their loans with deposits or liabilities rather lated inflation over the three-year period exceeds 100 per- than equity, the adjustment for inflation would be lower cent (26 percent per year compounded) (Goldschmidt, because they would have a higher debt-equity ratio. Shashua, and Hillman 1986). However, interest on debt in inflationary economies will To calculate the cost of inflation, the following for- be higher, so the more leveraged an MFI the greater the mula is used: potential impact on its actual financing costs.) Most 1 ­ (1 + inflation rate)p assets of an MFI are financial (loan portfolio being the largest) therefore, their value decreases with inflation where p = the number of periods (generally years as (that is, MFIs receive loan payments in currency that is inflation rates are most often quoted as annual rates). worth less than when the loans were made). (In addi- N For example, an annual rate of inflation of 10 percent tion, the value of the loan amount decreases in real results in a percentage loss in real terms in the value of terms, meaning it has less purchasing power for the a loan portfolio over a period of three years of 23 per- client. To maintain purchasing power, the average loan cent [1 ­ (1 + 0.1)3 ]. This means that for the portfo- size must increase.) At the same time, the price of goods lio to maintain its real value, it would have to grow by and services reflected in an MFI's operating and finan- 23 percent over the three-year period. cial costs increase with inflation. Therefore, over time an MFI's costs increase and its financial assets, on which REVALUATION OF ASSETS. To adjust nonfinancial assets, revenue is earned, decrease in real terms. If assets do not the nominal value needs to be increased relative to the increase in real value commensurate with the increase in amount of inflation. To make this adjustment, an entry is costs, the MFI's revenue base will not be large enough to recorded as revenue (credit) (note that this is not recorded cover increased costs. as operating income because it is not derived from normal An MFI should adjust for inflation on an annual basis business operations) on the income statement and an based on the prevailing inflation rate during the year, increase in fixed assets (debit) is recorded on the balance regardless of whether the level of inflation is significant, sheet. The increased revenue results in a greater amount because the cumulative effect of inflation on the equity of net income transferred to the balance sheet, which in of an MFI can be substantial. turn keeps the balance sheet balanced although the totals Note that similar to subsidy adjustments, unless an have increased. (The increase on the asset side of the bal- MFI is operating in a hyperinflationary economy, adjust- ance sheet equals the increase in the equity, the result of ments for inflation are calculated for the purpose of higher revenue. This increase in fixed assets would also determining financial viability only. Inflation adjust- have an impact on the annual depreciation, which would ments do not represent actual cash outflows or cash increase and, therefore, reduce income. For simplicity's inflows. The choice of whether or not an MFI records sake, it is possible to ignore this factor.) inflation adjustments directly on their financial state- N For example, in the sample financial statements in ments must be made by its board. appendix 2, the MFI has 4,000 in fixed assets recorded Unlike adjustments for subsidies, adjustments for on the balance sheet. Assume that the annual rate of inflation result in changes to both the balance sheet inflation for 1994 was 10 percent. This means that the totals and the income statement. The balance sheet fixed assets should be revalued at 4,400 [4,000 + (4,000 changes because fixed assets are increased to reflect the x 10%)]. An increase in revenue of 400 is entered on effect of inflation, and a new capital account is created the income statement (credit) and an increase in assets to reflect the increase in nominal equity necessary to (debit) of 400 is recorded on the balance sheet. maintain the real value of equity. The income state- ment is affected by an increase in expenses relative to CALCULATION OF THE COST OF INFLATION ON EQUITY. To the cost of inflation. account for the devaluation of equity caused by inflation, Audited financial statements may or may not include the prior year's closing equity balance is multiplied by adjustments for inflation depending on the rate of infla- the current year's inflation rate. This is recorded as an ADJUSTING FINANCIAL STATEMENTS 199 operating expense on the income statement. An adjust- each year or divided by the consumer price index (or ment is then made to the balance sheet under the equity GDP deflator) for each year. To do this, conversion fac- reserve account "inflation adjustment." tors must be calculated. For example, if the inflation rate N For example, in the sample financial statements in is used, the current year (say, 1997) has a factor of one, appendix 2, the MFI had 33,000 in equity at the end and the conversion factor for the previous year is simply of 1993. With an annual rate of inflation of 10 per- 1 plus the current year's inflation rate (for a 1997 infla- cent, the revalued equity should be 36,300 [33,000 + tion rate of 20 percent, the conversion factor would be (33,000 x 10%)]. An increase in operating expense of 1.2 for 1996 financial data). For the year previous to 3,300 is entered on the income statement (debit) and that (1995), the conversion factor is the previous con- an increase in equity (credit) of 3,300 recorded on the version factor (1.2) multiplied by 1 plus the rate of infla- balance sheet in the inflation adjustment account. tion for the prior year (assume a 1996 inflation rate of 30 percent so the conversion factor is 1.2 x 1.3, or 1.56), and so on. (Note that the inflation rate for the Restating Financial Statements in Constant earliest year that is converted is not referenced, since the Currency Terms conversion factor for each year uses the inflation rate for the following year.) Some MFIs may want to restate their financial statements To use the consumer price index or GDP deflator, the in constant currency terms (adapted from Christen 1997). prior year's consumer price index is divided by the cur- This is not considered an adjustment in the same way that rent year's consumer price index to arrive at the conver- financial statements are adjusted for proper accounting. sion factor (to convert 1996 data, if the 1996 consumer Constant currency terms means that, on a year-by-year price index was 200 and the 1997 consumer price index basis, financial statements are continually restated to reflect was 174, then the conversion factor would be 1.15). the current value of the local currency relative to inflation. Unlike the inflation method, the price index method Converting prior-year nominal local currency values to always uses the current year's index divided by the con- present-year values permits year-to-year comparisons of verted year's index. the real (that is, inflation-adjusted) growth or decline in Finally, MFIs may want to convert their local-currency key accounts such as loan portfolio or operating expenses. financial data into U.S. dollars or another foreign currency This conversion does not affect the financial results of the for the benefit of their donors or to compare their perfor- MFI, because all accounts are converted and no new costs mance with MFIs in other countries. This is simply done or capital accounts are created. by multiplying (or dividing, depending on how the To convert prior year data, the amounts are multi- exchange rate is stated) the financial data by the current plied by either one plus the annual rate of inflation for year's exchange rate. 200 MICROFINANCE HANDBOOK Appendix 1. Sample Financial Statements cial position as at December 31, 1994, with adjustments Adjusted for Subsidies for subsidies (950 donation for operating costs and 2,100 for imputed financial costs on concessional The following balance sheet represents an MFI's finan- loans). SAMPLE BALANCE SHEET (adjusted for subsidies) as at December 31, 1994 1994 Adjust 1994A ASSETS Cash and bank current accounts 2,500 2,500 Interest-bearing deposits 7,000 7,000 Loans outstanding Current 50,000 50,000 Past due 29,500 29,500 Restructured 500 500 Loans outstanding (gross) 70,000 70,000 (Loan loss reserve) (5,000) (5,000) Net loans outstanding 65,000 65,000 Other current assets 1,000 1,000 TOTAL CURRENT ASSETS 75,500 75,500 Long-term investments 11,000 11,000 Property and equipment Cost 4,000 4,000 (Accumulated depreciation) (300) (300) Net property and equipment 3,700 3,700 TOTAL LONG-TERM ASSETS 14,700 14,700 TOTAL ASSETS 90,200 90,200 LIABILITIES Short-term borrowings (commercial rate) 12,000 12,000 Client savings 0 0 TOTAL CURRENT LIABILITIES 12,000 12,000 Long-term debt (commercial rate) 15,000 15,000 Long-term debt (concessional rate) 30,000 30,000 Restricted or deferred revenue 0 0 TOTAL LIABILITIES 57,000 57,000 EQUITY Loan fund capital 33,000 33,000 Accumulated capital--financial costs 0 2,100 2,100 Accumulated capital--donation 0 950 950 Retained net surplus (deficit) prior years 0 0 Current-year net surplus (deficit) 200 (2,850) (2,850) TOTAL EQUITY 33,200 33,200 TOTAL LIABILITIES AND EQUITY 90,200 90,200 Source for unadjusted statements: SEEP Network and Calmeadow 1995. ADJUSTING FINANCIAL STATEMENTS 201 SAMPLE INCOME STATEMENT (adjusted for subsidies) for the period ended December 31, 1994 1994 Adjustment 1994A FINANCIAL INCOME Interest on current and past due loans 12,000 12,000 Interest on restructured loans 50 50 Interest on investments 1,500 1,500 Loan fees and service charges 5,000 5,000 Late fees on loans 300 300 TOTAL FINANCIAL INCOME 18,850 18,850 FINANCIAL COSTS Interest on debt 3,500 3,500 Adjusted concessional debt 0 2,100 2,100 Interest paid on deposits 0 0 TOTAL FINANCIAL COSTS 3,500 5,600 GROSS FINANCIAL MARGIN 15,350 13,250 Provision for loan losses 3,000 3,000 NET FINANCIAL MARGIN 12,350 10,250 OPERATING EXPENSES Salaries and benefits 5,000 5,000 Administrative expenses 2,500 2,500 Occupancy expense 2,500 2,500 Travel 2,500 2,500 Depreciation 300 300 Other 300 300 TOTAL OPERATING EXPENSES 13,100 13,100 NET INCOME FROM OPERATIONS (750) (2,850) Grant revenue for operations 950 (950) 0 EXCESS OF INCOME OVER EXPENSES 200 (2,850) Source for unadjusted statements: SEEP Network and Calmeadow 1995. 202 MICROFINANCE HANDBOOK Appendix 2. Sample Financial Statements cial position as at December 31, 1994, with adjustments Adjusted for Inflation for inflation (400 for revaluation of nonfinancial assets; 3,300 for devaluation of equity). The following balance sheet represents an MFI's finan- SAMPLE BALANCE SHEET (adjusted for inflation) as at December 31, 1994 1994 Adjustment 1994A ASSETS Cash and bank current accounts 2,500 2,500 Interest-bearing deposits 7,000 7,000 Loans outstanding Current 50,000 50,000 Past due 29,500 29,500 Restructured 500 500 Loans outstanding (gross) 70,000 70,000 (Loan loss reserve) (5,000) (5,000) Net loans outstanding 65,000 65,000 Other current assets 1,000 1,000 TOTAL CURRENT ASSETS 75,500 75,500 Long-term investments 11,000 11,000 Property and equipment Cost 4,000 4,000 Revaluation of fixed assets 0 400 400 (Accumulated depreciation) (300) (300) Net property and equipment 3,700 4,100 TOTAL LONG-TERM ASSETS 14,700 15,100 TOTAL ASSETS 90,200 90,600 LIABILITIES Short-term borrowings (commercial rate) 12,000 12,000 Client Savings 0 0 TOTAL CURRENT LIABILITIES 12,000 12,000 Long-term debt (commercial rate) 15,000 15,000 Long-term debt (concessional rate) 30,000 30,000 Restricted and deferred revenue 0 0 TOTAL LIABILITIES 57,000 57,000 EQUITY Loan fund capital 33,000 33,000 Inflation adjustment - equity 0 3,300 3,300 Retained net surplus (deficit) prior years 0 0 Current-year net surplus (deficit) 200 (2,700) (2,700) TOTAL EQUITY 33,200 33,600 TOTAL LIABILITIES AND EQUITY 90,200 90,600 Source for unadjusted statements: SEEP Network and Calmeadow 1995. ADJUSTING FINANCIAL STATEMENTS 203 SAMPLE INCOME STATEMENT (adjusted for inflation) for the period ended December 31, 1994 1994 Adjustment 1994A INCOME Interest on current and past due loans 12,000 12,000 Interest on restructured loans 50 50 Interest on investments 1,500 1,500 Loan fees and service charges 5,000 5,000 Late fees on loans 300 300 Revaluation of fixed assets 0 400 400 TOTAL FINANCIAL INCOME 18,850 19,250 FINANCIAL COSTS Interest on debt 3,500 3,500 Interest paid on deposits 0 0 TOTAL FINANCIAL COSTS 3,500 3,500 GROSS FINANCIAL MARGIN 15,350 15,750 Provision for loan losses 3,000 3,000 NET FINANCIAL MARGIN 12,350 12,750 OPERATING EXPENSES Salaries and benefits 5,000 5,000 Administrative expenses 2,500 2,500 Occupancy expense 2,500 2,500 Travel 2,500 2,500 Depreciation 300 300 Other 300 300 Revaluation of equity 0 3,300 3,300 TOTAL OPERATING EXPENSES