The World Bank Group Guarantee Instruments 1990­2007 An Independent Evaluation 48454 THE WORLD BANK GROUP WORKING FOR A WORLD FREE OF POVERTY The World Bank Group consists of five institutions--the International Bank for Reconstruction and Development (IBRD), the International Finance Corporation (IFC), the International Development Association (IDA), the Multilateral Investment Guarantee Agency (MIGA), and the International Centre for the Settlement of Investment Disputes (ICSID). Its mission is to fight poverty for lasting results and to help people help themselves and their environment by providing resources, sharing knowledge, building capacity, and forging partnerships in the public and private sectors. THE INDEPENDENT EVALUATION GROUP IMPROVING DEVELOPMENT RESULTS THROUGH EXCELLENCE IN EVALUATION The Independent Evaluation Group (IEG) is an independent, three-part unit within the World Bank Group. IEG-World Bank is charged with evaluating the activities of the IBRD (the World Bank) and IDA, IEG-IFC focuses on assessment of IFC's work toward private sector development, and IEG-MIGA evaluates the contributions of MIGA guarantee projects and services. IEG reports directly to the Bank's Board of Directors through the Director-General, Evaluation. The goals of evaluation are to learn from experience, to provide an objective basis for assessing the results of the Bank Group's work, and to provide accountability in the achievement of its objectives. It also improves Bank Group work by identifying and disseminating the lessons learned from experience and by framing recommendations drawn from evaluation findings. The World Bank Group Guarantee Instruments 1990­2007 An Independent Evaluation http://www.worldbank.org/ieg 2009 http://www.ifc.org/ieg The World Bank http://www.miga.org/ieg Washington, D.C. ©2009 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved 1 2 3 4 5 12 11 10 09 This volume, except for the "Management Response" and the "Chairperson's Summary," is a product of the staff of the In- dependent Evaluation Group of the World Bank Group. The findings, interpretations, and conclusions expressed in this vol- ume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. This volume does not support any general inferences beyond the scope of this evaluation, including any references about the World Bank Group's past, current, or prospective overall performance. The World Bank Group does not guarantee the accuracy of the data included in this work. The boundaries, colors, de- nominations, and other information shown on any map in this work do not imply any judgement on the part of the World Bank Group concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encour- ages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Cover: Image representing global partnership by Sandy Vaccaro. © Images.com/Corbis. ISBN-13: 978-0-8213-7679-9 e-ISBN-13: 978-0-8213-7681-2 DOI: 10.1596/978-0-8213-7679-9 Library of Congress Cataloging-in-Publication data have been applied for. World Bank InfoShop Independent Evaluation Group E-mail: pic@worldbank.org Knowledge Programs and Evaluation Capacity Telephone: 202-458-5454 Development (IEGKE) Facsimile: 202-522-1500 E-mail: eline@worldbank.org Telephone: 202-458-4497 Facsimile: 202-522-3125 Printed on Recycled Paper Contents v Abbreviations vii Acknowledgments ix Foreword xi Executive Summary xvii Management Response xxv Chairperson's Summary: Committee on Development Effectiveness (CODE) xxxi Definitions 1 1 World Bank Group Guarantee Instruments 3 Patterns in Private Investments and the Need for Risk Mitigation 4 The Nature of the Instrument 6 Evolution of WBG Guarantee Instruments 10 Mapping of WBG Guarantee Instruments 16 Summary of Similarities and Differences 18 Conclusion 19 2 Review of the WBG's Experience with Guarantees 21 MIGA Guarantees 35 World Bank Guarantees 51 IFC Guarantees 62 Conclusion 63 3 The Delivery of PRM Products across the WBG 65 The Market for PRM 66 WBG Political Risk-Mitigation Products 69 Institutional Organization of the Delivery of PRM Products 73 Overlaps, Competition, and Market Niches 77 Cooperation on Projects--Efforts and Results 80 Coordination Mechanisms 85 Improving Delivery: Some Organizational Realignment Options 89 Conclusion 91 4 Recommendations 93 WBG Management 95 MIGA Management 95 Bank Management 96 IFC Management 97 Appendixes 99 A: Evaluation Methodology 101 B: WBG Guarantees Staff Survey i i i T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 109 Endnotes 113 Bibliography Boxes 10 1.1: WBG Guarantee Products 58 2.1: GTFP: A New Product in Response to Existing Demand 88 3.1: The 2005 Proposal to Develop Synergies between the World Bank and MIGA Guarantees Figures 11 1.1 Structure of IBRD/IDA, IFC, and MIGA Contractual Framework 22 2.1 MIGA's Gross Exposure, by Country Risk Classification, Fiscal 1990­2007 23 2.2 MIGA Guarantees Issued and Trends 24 2.3 Distribution of Guarantees Issued by Sector (gross exposure), Fiscal 1990­2007 24 2.4 Distribution of Guarantees by Sector (number of contracts) Fiscal 1990­2007 26 2.5 Reasons Given by Clients for Canceling a Guarantee, Based on Guarantee Contract Cancellations, Fiscal 2001­07 29 2.6 Snapshot of the PRI Market: Total New PRI Business Volumes, Calendar Years 2001­07 32 2.7 Comparing MIGA's Market Share with Shares of Other BU-IIC Members 35 2.8 Total World Bank Net Guarantee, by Country Risk Classification, Fiscal 1990­2007 36 2.9 Distribution of IBRD PCGs by Region (net commitment), Fiscal 1990­2007 36 2.10 Distribution of IBRD PCGs by Sector (net commitment), Fiscal 1990­2007 37 2.11 Distribution of IBRD/IDA PRGs by Region (net commitment), Fiscal 1990­2007 37 2.12 Distribution of IBRD/IDA PRGs by Sector (net commitment), Fiscal 1990­2007 47 2.13 PPI Projects in Developing Countries by Sector (1990­2006) 52 2.14 IFC's Total Net Commitment (excluding GTFP), by Country Risk Classification, Fiscal 1990­2007 52 2.15 Distribution of IFC Guarantees by Sector (net commitment), Fiscal 1990­2007 54 2.16 Distribution of IFC Guarantees by Region (net commitment), Fiscal 1990­2007 Tables 17 1.1 WBG Guarantee Instruments 31 2.1 MIGA Pricing of Guarantees (2007) 39 2.2 PCGs Enabled Public Agencies to Access Commercial Finance at More Favorable Terms 43 2.3 Guarantee Facilities Supported by the Bank 48 2.4 IBRD/IDA Pricing of Guarantees (2008) 53 2.5 IFC's Local Currency Financing 67 3.1 The WBG's Political Risk-Mitigation Products 68 3.2 Comparisons of the Deployment of WBG Risk-Mitigation Instruments 71 3.3 Potential Synergies of Expertise across the WBG 74 3.4 Comparison of the Main Attributes of the WBG Political Risk-Mitigation Products 78 3.5 Joint Bank-IFC-MIGA Guarantee Projects 84 3.6 Comparison of WBG Pricing 85 3.7 Benefits of a More Uniform Approach to PRM Pricing 86 3.8 A Simple Comparison among Organizational Options i v Abbreviations ADB Asian Development Bank AfDB African Development Bank ATI African Trade Insurance Agency BOAD Banque Ouest Africaine de Developpement CAS Country Assistance Strategy CELT Credit-enhanced lending transaction EBRD European Bank for Reconstruction and Development FDI Foreign direct investment GOLF Global Offshore Liquidity Facility (IFC) GRC Guarantee Review Committee GRIP Guaranteed Recovery of Investment Principal GTFP Global Trade Finance Program IBRD International Bank for Reconstruction and Development IDA International Development Association IEG Independent Evaluation Group IFC International Finance Corporation IPRC Implicit political risk cover MIC Middle-income country MIGA Multilateral Investment Guarantee Agency (of the World Bank) MSME Micro, small, and medium-size enterprise OPIC Overseas Private Investment Corporation (US) PBG Policy-Based Guarantee PCG Partial Credit Guarantee PPP Public-private partnership PRG Partial Risk Guarantee PRI Political risk insurance PRM Political risk mitigation RRG Rolling Reinstatable Guarantee RSF Risk-Sharing Facility SIP Small Investment Program (MIGA) SME Small- and medium-size enterprise WAEMU West African Economic and Monetary Union WBG World Bank Group v Complex infrastructure projects that carry a higher perception of risk have been beneficiaries of WBG guarantee instruments. Photo by Arne Hoel, World Bank Group. Acknowledgments This report was prepared by a team led jointly by IEG colleagues in all three institutions. The team Asita De Silva (Independent Evaluation Group also gratefully acknowledges the work of former [IEG]­World Bank), Ethel Tarazona IEG Director Aysegul Akin-Karasapan in the initial (IEG­Multilateral Investment Guarantee Agency concept stage of the report, as well as Amitava [MIGA]) and Stoyan Tenev (IEG­International Fi- Banerjee, Shahrokh Fardoust, and Ali Khadr dur- nance Corporation [IFC]), under the supervision ing preparation of the report. The report also of Marvin Taylor-Dormond (Director, IEG­IFC). benefited from consultations with a range of It draws on research and contributions from World Bank Group clients, development part- Abdul Akande, Unurjargal Demberel, Tanakorn ners, and external stakeholders. Peer reviews Makaew, Bertrand Marchais, Carlos Nunez, Carla were provided by Theodore H. Moran (Marcus Pazce, Aurora Medina Siy, Victoria Viray-Mendoza, Wallenberg Professor of International Business Ian Webb, and Stephan Wegner. Yvette Jarencio and Finance, Georgetown University) and Augusto and Rosemarie Pena provided general adminis- de la Torre (Chief Economist, Latin America and trative support to the study team. Heather Dit- the Caribbean Regional Office, World Bank). The tbrenner edited the report, and Sid Edelmann, report benefited from comments from an expert Sona Panajyan, and Shunalini Sarkar managed its external advisory panel composed of Bob Chest- publication and dissemination. nut (Project Director, Aldwych International), James A. Hanson (consultant, World Bank), and The team gratefully acknowledges the contribu- Marshall W. Meyer (Richard A. Sapp Professor of tions of many present and former staff at IFC, Management, The Wharton School, University of World Bank, and MIGA and also of a number of Pennsylvania). Director-General: Vinod Thomas Directors: Marvin Taylor-Dormond (IEG­IFC) Christine Wallich (IEG­MIGA) Cheryl W. Gray (IEG­World Bank) Joint Task Managers: Stoyan Tenev (IEG­IFC) Ethel Tarazona (IEG­MIGA) Asita De Silva (IEG­World Bank) v i i The overall goal of WBG guarantee instruments is to facilitate investment in high-risk projects and countries. Terraced rice paddies near a Red Zao village, outside of Sapa, Lao Cai province, in northern Vietnam. Photo by Tran Thi Hoa, courtesy of the World Bank Photo Library. Foreword Foreign direct investment and private capital flows entry in the market for local currency finance and are highly concentrated geographically, with al- have helped improve access to finance for un- most half of them reaching the top five destina- derserved market segments. tions. These flows tend to evade many high-risk countries, with the exception of those directed to At the same time, the evaluation finds important extractive industries. Regulatory and contractual weaknesses in the delivery of political risk miti- risks, particularly in infrastructure, have inhib- gation instruments that constrain their deploy- ited investments in many parts of the developing ment. A range of policy or mandate restrictions world. A core objective of the World Bank Group holds back the use of the instruments in specific (WBG) has been to support the flow of private in- situations. There has been competition among vestment for development; guarantees and in- WBG institutions for the same clients, imposing surance have been among the instruments that transaction costs on clients and reputational risks the Group has used to pursue this objective. on the WBG. Weaknesses in marketing efforts for MIGA and World Bank products contribute to This evaluation assesses the effectiveness in the use limiting client awareness of the products. Inade- of guarantee and insurance products by the WBG. quate internal awareness of the instruments in the It finds that these instruments have effectively ad- World Bank and IFC reduces the potential for vanced WBG strategic objectives, in particular fa- their wider use. Inconsistent pricing of the Bank's cilitating the flow of private investment to high-risk partial risk guarantee instrument entails risks of sectors and countries. The diverse range of these creating market and product distortions. instruments has helped to meet the demand for risk mitigation under a variety of circumstances. The challenge for the WBG is to create an envi- The Multilateral Investment Guarantee Agency, ronment in which guarantee and insurance prod- (MIGA), a relatively small institution of 100 staff, ucts can be deployed--alongside other WBG has issued $17 billion in guarantees and meets a instruments--in a flexible and efficient manner gap in the provision of political risk insurance to meet client needs. The report concludes that that private providers are unable to meet. The maintaining the status quo, particularly in the de- World Bank's partial risk guarantee has supported livery of political risk mitigation products, ought large and complex public-private partnership in- not to be an option. It suggests that WBG senior frastructure projects in high-risk countries. Its management should decide whether to take a partial credit guarantees have introduced countries set of collective and individually tailored actions to commercial markets or reintroduced them fol- within the current institutional structure or to lowing a crisis. The International Finance Corpo- adopt a new organizational structure for product ration's (IFC) guarantee instruments have led its delivery. Vinod Thomas Director-General, Evaluation i x MIGA guarantees cover projects for noncommercial risk, such as infrastructure projects that may be interrupted by a political upheaval. Photo by Suzanne Pelland, courtesy of MIGA. Executive Summary N otwithstanding the dramatic growth in private financial flows to de- veloping countries, there is significant untapped potential for greater involvement of the private sector in financing development-oriented investments. Foreign direct investment and private capital flows are highly con- centrated geographically: almost half of foreign direct investment goes to the top five destinations (World Bank 2007a). These flows bypass many high-risk countries, with tives, particularly in facilitating the flow of in- the exception of some that are directed to ex- vestment to high-risk sectors and countries. The tractive industries. Regulatory and contractual additionality--or unique contribution of these risks, particularly in infrastructure, have inhib- guarantee instruments--has derived from the ited investment and threaten the sustainability of WBG's relationship with host countries, its ca- economic growth in large portions of the devel- pacity to absorb risks that the private sector is un- oping world. A core task of the World Bank Group willing or unable to bear, and its focus on the (WBG) is to support the flow of private invest- objectives of poverty reduction and sustainable de- ment: guarantees have been among the instru- velopment. Guarantee instruments remain im- ments used to achieve this. portant for the WBG's priorities. This evaluation reviews the WBG's experience The answer to the other two questions is essen- with guarantee instruments during 1990­2007.1 tially no, especially regarding the delivery of prod- Although these instruments have been an estab- ucts for political risk mitigation (PRM). The use of lished product line of the WBG for two decades, guarantee products in each of the three institutions they have not been rigorously evaluated across the has fallen short of reasonable expectations be- World Bank, International Finance Corporation cause of external and internal factors. Consider- (IFC), and Multilateral Investment Guarantee ing external factors, demand projections appear Agency (MIGA). to have been overly optimistic. This is particu- larly true in infrastructure, where the rapid growth The study asks three main questions: (1) Should of the mid-1990s has not been sustained. In ad- the WBG be in the guarantee business? (2) Have dition, the supply of risk mitigation offerings by guarantee instruments in the three WBG institu- other private sector providers in emerging mar- tions been used to their potential as reflected in kets has also grown in both products and markets. WBG expectations and perceived demand? (3) Is the WBG appropriately organized to deliver its The internal factors for the negative answers in- range of guarantee products in an effective and clude (1) competition among institutions for the efficient manner? same clients and of the kind that often imposes additional transaction costs on clients and adds The answer provided by the evaluation to the reputational risks to the Bank; (2) weaknesses in first question is yes. Guarantees have been ef- the marketing of WBG guarantees and PRM prod- fective in promoting key WBG strategic objec- ucts that limit client awareness and choice; (3) a x i T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 range of supply-driven policy and mandate re- WBG senior management consider at least three strictions that inhibit the deployment of WBG alternative perspectives for organizational re- guarantee instruments in response to evolving alignment: the client perspective, the country client needs; (4) limited internal awareness, skills, perspective, and the product perspective. If the or incentives in the World Bank and IFC to use current organizational structure is maintained, guarantee instruments in relevant situations; and the study recommends that the management of (5) inconsistent pricing of the Bank Group PRM each individual WBG institution enhance the de- products, which runs the risk of creating market livery of its own guarantee/insurance products distortions and product differentiation among by implementing specific measures designed to WBG instruments based on price. These suggest improve policies and procedures, eliminate dis- opportunities to productively expand the use of incentives, increase flexibility, and strengthen guarantees. skills for the deployment of the products. To overcome the current limitations of the deliv- Should the World Bank Group ery system of WBG guarantees and PRM instru- Be in the Guarantee Business? ments and to enhance their use and development Guarantee instruments have been largely potential, the evaluation recommends that WBG effective in supporting WBG strategic ob- senior management take a strategic approach and jectives. Across the WBG, guarantees have effec- decide whether to maintain the existing organi- tively promoted private investment. Since 1990, zational structure, while addressing some of its im- MIGA has issued 897 guarantees for a total of $16.7 portant problems, or to develop and propose an billion. Its guarantees supported investment flows alternative organizational structure to the Board. across a broad range of high-risk sectors and coun- tries and for small and medium-size investments. Under any scenario, WBG senior management needs to take actions to introduce greater flexi- The World Bank has issued 25 guarantees for bility in the use of guarantee instruments in re- $3 billion. Although limited in number, its Par- sponse to dynamic country and client needs and tial Risk Guarantees have facilitated the flow of market developments by taking several actions: investment in large infrastructure projects in (1) revising existing policies and regulations on high-risk countries, particularly by enhancing guarantees to minimize supply-driven product the credibility of untested regulatory regimes. restrictions and allow product differentiation on The International Bank for Reconstruction and the basis of value added; (2) ensuring that ade- Development's Partial Credit Guarantees have quate incentives exist for staff to offer the full been used to introduce well-performing coun- array of WBG guarantees and PRM products to tries to markets or to regain access following private sector clients; (3) establishing more sys- crises. tematic links between advisory services and the deployment of Bank Group PRM instruments and IFC has approved 196 guarantee operations for $2.8 other products, particularly in infrastructure; billion, including through its Global Trade Finance (4) following a consistent approach to pricing of Program. More than 30 percent of its guarantees PRM across its guarantee instruments to avoid have been used to support trade and investment potential distortions; and (5) strengthening in- flows in Africa. They have also enhanced access to ternal awareness of the guarantee instruments financing for micro, small, and medium-size en- and the skills for their use and reducing transac- terprises in low-income countries. tion costs where possible, keeping in mind the im- portance of maintaining adequate processes and Guarantees account for 1.6 percent of the Inter- regulations for risk management and safeguards. national Bank for Reconstruction and Develop- ment's loan portfolio and 6 percent of IFC's. That If a new organizational structure is developed compares with 2 to 4 percent of the portfolios in and proposed, this evaluation recommends that other multilaterals examined in this study. x i i E X E C U T I V E S U M M A RY The WBG's additionality in risk mitigation A large number of developing countries and sec- derives from its relationship with govern- tors do not receive enough funding because of ments and its contribution to broader perceptions of high risks that the private sector development objectives. Each institution is unable to mitigate. In light of significant potential has issued a substantial proportion of its guar- demand, and given the importance of the in- antees in high-risk countries: 45 percent in MIGA, strument for the WBG's strategic priorities, vari- 46 percent in IFC, and 48 percent in the World ous strategies and policies have anticipated a Bank. significant increase in the deployment of guar- antees. Relative to these expectations and to per- The WBG's additionality in mitigating risk is ceived demand, the use of WBG guarantees has largely derived from its special relationship with been modest. governments, which enables it to absorb higher risks than private sector providers can take on. Some external factors explain limited de- MIGA remains the largest multilateral provider ployment. The WBG has had overly optimistic of traditional political risk insurance in the world expectations, particularly in the case of public- and has filled a gap in the market by providing private partnerships across a range of infrastruc- longer-term insurance in higher-risk countries. ture sectors based on rapid growth in the World Bank guarantees have helped further mid-1990s. These have not been realized in part both policy reforms and the environment for because of perceived high regulatory and con- private investment. IFC guarantees have sup- tractual risks. Some studies indicate that 65 per- ported financial innovation and capital market cent of investors self-insure rather than taking development by introducing new financial in- third-party insurance, suggesting a more limited struments to new classes of investors. More effective demand. Private sector providers of risk than half of IFC's commitments under the Global mitigation products have expanded their cover- Trade Finance Program were in Africa, which age in terms of both products and markets. Liq- has helped address a gap in the market for trade uid markets in the 2000s have reduced the finance. demand for sovereign Partial Credit Guarantees. Have WBG Guarantees Been Used to Internal factors have also constrained the de- Potential as Reflected in WBG ployment of instruments. MIGA's Convention Expectations and Perceived Demand? and Operational Regulations limit its adaptability Whereas guarantee instruments remain an to new market trends. MIGA has also not been important tool for supporting WBG strate- sufficiently aggressive in innovating within the gic priorities, the use of the instruments flexibility allowed by current policies. Internal has fallen short of WBG expectations to constraints to the deployment of Bank guarantees varying degrees. Several factors contribute to include the application of standards designed for the perception that there is significant unmet de- public sector operations to private sector projects; mand for WBG guarantee instruments: (1) Polit- a depletion of skills; and lack of both internal ical risk is consistently ranked as a main constraint and external promotion of the instrument. IFC has to the flow of foreign direct investment to devel- tended to apply a traditional project financier's ap- oping countries; (2) regulatory and contractual proach to guarantee-type instruments. It has risks are perceived as the main reason for the taken an overly conservative stance toward risk- growing investment gaps in infrastructure; and sharing facilities, which has constrained their (3) abundant liquidity in emerging markets calls utilization. not for external funding but for enhancements that can help deepen the market, extend maturities, Although some progress has been made in inno- lower spreads, and redirect resources to under- vation, there has been limited replication and scal- served market segments and new areas unfamil- ing up. Processing requirements in terms of iar to financiers in emerging markets. safeguards and risk management in each of the x i i i T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 three institutions have added value and must be tarity exist among the WBG PRM instruments, maintained and strengthened. But inefficiencies that which implies that there are opportunities for encumber processes need to be improved upon. cooperation and a need for coordination. Is the Bank Group Appropriately The three WBG institutions have cooperated on Organized to Deliver the Range of guarantee projects in several instances, but such Guarantee Products? cooperation has often been associated with high There is an overlap in the provision of PRM transaction costs. In selected cases, the three products within the WBG. The WBG's guar- WBG institutions have cooperated on the same antee instruments were designed to be comple- projects, particularly large, complex projects in IDA mentary, not competitive. However, a range of countries. At the same time, this cooperation has both guarantee and nonguarantee products over- often been associated with higher transaction lap in the provision of PRM for private sector costs, for both the institutions and the clients. In clients, and this overlap has tended to expand over one project, the participation of the Bank, MIGA, time. and a third multilateral insurance provider re- quired the project sponsor to enter into three con- Flexibility of policies and innovation in guarantee tractual agreements, each with different structures, and nonguarantee products have expanded the coverage, and mechanics. scope for competition. In addition, several nonguarantee IFC products offer PRM to the mar- The IDA-IFC program to enhance small and ket. IFC's lending and equity investments carry a medium-size enterprise access to finance in Africa, degree of implicit political risk cover, and its B-loan although achieving development outcomes, has program can mitigate transfer and convertibility been difficult to replicate, because it has been dif- risk through the umbrella of IFC's preferred cred- ficult to mobilize IDA resources for first loss pro- itor status. The PRM products of the three WBG visions. The need to fit these operations into institutions serve the same broad group of clients, existing IDA operations has also been a limitation, and there is evidence that these overlaps have and different procurement and conflict of inter- caused confusion among clients and internal com- est policies within the WBG have added to trans- petition of the kind that often imposes additional action costs. transaction costs on clients and adds reputational risks to the Bank. The establishment of the joint IFC-Bank Subna- tional Finance Department has institutionalized At the same time, each institution's products synergies between IFC and the Bank in subna- carry distinct attributes that help define tional finance. However, it has been difficult to market niches. The package of services ac- come up with mechanisms to reward Bank staff companying Bank/International Development for contributions to an IFC project. And despite Association (IDA) Partial Risk Guarantees makes attempts to resolve the pledge of shares and shar- them appropriate in situations of large, complex ing of arbitral awards issues related to the allo- public-private partnerships in untested regula- cation of collateral assets between lenders and tory environments and difficult business climates. insurers, they continue to hinder direct IFC-MIGA IFC's PRM can be packaged with a full set of fi- cooperation. nancial services and combined with commercial risk cover. MIGA products can offer flexible, self- Mechanisms to enhance coordination across standing political risk coverage that is least dis- the WBG have had varying degrees of ef- ruptive of the project financial structuring process fectiveness. More systematic consultations be- and that can match the need for longer-term tween MIGA and Bank country and sector cover and reach smaller investors. Thus relation- departments have helped ensure that MIGA- ships of both substitutability and complemen- supported projects are consistent with the Bank x i v E X E C U T I V E S U M M A RY Group's strategy in a country. But the principles develop and propose an alternative or- that govern the relationship between MIGA and ganizational structure to the Board. IFC products have been unclear. The "hierarchy of instrument" principle--which dictates that 2. Under any scenario, take action to in- Bank products be deployed last--has provided troduce greater flexibility in the use of some guidance, but implementation has been guarantee instruments in response to difficult. A Group-wide Guarantee Review Com- dynamic country and client needs and mittee had limited success in harmonizing ap- market developments by-- proaches and added to transaction costs. Its · Revising existing policies and regulations on functions have now been transferred to the Op- guarantees to minimize supply-driven prod- erations Committee of the Bank. In some cases, uct restrictions where most needed and to informal information sharing about business op- allow product differentiation on the basis of portunities has been effective in leading to actual value added guarantee projects. · Ensuring that adequate incentives exist for staff to offer the full array of WBG guaran- There is limited coordination within the tees and PRM products to private sector WBG in developing new products and at clients within a single menu of options the business development stage. Lack of staff · Establishing more systematic links between incentives, inadequate skills, and poor familiarity advisory services and the deployment of with the products of the other institutions have PRM instruments and other products, par- prevented better exploitation of downstream syn- ticularly in infrastructure ergies in marketing WBG products. Significant · Following a consistent approach to pricing potential exists for more systematic links between of PRM across its guarantee instruments to Bank-IFC advisory services and the use of WBG avoid potential distortions risk mitigation instruments, particularly in infra- · Strengthening internal awareness of the structure, keeping in mind the need to manage guarantee instruments and the incentives potential conflict of interest issues. The Bank and and skills for their use and reducing trans- MIGA apply different approaches to pricing of action costs where possible, keeping in political risk in their products for private sector mind the importance of maintaining ade- clients, which limits the effectiveness of coordi- quate processes and regulations for risk nation through the market. Products are not of- management. fered as a single menu of options to prospective private sector clients. In sum, opportunities exist 3. If a new organizational structure is de- for improvement, and maintaining the status quo should not be an option. veloped and proposed, consider at least three alternative perspectives for orga- Recommendations nizational realignment: the client per- To overcome the current limitations of the spective, the country perspective, and delivery system of WBG guarantees and the product perspective. PRM instruments and to enhance their use · Under the client approach, all products and development potential, the Independ- for private sector clients, including guar- ent Evaluation Group recommends the fol- antees and PRM instruments, would be of- lowing to WBG senior management: fered through a single window. · Under the country approach, the deploy- 1. Take a strategic approach and make a de- ment of WBG guarantee and PRM products cision whether to maintain the existing would be made according to country needs, organizational structure while address- under a management arrangement com- ing some of the important problems, or mon for all three institutions. x v T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 · Under the product approach, the bulk of hance the delivery of its own guarantee/ guarantee/insurance products would be insurance products by taking actions to managed under one institutional roof. improve policies and procedures, elim- inate disincentives, increase flexibility, 4. If the current organizational structure is and strengthen skills for the deployment maintained, direct the management of of the products. each individual WBG institution to en- x v i Management Response M anagement welcomes the opportunity to provide its revised response to the evaluation of the use of World Bank Group (WBG) guarantee instruments, prepared jointly by the three Independent Evaluation Group (IEG) units of the WBG: IEG­World Bank, IEG­International Finance Corporation (IFC), and IEG­Multilateral Investment Guarantee Agency (MIGA). Management agrees with IEG's conclusions that to all parts of the WBG, when most are relevant WBG guarantee instruments have evolved over the to one, sometimes two, but seldom all three in- last 17 years; they now have the ability to provide stitutions. As a result, management believes the re- support to clients across a large and diverse spec- port missed an opportunity to appropriately inform trum of investments under a variety of institutional the WBG on how it could build on the strength in structures, and they have played an important the difference in the mandates, client bases, and role in mobilizing private capital. Management guarantee products of each institution. also agrees with the thrust of several of the rec- ommendations for exploring ways to ensure that Differences in mandates of MIGA, IFC, guarantees are used as effectively as possible, in- and the Bank cluding improving coordination and marketing. Each WBG institution has a mandate that is de- Management notes that these improvements need fined under its charter. According to these man- to be carried out in a manner that ensures that dates, each member of the group serves the needs Bank Group risk mitigation instruments com- of its clients. The International Bank for Recon- plement--not supplant--private sector risk mit- struction and Development (IBRD) and the In- igation instruments and provide true additionality. ternational Development Association (IDA) clients are, first and foremost, member governments. Management Comments IBRD and IDA guarantees, which require a sov- ereign counter-guarantee, play a role that is dis- Management is pleased that the IEG report reflects tinctly different from that of MIGA insurance and management's view that guarantee products have IFC guarantees. This difference is evident in the proven to be useful instruments in helping the structure and use pattern of these products: WBG fulfill its mission. Recent experiences tell us IFC guarantees do not have a sovereign counter- that more could be done to optimize the use of guarantee and are used chiefly as unfunded loans guarantee products to meet WBG clients' demand. to support primarily private sector clients; MIGA provides mitigation of political risk for cross- Building on the strengths of WBG institutions border investments. Bank guarantees are mostly Although management shares the report's view used for complex projects in high-risk situations that there are opportunities to increase the ef- that require government commitment, with the fectiveness of the WBG guarantee products, it has ongoing Bank engagement with the country and concerns about the manner by which the evalua- the sector as primary factors in choosing this in- tion was conducted. Management believes that in strument. In fact, only a small fraction of these an attempt to generate overarching conclusions, guarantee products have a common risk mitiga- the report has "oversynthesized" issues as relevant tion function. It also points out that multilaterals x v i i T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 that have unified structures have an equal or full potential of IBRD and IDA guarantee products. smaller share of guarantees in their product mix The effort is aimed at removing constraints to than the WBG does. the effective deployment of guarantees, looking at issues of policy, organization, and incentives. Summary of Management Actions Even prior to the IEG report, management has More effective collaboration been working to identify opportunities for in- At the same time, work is under way on more ef- creasing synergies in the WBG guarantee products. fective collaboration, notably between MIGA and Subsequent to the Committee on Development the Bank and between MIGA and IFC, given the Effectiveness discussion of the IEG report in May very different nature of Bank and IFC clients. For 2008, management has been engaged in a process MIGA and the Bank, the focus is on joint organi- of further analysis and has identified two broad zation and underwriting of large, complex infra- areas for improvement: (1) a set of issues specific structure projects. For MIGA and IFC, the focus is to each institution and (2) more effective collab- on working out a claims cooperation framework oration among the Bank, IFC, and MIGA on joint and on piloting cross-marketing of products and solutions in support of ultimate clients (essentially services. governments in the case of the Bank and private sector entities in the case of MIGA and IFC). Man- Conclusions agement is now working to put these improve- Again, management welcomes this evaluation ments in place. and finds its analysis useful for ongoing work on improving effectiveness of the use of its guaran- Institutional issues tee and insurance tools. Management agrees with MIGA has embarked on a wide-ranging set of ini- and is acting on several of IEG's recommendations. tiatives to better serve its clients. The Bank has also Detailed responses to IEG's recommendations launched a focused effort toward realizing the are given in the Management Action Record. x v i i i M A N A G E M E N T R E S P O N S E Management Action Record Recommendations Management response To overcome the current limitations of the delivery sys- Partially agreed with regard to incentives and staff awareness, as noted below. tem of WBG guarantees and political risk mitigation (PRM) instruments and enhance its use and develop- ment potential, IEG recommends the following to WBG senior management: 1. Take a strategic approach and make a deci- As noted in the Management Response, the WBG institutions have complementary but sion whether to maintain the existing orga- differing mandates that are defined under their respective charters. Under these man- nizational structure while addressing some dates, each member of the group has developed different products serving the differ- of the important problems, or develop and ent needs of its clients. Management therefore believes that current institutional propose an alternative organizational struc- arrangements are adequate, and issues of coordination and marketing can be addressed ture to the Board. without a change in institutional structure. 2. Under any scenario, take action to introduce greater flexibility in the use of guarantee in- struments in response to dynamic country and client needs and market developments by-- · Revising existing policies and regulations on · The specific characteristics of each entity's products are governed by the entity's guarantees to minimize supply-driven product re- respective charters and policies, based on the clients it serves. Each institution strictions where most needed and to allow prod- continues to work to eliminate unnecessary restrictions, if any, in the use of its uct differentiation on the basis of value added. products. Management does not agree that an across-the-board revision cover- ing very different guarantee products offered by the WBG members is needed. More effective coordination would better address these concerns (see below). Management would also like to point out that IFC has no specific policies restricting the offering of partial risk guarantees within its institutional boundaries and calls on MIGA to provide political risk insurance as needed. In accordance with its in- ternal guidelines, IFC does not offer guarantee products that replicate the offerings of MIGA. · Ensuring that adequate incentives exist for staff · Management will assess the feasibility of increased staff incentives in the con- to offer the full array of WBG guarantees and text of potential benefits. PRM products to private sector clients within a IFC is prepared to work with MIGA and the Bank with regard to marketing the single menu of options. various PRM products through IFC's channels with clients. For example, there could be scope to leverage IFC's industry departments' relationships with key global private sector players to offer PRM instruments along with other financing op- tions as appropriate. As the report notes, the newly established Client Relationship Management System could also be a vehicle for coordinated marketing efforts. The incentives for MIGA or Bank staff to utilize these channels could be considered. · Establishing more systematic links between · The World Bank and MIGA will explore ways to establish more systematic links advisory services and the deployment of WBG between advisory services and their PRM instruments. For MIGA, as solely a PRM instruments and other products, particularly provider of political risk insurance and with no commercial interest, close col- in infrastructure. laboration with Bank policy advice is possible and indeed encouraged, as there is full alignment of interests between sound policy advice and such guarantees. The infrastructure area is in fact a good example of WBG coordination on advisory and financing services. IFC's investment team, which is separate from the advisory team, can offer a financing package to the winning bidder subject to satisfactory due diligence. Such a package could include WBG guarantee x i x T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Management Action Record (continued) Recommendations Management response products as appropriate. Advisory teams working with government clients will, as a matter of course, need to advise a government on the best options for ensuring a successful and competitive bid, for a concession, build-operate- transfer, or other structure. Those options may lead to recommendations that ei- ther indications of interest from potential financiers (IFC or others) or of avail- ability of political risk reduction mechanisms (WBG or others) be included in bidding information packages to increase the prospects of the government achieving its objectives. Governments are of course always free to reject such recommenda- tions. Given the possible appearance of conflicts of interest, potential conflicts arising from such recommendations are fully disclosed to clients and mitigating measures as per WBG Conflict of Interest policies are put into effect if the gov- ernments choose to follow such recommendations. WBG units may not be able to offer financing or guarantee products if the winning bidder does not turn out to be acceptable to them. · Following a consistent approach to pricing of · Each of the Bank group institutions uses pricing methodology that reflects the PRM across its guarantee instruments to avoid unique characteristics of it charter and its clients. Management will provide guid- potential distortions. ance to staff to ensure the consistency of Bank advice to governments with re- gard to the "hierarchy of instruments" principle and the fee charged by governments to the private sector to offset the costs associated with issuing a counter- guarantee. Such a fee will be considered the default option and application of the guidance will be monitored through the Finance, Economics, and Urban De- velopment Department (the Sustainable Development Network department that supports the Regions' guarantee work). · Strengthening internal awareness of the guar- · Efforts to increase awareness and ensure adequate skills in different special- antee instruments and the incentives and skills izations are ongoing. MIGA and the Bank will review their procedures to address for their use and reducing transaction costs any specific issues identified in the report that lead to higher transaction costs. where possible, keeping in mind the importance IFC is prepared to provide training to staff on MIGA and/or Bank guarantee of maintaining adequate processes and regu- products it may be asked to promote. Within IFC guarantee products, IFC's In- lations for risk management. vestment Guidelines and Practices provide a detailed description of each prod- uct with a note to inform IFC Treasury as soon as is practical of plans to offer a guarantee product. Within the Finance/Treasury Vice Presidential Unit, the Struc- tured and Securitized Products Department is the center of knowledge and prac- tice on guarantee products in IFC, providing structuring guidance to investment staff as needed. IFC will continue to cover guarantee products in relevant train- ing modules for staff. IFC is also mainstreaming a range of innovative financing techniques to investment departments to the degree possible. 3 If a new organizational structure is devel- No changes to the current institutional structure are envisioned. oped and proposed, consider at least three alternative perspectives for organizational realignment: the client perspective, the coun- try perspective, and the product perspective. · Under the client approach, all products for pri- vate sector clients, including guarantees and PRM instruments, would be offered through a single window. x x M A N A G E M E N T R E S P O N S E Management Action Record Recommendations Management response · Under the country approach, the deployment of WBG guarantee and PRM products would be made according to country needs, under a man- agement arrangement common for all the three institutions. · Under the product approach, the bulk of guar- antee/insurance products would be managed under one institutional roof. 4. If the current organizational structure is main- tained, direct management of each individual WBG institution to improve the delivery of its own guarantee/insurance products by-- MIGA management · Proposing to MIGA's shareholders amendments · MIGA agrees on the desirability of amending the Convention to provide greater to its Convention to remain relevant and meet flexibility in political risk insurance coverage that MIGA would be permitted to its market potential. provide. However, this would be a major undertaking and would require a strong consensus among shareholders on such amendments, and MIGA would proceed to a formal proposal only with such a consensus. · Considering, in the meantime, alleviating sev- · Agreed. An in-depth review is under way to examine what changes might be eral constraints derived from its operational warranted in the Operational Regulations and Policies but that would still be con- regulations and policies. sistent with the requirements of the MIGA Convention. · Increasing its responsiveness to market demand · Agreed. A Business Process Review is under way to examine what measures by addressing internal weaknesses that reduce can be taken to improve operational efficiency and responsiveness. efficiency and slow responsiveness without lowering MIGA's financial, social, or environ- mental standards. · Improving its client relationship management, · Agreed. The Business Process Review noted above is also addressing client including aftercare, to enhance the value MIGA relationship management. adds and increase its client retention. Bank management · Maintaining and promoting the partial credit · Partially agreed at this time. Extending access of the Partial Credit Guaran- guarantee instrument as a potential effective tee instrument to IDA countries would be an option that will be discussed under countercyclical tool to leverage government ac- the IDA Guarantees Review to be presented to the Board by December 2008. cess to commercial funds and extending such ac- cess to IDA countries. · Creating awareness among Bank staff of the po- · Agreed. Banking and Debt Management currently provides training on the tential use and benefits of partial risk guaran- political risk guarantee and other IBRD/IDA guarantee instruments for opera- tees and building necessary skills. tional staff, and also includes a discussion of the guarantees in its general train- ing for task team leaders. The Finance, Economic, and Urban Department (the Sustainable Development Network department that supports the Regions' guar- antee work) will continue to maintain adequate capacity to respond to demand from task teams for specialized guarantee expertise. x x i T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Management Action Record (continued) Recommendations Management response · Developing a marketing strategy that encom- · Partially agreed. The potential use of guarantees is most usefully discussed passes both governments and the private sector as a part of Country Assistance Strategy preparation, thus making the govern- to better identify situations in which the role of ments fully aware of the availability of guarantees. Management will work to a partial risk guarantee can make a difference. ensure that the potential use of guarantees is discussed as part of the prepara- tory Country Assistance Strategy discussions for all countries. In addition, as part of outreach programs for IBRD financial products, Banking and Debt Manage- ment routinely includes material on the availability and potential for IBRD guar- antees. Management plans to undertake a similar outreach program for IDA guarantees and is exploring various institutional options. · Streamlining processing steps to reduce both in- · Partially agreed. Management is exploring ways to streamline the processing ternal disincentives to working on partial risk of guarantees but will not commit before identifying specific measures. Compliance guarantees and transaction costs for private with Bank policies pertaining to environment and social safeguards would not sector clients while ensuring that crucial mea- be affected by any changes in policies related to guarantees. sures for social and environmental safeguards and risk management are maintained. IFC management · Mainstreaming its guarantee products through · Already the practice. IFC guarantees are essentially unfunded loans, enabling its operations departments in the same manner IFC to extend credit to a client, but without the associated funding inherent in a that its equity and loan products are deployed. loan. The basics of the appraisal and processing are identical between the two products; however, there are differences in structuring and documentation. Therefore, a member of the Structured and Securitized Products Department joins the investment team to assist with these functions for structured guarantee trans- actions. IFC has the same relationship with the client whether offering a guar- antee or loan, with the same investment staff in IFC involved, regardless of the product offered. In both cases, staff from IFC's Investment Operations do a thor- ough due diligence up front and perform a monitoring function throughout. The representations, warranties, covenants, and so forth contained in the docu- ments are similar. Therefore, there is not an overall issue regarding the processing of guarantees versus loans. However, there are some broader, related issues with respect to improving processing of innovative financial products at IFC that are important and that management is addressing. IFC has been developing many innovative financial products in recent years, including various types of structured finance, RSFs, local currency facilities, securitizations, and other structures, using many guarantees. An important challenge is to mainstream the use of these products to invest- ment departments as they become established, to reduce processing times and improve efficiency, and to allow groups such as the Structured and Securitized Products Department to remain focused on innovation and product development. This mainstreaming requires training investment department staff and developing replicable financial models and documentation. Management is undertaking this process now, including implementing a new organizational structure for the Struc- tured and Securitized Products Department in which product development is cen- tralized and business development decentralized. This is expected to better facilitate ongoing development of innovative products by a core staff in head- quarters, with a focus on mainstreaming these products to field-based invest- ment staff as the products mature. x x i i M A N A G E M E N T R E S P O N S E Management Action Record Recommendations Management response · Assessing the extent to which it can bring its · Not agreed. Because guarantees are simply unfunded loans, to the degree that guarantee products closer to meeting Basel II loans are Basel II efficient, so are guarantees. As there is no underlying discus- and regulatory requirements in general, so that sion in this report on what issues specific to IFC guarantee products pertain to the guarantee beneficiaries can use IFC products Basel II, IFC cannot assess this recommendation. more effectively for capital, provisioning, and ex- posure relief. · Revisiting its approach to RSFs to increase flex- · Agreed with respect to simplifying the process. IFC's position with respect ibility and improve the attractiveness of the to first loss in risk-sharing facilities (RSFs) is not about willingness to take risk product. but about the appropriateness of the project structure. It is well understood in the market at this time that the decoupling of origination and risk is what led to the subprime crisis. Fortunately, IFC understood this risk early on and has insisted that clients who originate loans retain a first-loss position in the portfolio to en- sure an alignment of interest. Whether an off-balance sheet securitization or an on-balance sheet RSF, the approach to portfolio risk is the same. However, there are some broader issues with respect to the implementation and replication of RSFs in a timely and efficient manner that management is ad- dressing. As management develops more of these structures, it is becoming more efficient with respect to processing and documentation, and management is now working to simplify and standardize these structures to the degree pos- sible. IFC currently has more than 20 RSFs in the pipeline with a medium to high probability of closing, many in IDA countries. In addition, management continues to work with counterparties such as IDA to share in the first loss alongside clients in instances in which such participation enables a deal to happen that oth- erwise would not. · Scaling up successful models in energy effi- · Agreed on scaling up in these sectors; not agreed on limiting scope to ciency, education, and capital market develop- guarantee products. Management agrees that IFC continues to scale up suc- ment based on the use of guarantee structures. cessful models in these areas. However, guarantees are just a subset of the full range of financial products IFC offers, and scaling up is not necessarily de- pendent on the use of guarantees. As discussed in a recent planning document, IFC is pursing programmatic approaches as a way to increase the development impact and additionality of operations by extending IFC activities beyond the in- dividual project into a program of projects and advice. A key approach to this is through wholesaling, where IFC is combining its financial sector and industry ex- pertise to enable the wholesaling of IFC financial products for specific industries through local banks. This will allow IFC to reach smaller clients in smaller, harder-to-reach countries, in many cases IDA countries. The education sector is one of the major areas in which this is being done. In energy efficiency, IFC has been a leader in the development of financing programs through financial in- termediaries designed to deliver environmental benefits, including clean energy. Management expects to scale up in this area. Finally, a number of capital mar- ket projects that include guarantees are being explored that would address such needs as access to housing finance, trade finance, and agribusiness, and which will be facilitated using short-term finance, local currency financing, and risk- sharing products. x x i i i MIGA's guarantees supported the development of Mozambique's Temane and Pande gas fields and the construction of an 865-km cross-border gas pipeline in Mozambique and South Africa. Photo courtesy of Sasol, a project guaranteed by MIGA and the first cross-border initiative in Sub-Saharan Africa. Chairperson's Summary: Committee on Development Effectiveness (CODE) O n May 7, 2008, the Committee considered The World Bank Group Guarantee Instruments 1990­2007: An Independent Evaluation and the draft management response. Comments by the external reviewers were circulated to the Committee as background information. Background as incentive issues in the World Bank and IFC with The Board of Executive Directors considered sev- respect to guarantees; and (5) inconsistent pric- eral management reports on the World Bank ing of the WBG political risk mitigation products. Group's (WBG) guarantee activities including re- Third, the use of guarantee products of each WBG views in 1997 and 2000 and technical briefings in institution has fallen short of expectations be- 2005 and 2007. This Independent Evaluation cause of these weaknesses. Group (IEG) report is the first independent eval- uation of the WBG guarantee instruments, un- The IEG report's main recommendation is for dertaken at the request of the Board. WBG senior management to "take a strategic ap- proach and make a decision whether to maintain IEG's Main Findings and Recommendations the existing organizational structure while ad- The report assesses the effectiveness in the use of dressing some of the important problems, or de- the WBG guarantee instruments and the delivery velop and propose an alternative organizational system of guarantee products within the WBG. The structure to the Board." The Director-General, Director-General, Evaluations, highlighted three Evaluations, highlighted this recommendation key findings from the report in his opening state- for the Board's attention and emphasized that ment. First, guarantees have been effective in pro- maintaining the status quo, particularly in the moting key WBG strategic objectives, particularly delivery of political risk mitigation products, is not in facilitating the flow of investment to high-risk a tenable option and changes need to be made sectors. Second, several weaknesses are evident in with or without organizational adjustment. The the political risk mitigation activities, including IEG report recommends that if a new organiza- (1) competition among the WBG institutions that tional structure is proposed, management should imposes additional transactional costs on clients consider at least three distinct perspectives: client, and reputational risks to the WBG; (2) weaknesses country, and product. Regardless of any organi- in marketing; (3) supply-driven policies and re- zational change, the IEG report recommended strictions placed by the mandates of WBG insti- that WBG senior management take action to in- tutions; (4) limited staff awareness and skills, as well troduce greater flexibility in the use of guarantee x x v T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 instruments in response to dynamic country and was broad interest in this evaluation, and the need client needs and market developments. for more strategic thinking by senior management about the WBG approach to guarantee business in Draft Management Response the longer term was emphasized. In this regard, A joint Bank, International Finance Corporation there were diverse views on IEG's recommenda- (IFC), and Multilateral Investment Guarantee tion for the WBG senior management to "make a Agency (MIGA) response was presented, which decision whether to maintain the existing orga- noted that the IEG report provides analysis and rec- nizational structure while addressing some of the ommendations that may be drawn on to enhance important problems, or develop and propose an the effective use of WBG guarantees and insurance alternative organizational structure to the Board." products. Management agreed with the general The Committee remained neutral on the issue, nei- thrust of several IEG recommendations, such as ther recommending nor precluding organizational the need to improve coordination and marketing. change, which is a long-term matter to be con- However, it noted that in attempting to synthesize sidered by management. Members stressed the the issues and recommendations, the report may need for a more in-depth analysis by management have "oversynthesized" the findings at the WBG to determine the most appropriate way to address level, when most issues are not relevant across the the market challenges and the weaknesses iden- board. Management also opined that a greater tified in the IEG report, and for the effective de- focus on the differences in mandates of MIGA, livery and use of WBG guarantees. IFC, and the Bank (International Bank for Re- construction and Development [IBRD] and In- Several speakers underlined that maintaining the ternational Development Association [IDA]) and status quo was not an option and suggested that on the WBG's client-focused approach could have the Management Response could be deepened to offered a different perspective to the IEG evalua- include a review of the three perspectives for or- tion. In her opening statement, the MIGA Execu- ganizational realignment suggested by IEG. On the tive Vice President representing WBG management other hand, some members noted the difficulties summarized the distinct mandates and clients posed in organizational change, including impact served by each WBG institution. MIGA's mandate on staff morale and legal implications. Regarding is to foster development through the provision of the possibility of amending the MIGA Convention, political risk insurance for productive foreign in- the Committee reiterated the understanding of the vestment and to mitigate risks for cross-border April 30 CODE meeting that it is looking forward investments. IFC promotes private sector devel- to the management's assessment of the con- opment, and its guarantees are used chiefly as straints of the MIGA Convention before it could unfunded loans to private investors. IBRD and make any judgment on this issue. IDA serve client governments, and these guaran- tees require sovereign counter-guarantees. The Speakers stressed the importance of responding IBRD and IDA guarantees are more complicated to market changes, being client oriented, and instruments than the Bank's loans, credits, or promoting the full array of WBG guarantee prod- grants and are demanded mainly for complex ucts. In this regard, they noted a need for a higher projects in high-risk situations. Management em- level of collaboration among the WBG institu- phasized that the WBG as a whole can bring a tions, improved marketing, greater staff awareness rich diversity of instruments, which should be of the guarantee instruments, and appropriate used in a complementary fashion to support spe- staff incentives. Comments were also made on the cific development objectives and client needs. overlap and competition of political risk prod- ucts as raised by IEG and on the need to adjust Overall Conclusions and streamline internal policies and procedures, The Committee welcomed the timely discussion to consider the consistency of pricing policies of the high quality and comprehensive IEG eval- across the WBG, and to think about a "single win- uation of the WBG guarantee instruments. There dow" for guarantee products. x x v i C H A I R P E R S O N ' S S U M M A RY: C O M M I T T E E O N D E V E L O P M E N T E F F E C T I V E N E S S ( C O D E ) Next Steps of MIGA) but also said that management should The Committee supported the request made by not rule out the option for organizational change. some Executive Directors for a full Board dis- A few of them observed that changing the orga- cussion of this IEG report, together with the nizational structure will likely be even more dif- Management Response. The date of the Board dis- ficult than amending the MIGA Convention and cussion will be determined by the Corporate Sec- should be considered as a last resort. retariat after requisite consultation. Other speakers emphasized that maintaining the Management was asked to revise its response to status quo is not an option and urged management the IEG report for the Board discussion, taking to quickly review different alternatives. Likewise, into consideration the comments raised at the they requested management to revise and deepen meeting. A WBG legal opinion on the feasibility its response to the IEG report with more analysis of a "single window" for guarantees, as well as any on the issue of organizational change, including other legal questions that may arise during the the three options suggested by IEG (for exam- forthcoming Board discussion related to the WBG ple, client, country, or product approach). A few delivery of guarantees, should be prepared in of them also remarked that IEG could have been due course. The Committee Chairperson con- more specific in its recommendation for organi- cluded that, in accordance with the approved zational change. A member questioned the ap- IEG Disclosure Policy Statement, this evaluation propriateness of the IEG recommendation and report will be made publicly available following the noted that it was not clear from the IEG report Board discussion, unless the Executive Directors whether IEG found any fundamental deficiencies decide not to disclose. that could only be addressed through a change in organizational structure. In this regard, some The following main issues were raised at the members clarified that this IEG report should be meeting: considered a fact-finding paper as requested by the Executive Directors and that the idea of an orga- Organizational change nizational change for the WBG guarantee busi- Members considered the IEG report a good basis ness had been actually raised by the management for considering longer-term improvements in the itself in past meetings. The negative impact on staff effective use of WBG guarantees in a changing mar- morale in raising the issue of organizational change ket context, beyond the short- to medium-term was also noted by a few members. focus of MIGA Operations Directions. In this con- text, IEG's main recommendation calling on the Management noted that multilateral develop- WBG senior management to decide whether to ment banks that have unified structures but no maintain or suggest changes to the existing or- specialized guarantee institution do not have a ganizational structure elicited a range of com- larger share of guarantees in their product mix ments. Members and speakers noted the need for than the WBG, and the use of guarantees is not management to analyze and propose concrete necessarily greater at these other multilateral de- solutions, including a timetable, to address the is- velopment banks. Furthermore, none of the other sues raised by IEG (for example, overlaps, compe- multilateral development banks has had the tition, and pricing) and to streamline the delivery same success in extending political risk insur- of WBG guarantees, increase competitiveness of ance as MIGA. Accordingly, management was not guarantee products, and introduce more flexi- recommending a restructuring of the organi- bility to adapt to client needs. Several speakers zation but was focusing on strengthening WBG noted that the effective use and delivery of WBG collaboration. At the same time, management guarantees may be enhanced to a certain extent welcomed a Board discussion on guarantees, un- through organizational adjustments within each derlining that the issue of changing the organi- WBG institution (for example, internal policy zational structure required the involvement of changes or updating the Operations Regulations the President, senior management of WBG, and x x v i i T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 the Board. The MIGA General Counsel added greater flexibility in its products by overcoming that any changes in organizational structure some of the constraints through the revision of will probably necessitate an amendment to the its current Operational Regulations. MIGA Convention. Such an amendment would require the affirmative vote of 60 percent of the Cooperation within the WBG Governors exercising 80 percent of the total vot- Several speakers stressed the need for greater ing power of the institution. Noting that the IEG collaboration among the three institutions based report is expected to be disclosed, management on their comparative advantages, and strength- commented on the need for a good communi- ening the coherence of the products offered, cation strategy by the Board, particularly be- including their pricing. Some members echoed cause the Committee discussions focused on IEG's remark that overlaps and competition did organizational issues. It cautioned that the ex- not necessarily imply redundancies, but required ternal public may have different views on the better WBG coordination. One speaker noted matter, which may affect the options available that there was sufficient external market compe- to management and the Board. tition and that competition within the WBG should be minimized. They and others called for more co- Management concurred that much could be ordinated WBG efforts for marketing, increased done to strengthen the delivery of guarantees, staff knowledge of the guarantee products, and which are part of a portfolio of services offered appropriate staff incentives. to clients. It committed to consulting senior man- agement on how to deepen the Management Re- A speaker urged management of IFC and the sponse. IEG reiterated that its evaluation findings Bank to make better use of staff knowledge and indicate that substantial business opportunities skills available at MIGA. Regarding IFC's readi- are missed with the status quo, and so a more in- ness to work with MIGA and the Bank to market depth review discussion on the options for various political risk mitigation products, one strengthening the delivery and use of guarantee member cautioned about the need for staff ex- products in a rapidly changing market situation pertise in these products and the risk of conflict were merited. of interest. Another member suggested the pos- sibility of placing MIGA under the umbrella of Client-focused approach IFC, because IFC has a more extensive network In considering the options for enhancing the ef- for marketing and MIGA guarantees can comple- fective use and delivery of WBG guarantees, sev- ment IFC products. In this context, he also re- eral speakers urged WBG management to consider quested clarification about the issue of conflict of the client perspective option. They also encour- interest. Still another member cautioned against aged WBG management to seriously consider the this idea. There was a question raised about the possibility of a "single window" for guarantees. A limited number of joint Country Assistance Strate- member stressed that the issue was not whether gies that included MIGA. guarantee products were marginalized, but whether WBG clients' need for different products Management agreed on the need for more WBG (for example, loans, equity, guarantees) is being coordination. MIGA management noted that po- met. A speaker understood that if guarantees are litical risk insurance requires specialized skills, underutilized, it also meant that clients are un- which are available in MIGA, but integration of derserved. In their opening statement, IEG staff MIGA work with IFC can lead to conflicts of in- clarified that its recommendations do not ad- terest, especially when there are claims that vocate that guarantee products should take MIGA needs to mediate. The MIGA General Coun- precedence over various WBG instruments or sel noted that MIGA has a strict fiduciary duty em- propose a target level in use of WBG guarantee bodied in its Convention that does not allow products. Referring specifically to MIGA, another staff to take into consideration other interests out- member expressed support for introducing side the institution in the discharge of their duties. x x v i i i C H A I R P E R S O N ' S S U M M A RY: C O M M I T T E E O N D E V E L O P M E N T E F F E C T I V E N E S S ( C O D E ) The IFC Acting General Counsel pointed out that antees were incorporated into its outreach efforts the overall objective of increased WBG coordi- to offer customized financial solutions focusing nation is probably less problematic from IFC's on IBRD countries, as well as training of both perspective, and the IFC Articles of Agreement (es- clients and Bank staff on the different financial pecially Article I , Purpose) are more flexible in products. the sense that they expressly recognize that IFC's role is to complement the activities of the Bank. Bank management briefly described how guar- Jiayi Zou, Chairperson x x i x Vietnamese woman riding a bike. Photo by Suzanne Pelland, courtesy of MIGA. Definitions Acceleration Making payments due immediately in specified circumstances. Assignment The complete transfer of the rights under a contract to one party in that contract. Credit-default swap A financial contract under which an agent buys protection against credit risk for a periodic fee in return for a payment should there be a credit-default event. Credit insurance A form of guarantee against loss from default by debtors. Credit-Linked Credit guarantee conditional on an event not occurring. Guarantee Derivatives Financial contracts whose value derives from underlying securities prices, interest rates, foreign exchange rates, market indexes, or commodity prices. Direct debt substitute Credit enhancement guarantee that covers some or all principal and interest payments and that may be applied to, among other in- struments, loans, bond issues, commercial paper facilities, note is- suance facilities, revolving credits, and portfolios of credit. Direct Debt Substitute Guarantees are general guarantees of financial indebted- ness and function as a debt substitute. The most common example is the financial guarantee of indebtedness to domestic banks. Double default Both the obligor and the guarantor failing to meet their obligations. Financial guarantee A form of coverage in which the insurer guarantees the payment of interest and/or principal in connection with debt instruments is- sued by the insured. First-loss tranche A class of securities that ranks last in priority of payments. It is gen- erally structured as the most junior claim on the borrower or collateral assets, absorbing losses in a manner similar to equity capital. Fortuitous event Any occurrence or failure to occur that is or is assumed by the par- ties to be to a substantial extent beyond the control of either party. x x x i T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Full Credit Guarantee Unconditional guarantee of 100 percent of the principal in present value terms (with the coupon rate used as the discount factor) for all categories of risk. This is economically equivalent to a guarantee of all principal and interest payments on their due dates. Guarantee The agreement by a guarantor to assume the responsibility for the performance of an action or obligation of another person or entity and to compensate the beneficiary in the event of nonperformance. Guarantee for Credit enhancement guarantee in a nonlending situation, where Commercial the objective is to back up a client's performance in a commercial Operations transaction involving the provision of goods and services, such as guar- antees of bid or performance bonds (called standby letters of credit in the United States); a guarantee that facilitates commercial trans- actions between the associated parties. Insurance A practice or arrangement by which a company provides a guarantee of compensation for specified loss in return for payment of a premium. Nonhonoring of Failure of sovereign or subsovereign entities and some state-owned sovereign guarantees enterprises to satisfy direct debt obligations or guarantees. Novation Term used in contract and business law to describe the act of either replacing an obligation to perform with a new obligation or replac- ing a party to an agreement with a new party. In contrast to an as- signment, all parties to the original agreement must agree on a novation. Partial Credit Guarantee An unconditional guarantee of a portion of the principal and/or interest in present value terms for all categories of risks. Partial Risk Guarantee Conditional guarantee of 100 percent of principal in present value terms for specific categories of risk (such as devaluation, breach of off-take agreements, labor unrest, and technology failure). Partial Credit and Conditional guarantee of a portion of the principal and/or interest Partial Risk Guarantee in present value terms for specific categories of risks (for example, devaluation, off-take agreements, labor unrest, and technology failure). Performance Bond Guarantee of a bond issued by the client to guarantee satisfactory Guarantee completion of a project by a contractor. Reinstatable guarantee A guarantee (coverage) that can be restored after the client has failed to perform and the guarantee has been called if the client re- pays the guarantor within a specified number of days, or after the client has repaid the guarantor. x x x i i D E F I N I T I O N S Risk-sharing facility A facility that allows a client to sell a portion of the risk associated with a pool of assets. In this case, the assets typically remain on the client's balance sheet and the risk transfer comes from a partial guarantee provided by the guarantor. In general, the guarantee is avail- able for new assets to be originated by the client using agreed-upon underwriting criteria, but in certain situations it may also be used for assets that have already been originated. Rolling guarantees Guarantee of debt service payment(s) that moves or "rolls" to cover new debt service payment(s) on the client's timely payment of the previously guaranteed debt service payment(s). Securitization A form of financing that involves the pooling and true sale of finan- cial assets and issuance of securities that are repaid from the cash flows generated by such assets. Structured finance A broad term used to describe a sector of finance that was created to help transfer risk using complex legal and corporate entities; in- cludes securitization. Subrogation An accepted principle in insurance law that provides for the assign- ment of an existing claim from the guaranteed investor to the insurer. The insurer, as the subrogee, acquires the same rights the investors had. Subrogation in The guarantor's recovery of the claim from the client in local currency local currency after the guarantee has been called and the guarantor has disbursed the loan. Transfer restriction An action taken by a government to prevent conversion of local and currency currency to some form of foreign exchange. inconvertibility Underwriting The process of selecting risks and classifying these risks according to their degrees of insurability so that the appropriate rates may be as- signed; includes rejecting risks that do not qualify. x x x i i i Chapter 1 WBG support through guarantees aims to boost investor confidence and catalyze investment in complex infrastructure projects, such as this Bujagali hydropower project in Uganda, which was guaranteed by MIGA and the World Bank and financed by IFC. Photo courtesy of MIGA. World Bank Group Guarantee Instruments W orld Bank Group (WBG) guarantee instruments have the poten- tial to mobilize private sector financing for development. Since the 1980s, the WBG has been expanding the menu of guarantee in- struments, and it currently offers a range of products to member governments and the private sector. Patterns in Private Investments current account surpluses have contributed to and the Need for Risk Mitigation abundant liquidity in the domestic financial mar- Private capital flows to developing coun- kets of many of those countries. tries have increased dramatically, but ac- cess remains uneven. Private capital flows to Despite the growth in private finance to de- developing countries increased from $165 billion veloping countries, the consensus in the in 2001 to $647 billion in 2006, but these flows development community is that there is sig- have been highly concentrated in a few large nificant untapped potential for greater middle-income countries (MICs). Almost half of private sector involvement in meeting the the foreign direct investment (FDI) and bank need for development-oriented investments. flows have gone to the top five destinations. Al- In 2006, the United Nations Millennium Project though developing countries' access to interna- estimated the financing gap for the Millennium De- tional financial markets has improved since the velopment Goals at $73 billion a year for low- 1980s, it remains uneven. Sixty percent of all de- income countries and $10 billion a year for MICs. veloping countries (79 of 135) did not access the Needs are particularly large in infrastructure--es- external bond market between 1980 and 2006, and timated at about 5.5 percent of the annual gross only eight did so frequently. Most low-income domestic product of developing countries to main- countries lack ready access to private debt mar- tain projected gross domestic product growth lev- kets, and many continue to depend very heavily els, compared with about 3.7 percent of gross on concessional loans and grants to meet their fi- domestic product actually invested (World nancing needs (World Bank 2007a). Economic Forum 2006). Abundant liquidity in many emerging markets is accompanied by a In many emerging markets, domestic li- paucity of long-term capital and poor access to fi- quidity has increased in recent years. Driven nancial services for large segments of the business in part by the commodity boom, current account community. Access to finance is consistently ranked surpluses in developing countries reached a among the top constraints to doing business in de- record 3.1 percent of gross domestic product in veloping countries, particularly low-income coun- 2006. Since 2000, developing countries as a group tries. This constraint is especially strong for micro have been a consistent net exporter of capital, and and for small and medium-size enterprises (SMEs). 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Risk perceptions are a significant factor in ments, the WBG has a comparative advantage in explaining patterns of private investment. mitigating certain political risks, particularly reg- A recent survey of more than 600 executives of ulatory and contractual risks. The WBG is ex- multinational companies conducted by the Econ- pected to act not merely as a supplier of capital, omist Intelligence Unit revealed that political risk but also as a catalytic agent in stimulating invest- is "the main investment constraint" in emerging ment in developing countries. Its policies mandate markets. The survey also found that all forms of that, in working with the private sector, it needs political risk in emerging markets are seen as in- to limit its own participation to the minimum re- creasing over the next five years. quired to secure satisfactory financing from private risk-taking sources. In this context, guarantee in- Regulatory and contractual risks are particularly struments are an important tool in the Bank high in sectors such as infrastructure and oil, gas, Group's arsenal of instruments for the pursuit of and mining. Political risk is a major constraint for its development objectives. domestic investors as well. Political uncertainty and unpredictability of economic policies are consis- The Nature of the Instrument tently viewed by entrepreneurs in developing A guarantee is a financial instrument for the countries as a major constraint and contribute to transfer of risks. A guarantee is the agreement the reluctance of financiers in these countries to of a guarantor to assume the responsibility for the extend long-term financing to private borrowers. performance of an action or obligation of another Risks perceptions are also a factor in explaining person or entity. The guarantor agrees to com- the limited flows of financing to micro and SMEs pensate the beneficiary in the event of nonper- and to new sectors of the economy. formance. Guarantees fall into the broad category of risk-transfer instruments such as collateral, in- Guarantees and other risk-mitigating prod- surance, and derivatives. They are unfunded trans- ucts are viewed as particularly well suited actions and thereby distinct from some funded to unlocking the potential of private in- transactions such as direct loans or loan syndica- vestment to contribute further to develop- tions that may also serve to transfer risk. One way ment. It is estimated that in 2006, between $225 to characterize the instrument is to pinpoint some and $439 billion FDI and debt to developing coun- of its similarities and differences with these other tries was covered by some type of political risk mit- mechanisms for risk transfer. (See Definitions on igation (PRM) instrument. The market for political page xxxi for further related definitions.) risk insurance (PRI) has grown rapidly in recent years--PRI exposures of Berne Union members Notwithstanding technical differences be- grew by 47 percent between 2001 and 2005. Analy- tween insurance and guarantees, for the sis of market data indicates that about a third of purposes of this study the terms are used in- loans to developing countries were guaranteed in terchangeably, because they perform es- 2005 and that there are significant fluctuations in sentially the same functions. With respect to the share of guarantee loans from year to year. collateral, a guarantee provides third-party se- Many governments in developing and developed curity against a failure to perform a duty and is countries alike have established credit guarantee often referred to as external collateral.1 Like in- schemes to support bank lending to micro and surance, a guarantee provides protection against SMEs (Levitsky 1997). possible eventualities. It typically refers to com- pensation for nonperformance, whereas insur- The WBG has a critical role to play as a cat- ance refers to compensation for specified loss.2 alytic agent to stimulate investment in de- veloping countries, and guarantees are an When the loss arises from nonperformance of important tool in its arsenal of instruments contractual obligations, the two terms are often to support private investments. Given its sta- used interchangeably, as in the case of financial tus and special relationship with world govern- guarantees and credit insurance (Ernst and Young 4 W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 2005). Unlike guarantees, insurance is typically a creation of more of the underlying assets, which bilateral rather than a three-party contractual re- translates into more investments and trade. lationship. Insurance also tends to work through the asset rather than the liability side of the un- Guarantees, if appropriately structured, could re- derlying relationship, although there are excep- duce expected loss from certain transactions by tions and the two are economically equivalent. lowering the probability of default, the severity of loss, or both (see Fitch Ratings 2005). For exam- Insurance agreements tend to be more stan- ple, guarantees can provide liquidity or absorb a dardized than guarantee contracts and typically are certain level of loss on the underlying assets, thus part of pools of contracts big enough to allow the reducing the probability of default. Guarantees law of large numbers to operate (except for in- could reduce credit risks by helping avoid currency vestment insurance, where this is generally not or maturity mismatches. They could lower the risk possible). There may also be differences in how profile of investments by making the expected loss much control the parties involved have over the dependent on the joint probability of default of event that triggers the loss.3 Despite these tech- two independent credits (the credit of the bor- nical differences, insurance and guarantees share rower and of the guarantor in the case of a bank the same functions and are treated interchange- loan, for example), assuming the guarantor and ably in this report.4 the issuer are not overly correlated. This is referred to in the literature as the double-default effect A guarantee, like most financial instruments, (Heitfield and Barger 2003). can be viewed as a form of derivative. The "easiest and most traditional form of a credit Some experts have argued (Heitfield and Barger derivative is a guarantee," where the guarantor 2003) that double-default effects argue for lower (seller of protection) provides protection to the capital requirements for such transactions. How- beneficiary (buyer of protection) with respect to ever, recognizing that although the use of guar- the performance of a third (reference) party or antees and other risk management techniques obligor (Kothari 2007). reduces or transfers credit risk, it may simulta- neously increase other risks (see BIS 2006, para- There are key distinctions between a guarantee and graph 115), Basel II takes a conservative approach a derivative contract: First, a loss may not be re- and instructs against taking double-default ef- quired for payment to be made from the seller to fects into account, opting instead for the substi- the buyer of protection under a derivative contract; tution approach. Under that approach, the lower second, the relationship between the seller of risk of the guarantor (protection provider) is sub- protection and the obligor (reference party) is stituted for the exposure to the guaranteed entity typically more remote under a derivative (the ref- (BIS 2006). erence party may not even know about the exis- tence of the derivative contract). Derivatives, The use of guarantees also entails certain so- unlike guarantees, are typically concluded under cial costs.The involvement of a third party implies standardized master agreements and are tradable. potentially higher transaction costs. Other things being equal, the presence of a guarantor tends to Guarantee instruments can bring social gains. increase complexity and add to the fragmentation Guarantee instruments help to complete markets of financing. It may also increase the social cost of by allowing participants to isolate and trade certain investment by adding the guarantor's risk to those risks as distinct from other types of risks. This cre- of the entrepreneur and the lender (Keynes 1936, ates markets that lead to better allocation of risks p. 145), unless the guarantor is sufficiently strong in line with the preferences and comparative ad- to offer an exceptional margin of security. vantages of the different market players in as- sessing, managing, and bearing these risks. By In addition, guarantees may exacerbate ineffi- providing asset protection, guarantees stimulate the ciencies created by information asymmetries and 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 influence the severity of the adverse selection economic/political/social externalities in the coun- problem. For example, a lender may have less or tries involved. (3) Where there are information no incentive to screen and/or monitor a borrower asymmetries and problems of moral hazard. Many because the lender will not bear the full conse- private sector activities--particularly some fi- quences of his action or inaction in the presence nancial sector activities--are rife with informa- of a guarantee. But if the guarantor conducts its tion asymmetries and problems of moral hazard, own screening before agreeing to sell protection which WBG guarantee instruments can help re- on an exposure, and if it has the same access to pair. (4) Where a demonstration effect is needed information and screening technologies as the after a breakdown in public order. In countries re- lender does, then the problem of weakened lender covering from civil strife and/or humanitarian incentives to screen may not arise. crises, WBG guarantee instruments can play a key role in turning conditions around so that pri- Finally, a guarantee may introduce new incentive vate markets are able to function once again. problems by adding the moral hazard of the guar- antor and the guarantee beneficiaries (see Kiff, It is important to note, however, that public Michaud, and Mitchell 2002). Many of the standard institutions, including the WBG, have their features of guarantee contracts have emerged as own weaknesses, and their interventions may at attempts to address such issues. The guarantee times exacerbate rather than alleviate market instrument is therefore no magic bullet against imperfections. market imperfections. Evolution of WBG Guarantee Instruments Public sector institutions continue to have Tapping private initiative to promote growth, re- a role in mitigating risks through guarantee duce poverty, and help people improve their qual- instruments. The outcome from private actor ity of life has always been at the center of WBG transfer of risks could be suboptimal when there activities. The three WBG institutions share this are market failures, economic/social/political ex- objective, and all three use guarantee instruments ternalities, or public goods. In such circumstances, in pursuit of these goals. there is a role for public sector institutions to step in. Guarantees are one tool that can be deployed Multilateral Investment Guarantee Agency to correct for market failures and promote the The Multilateral Investment Guarantee realization of positive economic/social/political Agency (MIGA) was conceptualized in 1948, externalities or the supply of public goods. but it was not established until 1988. MIGA's origins go back to 1948, when Bank staff realized In this context, there are at least four areas in that to achieve the objective of promoting pri- which WBG guarantee instruments can fill gaps vate foreign investment, guarantees were needed that the private sector cannot overcome on its against noncommercial risks. There were sug- own: (1) Where there are failures in making cred- gestions in the early days of the Bank that private ible commitments, honoring pledges, enforcing investments could be stimulated if the Bank contracts. Here the deterrent effect of WBG in- adopted measures that included "the investment volvement, and the halo or comfort that this of a certain percentage of its earnings in an `in- gives to investors and lenders, could provide the surance fund' to guarantee foreign investments underlying rationale for these kinds of WBG guar- against risks such as nationalization without com- antees. (2) Where there is a need for provision of pensation, war, and restrictions on the conver- public goods in governance, transparency, and so- sion of currencies" (Shihata 1998, p. 41; for details cial and environmental performance. Here the of MIGA's establishment, see Shihata 1988). presence of WBG guarantee instruments could en- sure observance of international standards on It took nearly 40 years--from the 1948 confi- the part of private sector actors, with consequent dential report on the proposed plan for guaran- 6 W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S teeing investment against transfer risk and certain settlement technique, as opposed and comple- other risks through numerous reports, memos, mentary to formal means of dispute resolution, studies, draft articles of agreement, and work to arbitration, and conciliation provided by the In- overcome the skepticism of World Bank Board ternational Centre for Settlement of Investment members, management, and staff--for MIGA's Disputes. It was also envisioned that MIGA's guar- Convention to be opened for signature on Octo- antee products and technical and legal services ber 11, 1985. Finally, in April 1988, all require- would help strengthen international standards ments for member states to join were completed, of fair treatment of the rights of investors. and MIGA started business. The first contract was issued in 1990. MIGA's Convention was set up to comple- ment, not compete with, national invest- Low foreign private investment flows and po- ment insurance programs and private litical risks had underpinned the establish- insurers of political risk. National insurance ment of MIGA. After a period of diminished agencies tended to "act as an agent of the coun- interest, discussions of MIGA's establishment were tries sponsoring investments of their nationals revived in the early 1980s. FDI had remained at very abroad" and to exclude certain investors, countries, low levels compared with official development and projects from their eligibility criteria (Shihata assistance, averaging $19 billion annually in the first 1988, p. 18). Private insurers offered limited cov- half of the 1980s. Moreover, although cases had de- erage, short-term duration, and few players. clined in frequency since the 1970s, there were 42 expropriations and nationalizations in Asia, Africa, It was envisioned that MIGA would fill the gaps and Latin America during 1978­83. in the market, particularly for investments in host countries that were ineligible for coverage by The early 1980s also saw the twin debt crises that other programs and for investors in projects that strengthened the resolve of MIGA's founding gov- made them ineligible because of ownership, res- ernments that heavily indebted countries must rely idence, or sources of procurement. Moreover, on the private sector and foreign private invest- Article 21 of the MIGA Convention called for co- ment to avoid the debt trap (MIGA 1994, p. 1). In operation with other insurers to encourage them the mid-1980s, there were few private political risk to provide coverage of noncommercial risks in de- insurers, the types of coverage were limited, and veloping member countries on conditions simi- coverage was mostly short term. Whereas the lar to those applied by MIGA. Bank's Articles of Agreement allowed it to make guarantees directly for private loans against any In 1997 MIGA launched two mechanisms for type of risk, guarantees for equity investments coinsurance and reinsurance: the Cooperative were overlooked. Underwriting Program, designed to encourage private insurers to cover projects in developing MIGA's objective was to encourage the flow countries whose risks they might otherwise be of private resources among members for reluctant to assume, and a quota share treaty or productive purposes. MIGA's instruments com- whole portfolio reinsurance agreement to ex- prised both the issuance of guarantees against pand MIGA's per project coverage and country noncommercial risks and the carrying out of other coverage limits. activities (mediation services and technical assis- tance) needed to allow developing countries to MIGA's shareholder composition of both de- attract increased FDI. The MIGA Convention gave veloped and developing member countries the agency a specific mandate to use its facilities was intended to make it an honest broker to encourage the amicable settlement of disputes that could guide "all concerned parties toward a between investors and member countries. Me- common definition of fairness and equitable treat- diation was chosen as the most current amicable ment . . . and help avert disputes . . . or provide 7 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 a channel for impartial mediation and amicable set- egory II countries have requested a change in tlement when they did arise" (MIGA 1998, p. 6). status to Category I and vice versa. MIGA is the only WBG institution with a special mandate to mediate disputes--including those un- International Bank for Reconstruction and related to a MIGA contract. In disputes unrelated Development/International Development to a MIGA contract, the host government must ex- Association press its consent to participate in the mediation, Formal World Bank guarantees were not and the investor pays MIGA's out-of-pocket ex- used until the early 1990s. The provision of penses for the mission. guarantees has been part of the Bank's mandate since its incorporation. Article I of the 1945 Arti- MIGA was to also have a deterrence effect. cles of Agreement states, "The purposes of the MIGA's shareholder structure and its status as a Bank are . . . to promote private foreign invest- member of the WBG allowed it to be "particularly ment by means of guarantees or participations in effective in pursuing salvage" and thus to have a loans and other investments made by private in- deterrent effect, which was deemed highly valu- vestors." Until the early 1990s, however, formal able to investors in projects that were vulner- guarantees were not utilized as a means to en- able to changes in host government actions and courage commercial finance to member coun- policies.5 tries. Instead, several cofinancing programs were implemented. Efforts through the early 1980s en- Its membership in the WBG gives MIGA leverage couraged commercial banks to cofinance Bank- over host countries that may be sufficiently con- funded projects but did not specifically enhance cerned about their current and future guarantee the credit terms available to governments. In- and lending programs to maintain appropriate stead, cofinancing of Bank-funded projects was host country policies toward foreign investment. seen as giving commercial lenders an assurance MIGA's participation would potentially deter any that loan proceeds were being used for priority abrogation of promises entered into by the host purposes, enabling them to develop broader and country. This deterrence role would likely sur- longer-term perspectives on member countries. vive the transition from one host government to another and strengthen MIGA's role in mediation In 1983, to increase the volume of cofinanced and in claims recovery. operations as well as to enhance commercial bor- rowing terms for governments, the Bank estab- MIGA's insurance products have been lim- lished its B-loan program. Under this program, the ited by what is allowed under its Convention, Bank would either guarantee or participate directly but changes can be made within the current in commercial bank loans for public investment authorizing environment. A change in the projects that it was also financing directly. Direct Convention would require action by its Council Bank participation in B-loans was subsequently of Governors and ratification by the member dropped, however, following a 1988 review that states.6 However, in the short and medium term, raised concerns that the Bank would dilute its pre- there are policy changes that MIGA could propose ferred creditor status. Instead, under the Ex- that would remain within the authorizing envi- panded Co-financing Operations Program of 1989, ronment of the Convention. A special majority the guarantee aspects of the program were to be vote by the Board is required for such changes. enhanced. In the past, amendments to Schedules A and B The coverage of guarantee operations was (membership and subscriptions and election of eventually expanded to include borrowing officers) have been made in accordance with by the private sector in both International Article 59(b) of the Convention, with Board ap- Bank for Reconstruction and Development proval. The amendments have been effected to (IBRD) and International Development include new states in the schedules or when Cat- Association (IDA) countries. In 1991 the Ex- 8 W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S panded Co-financing Operations Program was ways that could not be met through direct lend- broadened to include guarantees to support ing. The policy emphasized the importance of private commercial financing for private sector guarantees in promoting the development of local projects. In 1994 the Expanded Co-financing Op- capital markets. It made a choice in favor of IFC erations Program was then replaced by the "main- providing full-risk coverage instead of PRGs. streamed" guarantee program, which divided guarantees into Partial Risk Guarantees (PRGs) to In response to the limited use of guarantees, support private sector investment in member IFC revised its guarantee policy in 1997. countries by mitigating political risk, and Partial The revision recognized that in the nine years Credit Guarantees (PCGs) to support govern- since the policy had been adopted, and in contrast ment commercial borrowing by enhancing credit with the rapid growth experienced by IFC's direct terms. investments, IFC's use of guarantees--although successful in individual cases--had been very lim- Initially, only IBRD and blend countries were el- ited in aggregate volume. The document ad- igible for both products. To broaden the eligibil- dressed two reasons for the limited use of the ity for guarantee instruments, in 1997 the Bank instrument: the inherent disadvantages of full- introduced an IBRD PRG "enclave" program, risk guarantees versus direct lending, resulting in which made export-earning projects in IDA-only higher all-in-cost for the client, and the policy re- countries eligible for IBRD PRGs. Shortly there- quirement that, in the event of a call, the guarantee after, the Bank introduced a pilot program of be converted into a loan denominated in one of IDA PRGs for use in IDA-only countries, with an IFC's standard lending currencies. initial exposure cap of $300 million (subsequently increased to $500 million). In 1999 a further The revised policy allowed for flexibility in front- instrument was added--the IBRD Policy-Based end fees but kept commitment fees unchanged. Guarantee (PBG)--that extended the Bank's ex- It also amended the policy to allow for greater flex- isting PCG instrument beyond projects to in- ibility for IFC's subrogation in the local currency clude sovereign commercial borrowing in support in the event of a call on a local currency guaran- of structural and social policy reforms. tee. The changes were not expected to lead to a quantum increase in the volume of IFC's full-risk International Finance Corporation guarantee business, but improved applicability The International Finance Corporation (IFC) was expected in projects with access to funding began providing guarantees in 1982 in re- in local currency. sponse to demand for local currency fi- nancing. IFC's first guarantees were issued in Although there have been no major policy response to client needs to have access to long- changes in guarantees since 1997, product term local currency funding. During the next five innovation to meet client needs has led to years, IFC approved a limited number of guar- expansion in the range of IFC's guarantee antee transactions, mostly for debt, but also for instruments. The major shift in the use of guar- some trade obligations and equity-related in- antees has been from single-credit guarantees to struments. An IFC guarantee instrument in the guarantees on portfolios of credits, such as the 1980s was the equity-related guarantee, which Risk-Sharing Facilities (RSFs). was discontinued in the late 1980s.7 In 1988, the year MIGA was established, IFC developed its IFC currently offers a variety of guarantee prod- first guarantee policy in response to the per- ucts. Its nomenclature of products includes dif- ceived growing demand for guarantees. The pol- ferent forms of direct debt substitutes, trade icy recognized that guarantees were special facilities, commercial operations, and the Global purpose instruments that give IFC the flexibility Trade Finance Program (GTFP). IFC has also in- to respond to a broad range of specific client troduced some partial-risk products, such as the needs and to fulfill its development objectives in Guaranteed Offshore Liquidity Facility (GOLF) 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Box 1.1: WBG Guarantee Products MIGA PRI. MIGA offers PRI coverage to foreign direct investors other risks to the extent they are covered by a contractual obliga- for any combination of the following political risks: transfer re- tion of a government entity, and noncompliance with an agreed dis- striction, expropriation, war and civil disturbance, and breach of pute resolution clause. PRGs can be provided in both IBRD and IDA contract. MIGA can insure direct equity, quasi-equity, nonequity countries and require a government counter-guarantee. direct, and other investments. To insure debt, however, it must have an equity link. MIGA guarantees cover new foreign- IFC Direct Debt Substitutes. IFC PCGs are a credit-enhance- currency-denominated investments, including "new" invest- ment mechanism for debt instruments (bonds and loans). It is an ments to existing investments, investments by private for-profit irrevocable promise by IFC to pay principal and/or interest up to and nonprofit organizations, and publicly owned investors and or- a predetermined amount, irrespective of the cause of the pay- ganizations that operate on a commercial basis. MIGA can cover ment default. It can be applied to a single credit or to a portfo- any freely usable currency, which may include local currency lio of credits. investments/loans. Under certain circumstances, MIGA can cover investments by local investors. IFC Commercial Operation. This provides credit enhancement guarantee in a nonlending situation where the objective is to World Bank PCGs. PCGs support government borrowing from back up a client's performance of its obligation in a commercial commercial lenders or government bond issues to finance pub- transaction that involves the provision of goods and services, such lic investment projects. They provide comprehensive cover as guarantees of bid or performance bonds (called standby let- against all risks. PBGs are a type of PCG that are not associated ters of credit in the United States). with specific public investment projects, and instead support agreed policy reforms. Both PCGs and PBGs are available only IFC GTFP. The GTFP supports trade transactions by offering con- to IBRD countries and require a government counter-guarantee. firming banks partial or full guarantees that cover payment risk on issuing banks in emerging markets. Guarantees issued under World Bank PRGs. PRGs cover commercial lenders for a private the GTFP cover import and export transactions and extend to both sector project against default arising from a government-owned en- political and commercial payment risks. tity failing to perform its obligations. PRGs can cover changes in law, failure to meet contractual payment obligations, expropriation and IFC GOLF. IFC's GOLF provides single risk coverage for transfer nationalization, currency transfer and convertibility, nonpayment of and convertibility risk. a termination amount, failure to issue licenses in a timely manner, Source: WBG. and Credit-Linked Guarantees. IFC and IBRD tional range of PCG instruments for single cred- have established a joint Subnational Finance its and portfolios has been augmented with some Department that can offer IFC guarantees to sub- PRGs in recent years. Box 1.1 presents a more de- national entities and parastatals.8 tailed list of products, and Figure 1.1 illustrates the contractual framework under each of the WBG's Mapping of WBG Guarantee Instruments instruments, using a public-private-partnership The WBG offers a wide range of guarantee (PPP) project as an example. instruments. MIGA offers PRI against the specific risks of transfer and convertibility, expropriation, Guarantee transactions have a number of war and civil disturbance, and breach of contract. common elements. The basic structure of guar- IBRD offers PCGs that support sovereign bor- antee transactions includes the following elements: rowing, and both IBRD and IDA offer PRGs that provide PRI for breach of contract as well as tra- · Guarantor (or insurer)--The entity that ditional political risks. IFC offers credit guarantees provides the guarantee; equivalent to the "in- of performance of private borrowers. Its tradi- surer" in MIGA's terminology 1 0 Figure 1.1: Structure of IBRD/IDA, IFC, and MIGA Contractual Framework (using PPP as example) IBRD partial credit guarantee/ IBRD/IDA partial risk guarantee/ traditional PCG structure traditional PRG structure (PPP) (public investment) Concession agreement · Ongoing obligations Member SOE/ Member · Upon termination, government Project country government country owes (among other things) "termination company agency payment" to project company (borrower) Project agreementa · Reps and warranties IBRD- Project agreement Indemnity IBRD- Indemnity · Covenants guaranteed guaranteed agreement · Reps and warranties loan/bonds agreement · Covenants loan IBRD guarantee agreement IBRD guarantee agreement Commercial World Bank Commercial · Guarantee of government ongoing World Bank · Guarantee of partial debt payment lenders (guarantor) lenders obligation and/or payment of (guarantor) (loan or bonds) obligation (beneficiaries) (beneficiaries) "termination payment" WORLD MIGA political risk insurance (PPP) IFC partial credit guarantee (PPP) Concession and other agreements Concession agreement Member (breach of contract risk covered Member Project BANK Project · Ongoing obligations country by MIGA guarantee contract) country company company · Upon termination, government owes (borrower) · Ongoing obligations (among other things) "termination · Upon breach, government owes payment" to project company GROUP Host country (among other things) "termination Host country payment" to project company approval (no MIGA- notice IFC- · Requires an arbitral award guaranteed Agreement to issue a guarantee objection basis) guaranteed debt/equity GUARANTEE · Reps and warranties loan/bonds · Covenants MIGA guarantee contract Lender and/or · Single or a combination of breach of equity Commercial MIGA contract, currency transfer restriction and investors IFC lenders convertibility, expropriation, and war and IFC guarantee agreement (insurer) (guarantee (guarantor) (beneficiaries) civil disturbance risk coverage of underlying holders/ INSTRUMENTS assets, retained earnings, dividends, and so beneficiaries) forth Source: World Bank and IEG adaptations. Note: IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency; PCG = partial credit guarantee; PPP = public- private partnership; PRG = partial-risk guarantee; SOE = state-owned enterprise. 11 a. If there is no implementing agency involved, the respective project agreement provisions are included in the indemnity agreement with the member country. T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 · Beneficiary (or guarantee holder)--The en- to the client. MIGA and the Bank can also insure tity in whose favor the guarantee is issued, public corporations that operate on a commercial such as a bank in the case of a debt obligation basis and issue guarantees to nonprofit organi- · Obligor--The party that undertakes to per- zations (MIGA Operational Regulations 2007b, form an obligation, typically through a con- §1.19). MIGA beneficiaries must be foreign in- tract; that is, a borrower in the case of a debt vestors (in certain cases, local investors) from a obligation MIGA member country, and the project must be · Guaranteed asset/obligation--The under- located in a Part II country. IFC and IDA/IBRD lying obligation that is the subject of the guar- can issue guarantees to foreign as well as local in- antee, such as a loan in the case of a loan vestors without restriction.9 guarantee; equivalent to a "guaranteed in- vestment" for MIGA Eligible obligors · Events covered by the guarantee--Events WBG guarantees can have both private and that may cause the obligor to fail to perform public entities as obligors. In the case of World the obligation and that can trigger a call on the Bank and IFC transactions, the third party on guarantee; also referred to as risks whose balance sheet the guaranteed obligation ap- · Guarantee percentage/extent of coverage/ pears as a liability is the obligor or "client." MIGA percentage of cover--Percentage of the un- does not typically have a direct contractual rela- derlying asset/liability being guaranteed/ tionship with the obligor, unlike IFC and World insured; in MIGA, this refers to the portion of Bank guarantee transactions. the loss to be paid by the agency to the guaran- tee holder in the event of a claim WBG guarantee clients are private institutions in · Calls and claims mechanism--The steps the the case of IFC PCGs, IBRD/IDA PRGs, and MIGA beneficiary must follow to be able to call the PRI, and public institutions in the case of IBRD guarantee. PCGs and the PCGs issued by the IFC­World Bank Subnational Finance Unit and MIGA PRI (public Guarantors institutions operating on a commercial basis). Pub- Each of the three WBG institutions can be lic institution clients can be national and subna- a guarantor. Each WBG institution may provide tional entities as well as parastatals and state-owned a guarantee as a sole participant or may coguar- enterprises. The World Bank adopts the approach antee a transaction with other guarantors. MIGA that the government is the primary client for both and IFC can also syndicate (or reinsure) guaran- PRGs and PCGs. In the case of PRGs, however, as tees with other guarantors to leverage under- it is the private company that typically pays the guar- writing capacity and diversify risk. antee fee and in whose name the guarantee is is- sued, this evaluation considers the client for a Eligible beneficiaries Bank PRG to be the private company on whose be- The beneficiaries of WBG guarantees are half the guarantee is issued. typically private entities, although there are provisions for public institutions. The ben- Eligible underlying obligation/assets eficiaries of guarantees issued by the WBG are typ- The menu of WBG guarantees can cover al- ically private institutions, most often financial most any asset or liability. The guarantee in- institutions, but also commercial firms (as in the struments have established eligibility criteria case of IFC's commercial operations or MIGA's eq- according to several characteristics of the un- uity insurance) or institutional and individual in- derlying assets. These include the identity of the vestors, as in bond enhancements. assets as defined by whether they are debt, equity, or quasi-equity; are a single asset or a portfolio of In certain situations IFC can issue guarantees to assets; are in local or foreign currency; originate government-owned financial institutions if the locally or from foreign sources; are existing or new result would give a significant financial advantage items; are of short-, medium-, or long-term ma- 1 2 W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S turity; or are a financial obligation or a nonfinan- project, and the guarantees are available up to cial contractual obligation (such as to deliver the maximum maturity for Bank loans for partic- goods or services or a management contract ular countries, which may be up to 20 years. arrangement). Eligible investments for MIGA coverage in- IFC can offer guarantees on a broad range clude direct equity investments, equity-type of asset and liability classes, with the ex- loans, nonequity direct investments, and ception of straight equity. IFC guarantees can debt linked to equity. Eligible investments10 for be provided for senior or subordinated loans; a MIGA guarantee include the following: (1) eq- senior or subordinated bonds; senior, mezzanine, uity investments,11 (2) equity-type loans and guar- or first-loss securitization tranches; and a single antees (for example, shareholder loans or loan credit or a portfolio of credits. IFC can also guar- guarantees), (3) nonequity direct investments,12 antee long-, medium-, or short-term instruments, and (4) other kinds of investments.13 MIGA guar- such as bills of exchange, promissory notes, or antees equity investments, long-term loans and contractual obligations, including bid or per- guarantees made by equity holders in the project formance bonds, delivery of carbon emission enterprise, and loans made by institutions with no credits, and so on. equity involvement, provided MIGA also insures participation by equity holders. Since its Guaranteed Recovery of Investment Prin- cipal (GRIP) program was discontinued in the In general, MIGA guarantees nonequity direct late 1980s, IFC has not provided guarantees for type interests where the returns to the investor straight equity or equity-like instruments, but it depend on the production, revenues, or profit of can provide enhancement to first-loss securitiza- the enterprise.14 Other investments can be cov- tion tranches, for example. In terms of currency, ered, "provided that the Board so approves by a IFC can guarantee both foreign and local cur- special majority [and] . . . may be eligible only if rency instruments. It can also provide guaran- they are related to a specific investment covered tees for greenfield, existing investments, and or to be covered by the Agency" (MIGA 2007b, restructurings. p. 5). Eligible investments may be in monetary form, or they may be any tangible or intangible as- IDA/IBRD guarantees are restricted to debt sets that have a monetary value, such as machin- obligations, although policy flexibility al- ery, patents, processes, techniques, technical lows for effective coverage of equity as well. services, managerial know-how, and trademarks.15 IBRD PCGs can guarantee commercial loans to Examples of this practice include coverage of the governments as well as government bond issues. management agreement of a major hotel for its PRGs can be used to guarantee commercial loans investments in Peru and Costa Rica and of an oil to private investors but not equity investments. In production­sharing agreement in Equatorial one case, however, IDA was able to effectively Guinea. guarantee an equity investor by stipulating that if the government failed to make a termination pay- MIGA can offer a guarantee in any currency ment to the project company, the company would that at the time of the decision on the is- be deemed to have made a loan to the government suance of the guarantee is freely usable. and it would be this loan that would be guaranteed This implies that a guarantee in the host country's by the Bank PRG. The case illustrates the poten- currency is allowed if this currency is freely usable. tial flexibility of the PRG instrument in accom- modating different types of assets and obligations MIGA can guarantee a single investment, but its by transforming them into a loan equivalent guarantees are restricted to investments that will following a government failure to fulfill an obli- be implemented after the registration of the ap- gation. The Bank can guarantee local currency­ plication for the guarantee (with the exception of denominated loans to support local lenders to a expansions, restructurings, or privatizations). 1 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 MIGA can guarantee investments with terms from ment) can also cover other risks, including de- 3 to 15 years, although some flexibility exists to valuations, market risk, or corporate misman- underwrite longer- (up to 20 years) or shorter- agement. That is, any risk can be covered if it is term guarantee contracts (MIGA 2007b, §2.04). reflected in a contractual obligation of a govern- ment entity (Delmon 2007). As such, PRGs have Events covered by guarantees also covered natural hazards that are otherwise IFC traditional guarantee instruments offer uninsurable. comprehensive full-risk protection. Full-risk guarantees provide cover against all risks, and in MIGA's guarantees cover traditional PRI and this sense they are unconditional guarantees. breach of contract. MIGA can guarantee a sin- IFC's traditional guarantees are unconditional: gle risk or any combination of specific political they cover nonperformance for any reason, risks. These include transfer restrictions and non- whether commercial or political, by an IFC client convertibility, expropriation and similar mea- or by a portfolio of assets.16 sures, war and civil disturbance, and breach of contract. Article 11(a)(iii) of the Convention al- IFC has occasionally provided PRGs, as in the lows MIGA to provide breach of contract cover- case of credit-linked guarantees, where certain age when (1) the holder does not have a recourse credit events are excluded from IFC's cover. Be- to a judicial or arbitral forum to determine the cause these have typically been very low proba- claim; (2) such decision is not rendered within a bility events (such as defaults by the sovereign on reasonable amount of time; or (3) such decision its local currency obligations), credit-linked guar- cannot be enforced. At present, only the third antees have been a way for IFC to provide as condition is authorized by the Board, which re- close a substitute for full credit guarantees as quires an arbitral award or a court decision. MIGA's possible without violating the Treasury's policy Convention allows other types of specific non- against provision of such guarantees (see below). commercial risks to be covered if the Board so ap- Another example of a PRG offered by IFC is the proves. The Convention specifically excludes GOLF, which covers transfer and convertibility devaluation or depreciation risks from MIGA's risks only. coverage. Bank PCGs cover all risks, and PRGs cover Extent of coverage traditional political risk and breach of con- A common feature of WBG guarantees is tract risks, which in some cases include com- the intent to limit the extent of coverage to mercial and natural force majeure risks. the minimum amount necessary "to achieve IBRD PCGs are also unconditional: they guarantee a successful transaction." Guarantees can government repayment of a commercial loan or cover either part or all of an asset's value. Full bond regardless of the cause of default. IBRD/IDA credit guarantees provide complete coverage of PRGs are triggered when a covered event causes all principal and interest due on a financial obli- a debt service default. Coverage can include war gation, or the entire amount of investment plus and civil disturbance; expropriation and national- some specified return in the case of equity in- ization; foreign currency transferability, availability, vestment. Full (or comprehensive) risk guarantees and convertibility; failure to meet contractual pay- cover all risks, both commercial and political. IFC ment obligations; nonpayment of a termination initially offered full credit guarantees, but moved amount or an arbitration award following a covered over the years to provide either full coverage for default; obstruction of an arbitration process; fail- PRGs or partial credit for full-risk guarantees. It ure to issue licenses, approvals, and consents in a does not now, as a rule, provide full credit com- timely manner; and changes in laws. prehensive guarantees. Guarantees that cover government contractual This policy decision was intended to enhance payment obligations (such as a purchase agree- the catalytic effect of guarantees, align incentives, 1 4 W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S and avoid the creation of a surrogate for IFC's risk nonpayment by the obligor, actions of notice, ev- in the market. However, there are no strong the- idence of efforts by the beneficiary to remedy oretical or practical reasons for an absolute rule the situation, or arbitration award in the case of against full credit guarantees,17 and in several in- disputes between the beneficiary and the obligor. stances IFC has provided or has moved close to By paying the claim, the guarantor generally ac- providing full credit guarantees, as in the case of quires the rights of the guarantee holder in the some credit-linked guarantees. form of a novated or subrogated loan or claim. IFC's GTFP offers full credit unconditional guar- An important concern for IFC is to ensure that the antees on trade instruments. The same stance loan or claim obtained in this fashion enjoys pre- against the provision of full credit guarantees is ferred creditor status. Such a concern creates a seen in IBRD/IDA guarantees. IFC and World Bank preference in IFC toward novation, or the use of PCGs can cover a percentage of the principal standby loans, instead of a straightforward guar- and/or interest or apply to only a specified period antee as far as documentation is concerned. of the loan maturity. Under late maturity guaran- tees, IFC and the Bank cover payments in the In MIGA's case, the guarantee holder assigns an later stages of the life of the instrument--say, be- existing claim against the host government to tween years 10 and 15 in the event of a debt ser- MIGA, which is subrogated to the rights of the in- vice default on a loan with a final maturity of 15 vestor upon payment of compensation. As a sub- years. Any default before that time would be the rogee, MIGA acquires the same right that the lender's risk. MIGA can cover amounts of up to investor had. MIGA can--and does--negotiate 99 percent for loans and 95 percent for all other before paying compensation, which is one of the types of investments for political risk only. reasons for the waiting period. If the government refuses to negotiate and subrogation fails, MIGA Bank and IFC PCGs can be rolling and re- has direct recovery rights and a system to activate instatable. Both IFC and IBRD have used rolling them under the MIGA Convention. and reinstatable PCGs. Rolling guarantees are PCGs of debt service payments that move, or Until 1997, IFC did not allow subrogation in the "roll," to cover new debt service payments on local currency if there was a call on a local currency the client's timely provision of the guaranteed guarantee. Flexibility was introduced in the 1997 debt service payment. The rolling period can amendment of the guarantee policy, with re- cover all or part of the obligation. In reinstatable strictions on exposure to countries with under- guarantees, after the client has failed to perform developed derivative markets. A guarantee is and the guarantee has been called, it can be re- accelerable if payments due under the underlying instated (that is, coverage can be restored) if the obligation can be made to fall due immediately in client repays the guarantor within a specified specified circumstances. In virtually all cases, IFC amount of time. IFC has offered rolling reinstat- retains the right to accelerate the guarantee (ac- able guarantees. The IBRD has used the instru- celeration is defined as a full, immediate payout). ment in Argentina, Thailand, and Colombia, IBRD enclave guarantees and IDA guarantees are although policy guidelines now restrict the use of generally not accelerable. reinstatable guarantees, as discussed below. Although the Bank subrogates investors' Calls and claims procedures rights, it also acquires a claim on the gov- When paying a claim, IFC and MIGA get the ernment through a counter-guarantee. For investors' rights through assignment, no- all guarantees, the Bank enters into an indemnity vation, or subrogation as main vehicles for agreement with the member country in which the recovery. All guarantees specify conditions under project is located. Under this agreement, the which the guarantee may be called. Such condi- Bank is counter-guaranteed by the member coun- tions may include a minimum number of days of try. Once the Bank has paid out the amount owed 1 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 under the guarantee, it has the right to demand The ease of obtaining a government counter- immediate repayment of that amount from the guarantee for private transactions can vary, de- member country or schedule normal IBRD/IDA re- pending on the government's relationship with payments terms. the Bank, the strategic importance of the project, and the extent to which the counter-guarantee Unlike the IBRD, IDA's Articles do not require that may make explicit some implied government ob- it obtain a government guarantee. But as a busi- ligations embedded in its contract with the private ness practice, Bank policies require IDA to obtain firm. MIGA requires host country approval for a counter-guarantee from the host government. As both the issuance of MIGA's guarantee and the with IFC and MIGA, the Bank is also subrogated risks to be covered. Host country approval is to the rights of the beneficiary of its guarantee fol- sought at two levels: first, whether the proposed lowing a payment on a claim, which can provide investment conforms to the host country gov- recourse if the government fails to discharge its ob- ernment's laws, regulations, and declared devel- ligation under the indemnity agreement. opment objectives, and second, approval of MIGA's issuance of the guarantee against the risk Summary of Similarities and Differences designated for coverage. Should the host coun- The WBG offers a diverse set of guarantee try limit its approval to a certain type of risk, products (table 1.1). All guarantees enable the MIGA will have to reflect that limitation in its con- sharing of risk through unfunded transactions. tract of guarantee. Private entities can benefit from all WBG guaran- tee products and--with the exception of IBRD's MIGA, the only WBG institution established PCGs and the subnational finance PCGs--all WBG to cover political risk, has the most un- institutions have private entities as clients (or bundled and narrowly defined product space obligors). IFC PCGs and IBRD PCGs are full-risk of all WBG institutions. Because MIGA was guarantees, and as such they cover defaults caused the last of the WBG institutions to be established, by all political and commercial risks. All the other and given the prevailing economic environment WBG guarantees are PRGs. of the time, as well as a desire among the founders to ensure complementarity and avoid overlaps By type of underlying assets, all three institutions with national insurers, MIGA was given a specific can guarantee debt, although MIGA has some re- role and a narrowly defined product. As a result, strictions for nonshareholder loans. With minor MIGA's PRI product is subject to the most eligibility exceptions, such as IFC's Sovereign-Linked Credit limitations among the guarantee instruments of Guarantee, which can specifically exclude cer- the WBG and can guarantee against four defined tain types of sovereign/political risks, all WBG types of political risks. guarantees provide PRM (this is further exam- ined in chapter 3). MIGA's Convention does not allow it to cover commercial risks because it was designed to "pro- The unique features of IBRD/IDA guaran- vide guarantee services to investors in projects of tees are that governments are clients in the acute vulnerability to changes in host govern- case of PCGs and that sovereign counter- ment policies or commitments" to avoid compe- guarantees are required for all Bank guar- tition with national export credit agencies. It is antees. A sovereign counter-guarantee is required generally unable to guarantee against breach of for all IBRD guarantees by the IBRD Articles of contract by state-owned enterprises that operate Agreement, and IDA has adopted the practice as on a commercial basis because of its inability a principle of business prudence. Clients may to hold the governing authority liable for the value the presence of a counter-guarantee as a obligations of the enterprise. MIGA's most dis- further signal of the government's commitment or tinctive feature is its ability to insure equity, quasi- as indication of the Bank's ability to influence the equity, and other forms of nondirect investments, government or deter adverse government actions. although the Bank has synthetically replicated 1 6 W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S Table 1.1: WBG Guarantee Instruments IFC IFC IFC IFC IFC/Bank IFC IBRD IBRD IDA MIGA Product attributes PCG SCLG GOLF RSF SNF PCG COP* PCG PRG PRG PRI Risks covered Comprehensive risk + + + + + Commercial risks only + + + + + + Political risks + + + + + Transfer and convertibility + + + + + + + + + + Breach of contract + + + + + + + + + Expropriation + + + + + + + + + Political violence + + + + + + + + + Conventional terrorism + + + + + + + +b Non-honoring of sovereign guarantee + + + + + + + Non-honoring of arbitration award + + + + + + + + Wrongful calling of guarantee + + + + + + + + License cancellations + + + + + + + + + Sovereign default + + + + + + + Beneficiary type Privately owned + + + + + + + + + + Government owned + + + + + + + + Client type Privately owned + + + + + + + + + Government National + + + +** Subnational + + + + +** Parastatal + + + + +** Non-profit organization + + + + +*** Underlying assets/obligations Equity + Portfolio equity + Quasi-equity + + + + + + Subordinated loan + + + + + + Senior loan + + + + + + + Bonds + + + + + + + First loss tranche + + + + Mezzanine tranche + + + + Delivery of goods and services + Security/enhancement + Short-term instruments + + + + Trade credit + + + + Portfolio of assets + + Production sharing contracts + + + + Profit-sharing contracts + + + + Management contracts + + + + Franchising agreements + + + + Licensing agreements + + + + Turn-key contracts + + + + Operating leasing agreements + + + + Subordinated bonds + + + + Guarantees or securities + + + + Guaranteed percentage Full + + + Partial + + + + + +c Others Government counter-guarantee + + + + In kind contributiona + Local currency + + + + + + + Source: WBG. Note: *COP can be conditional or unconditional. **Provided that the government-owned investor operate on a commercial basis. ***Provided that the specific investment for which the coverage is being sought will be carried out on a commercial basis. Shaded cells are distinctive features to the referenced guarantee instrument. COP = Community of Practice; IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; GOLF = Global Offshore Liquidity Facility; MIGA = Multilateral Investment Guarantee Agency; PCG = Partial Credit Guarantee; PRG = Partial Risk Guaran- tee; PRI = political risk insurance; RSF = Risk-Sharing Facility; SCLG = single-currency loan guarantee; SNF = subnational finance. a. Tangible or intangible assets that have monetary value such as machinery, patents, processes, techniques, managerial know-how, trademarks, and marketing channels. b. Covered under war and civil disturbance. c. The portion of loss to be paid by MIGA in the event of a claim shall not exceed 99% for loans and 95% for all other instruments. 1 7 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 the political risk coverage of equity in one infra- a broad range of assets, including debt and equity structure project in Africa. Whereas its scope of or quasi-equity; a single asset or a portfolio of as- operations is limited by the Convention, it was also sets; local or foreign currency; local or foreign envisaged that the provisions could be elabo- funding; existing or new investments; of short-, rated, as the need arose, in the regulations, poli- medium-, or long-term maturity; and financial ob- cies, and rules put forward by MIGA's Board of ligations or nonfinancial contractual obligations. Directors, which would allow room for innovation and flexibility. Virtually all types of commercial entities--pub- lic or private, profit or nonprofit, domestic or Conclusion foreign--could access WBG guarantees. Risk The guarantee instruments of the three WBG coverage ranges from a single risk to compre- institutions have both clearly defined, dis- hensive political and commercial risks. The in- tinctive features and overlaps in terms of struments of the three institutions also have types of risks covered, underlying assets, clearly defined, distinctive features determined and beneficiaries. The product comparison is by their policies and regulations. In addition to based on the authorizing environment for the de- policies, however, other factors influence the ployment of the instrument as specified by their actual deployment of the instruments. These charter, internal policies, and regulations. WBG include market demand, strategic priorities, and guarantee instruments as a group can guarantee internal incentives. 1 8 Chapter 2 The gas pipeline between Mozambique and South Africa was the first cross-border pipeline in Sub-Saharan Africa. The project was guaranteed by MIGA and the World Bank to mitigate political risks. Photo courtesy of Sasol. Review of the WBG's Experience with Guarantees I n chapter 1 we presented the range of WBG guarantee instruments. We now turn our attention to the way these instruments have been deployed in practice by the three WBG institutions. MIGA Guarantees places that were largely ignored by international investors. Patterns of use Within the limits imposed by its Convention, Its client base also expanded--from three clients MIGA has made substantial strides in fulfilling in fiscal 1990 to 371 in fiscal 2007. MIGA's client re- itsmandate, expanding itsoperationsand client tention has also improved: 56 percent of its gross base, and diversifying its portfolio. MIGA has exposure is associated with repeat clients, mainly channeled an estimated $56 billion of investments for projects in the financial and infrastructure in high- and medium-risk countries through its PRI sectors. Most of MIGA's clients come from Part I products. Using the Institutional Investor Coun- countries, although MIGA has also focused on try Credit Rating index, it can be determined that promoting South-South investments, which rep- 45 percent of MIGA guarantee projects are lo- resent nearly 14 percent of the total amount of cated in high-risk countries and 31 percent in guarantees issued since fiscal 1990. MIGA has also moderate-risk countries; about one-sixth of MIGA achieved portfolio diversification in its Regional and projects are located in low-risk countries (figure sectoral distribution. Ex post evaluations of a sam- 2.1). ple of 21 MIGA guarantee projects during fiscal 1996­2002 show that about half (48 percent) of the MIGA's gross and net exposures follow the same projects attained satisfactory or better development trend. Frontier countries (high-risk and/or low- outcomes and made a positive contribution to income countries/markets) were also the recip- the achievement of MIGA's development man- ients of 56 percent of all MIGA guarantee date. These projects provided extensive benefits contracts issued, corresponding to $7 billion in to the host countries and communities. gross exposure. PRI exposure to IDA countries represented 18 percent of its overall portfolio, al- MIGA's PRI business expanded rapidly, espe- though the level of exposure to MICs accounts cially in the second half of the 1990s but was for a higher percentage (76 percent). This figure negatively affected by post-2001 events. It has reflects investor interest in emerging markets. now started to recover. Between fiscal 1990 and Most of the guarantee projects were located in 2007, MIGA issued $16.7 billion in guarantee cov- developing countries that did not receive large erage for 897 guarantee contracts in support of FDI inflows, indicating demand for MIGA PRI in 556 projects in 96 countries. Latin America and 2 1 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Figure 2.1: MIGA's Gross Exposure, by Country Risk Classification, Fiscal 1990­2007 2,500 2,000 exposure 1,500 million) guarantee (US$1,000 MIGA 500 New 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Fiscal year Low Medium High Source: MIGA. Note: Based on Institutional Investor Country Credit Rating. MIGA = Multilateral Investment Guarantee Agency. the Caribbean and Europe and Central Asia Re- gion by number of MIGA contracts, which also re- gions account for nearly two-thirds of MIGA con- flects investor interest in the continent. Contracts tracts issued during this period; its presence in issued for investments in the Middle East and the Middle East and North Africa Region has been North Africa, a current WBG strategic priority, rep- intermittent. resented 4 percent of total contracts issued; MIGA's presence there has been intermittent. The Regional and sectoral trend of MIGA's port- folio closely tracks FDI trends. In the early 1990s, Most of MIGA's clients purchase transfer re- the demand for MIGA coverage was spurred by striction coverage followed by expropriation investments in the mining sector in the Latin coverage, which is reflected in the large share America and the Caribbean Region. From the offinancialand infrastructure projectsinMIGA's mid-1990s until the end of the decade, the pri- portfolio. Financial and infrastructure sectors ac- vatization of state-owned utility companies and fi- count for about two-thirds of MIGA guarantees by nancial liberalization in the Region provided the volume or number of contracts. Investors in fi- impetus for foreign investors and lenders to ob- nancial sector projects tend to seek coverage tain MIGA coverage (figure 2.2). against currency transfer restrictions, whereas in- vestors in infrastructure projects generally pur- By 2003 and 2004, the rush of investment activities chase PRI against expropriation and breach of in the Russian Federation and the Commonwealth contract. Stand-alone breach of contract coverage, of Independent States group, as well as in coun- often combined with coverage for expropriation, tries preparing for accession into the European accounts for about 10 percent of overall volume Union, shifted MIGA's portfolio allocation to the of coverage issued, but demand has rapidly grown Europe and Central Asia Region. During the last two for these products because of increased invest- years, however, Africa has become the leading Re- ments in PPP projects. 2 2 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S Figure 2.2: MIGA Guarantees Issued and Trends 2,500 Privatizations and financial liberalization mostly in LAC PPP investments Accession to EU in ECA and Sub-Saharan Africa focus then South Asia, ECA, and Sub-Saharan Africa After Sept. 11 event, Argentina Bolivia, Ecuador, Venezuela, Russian Federation, default and Afghanistan invasion R. B. de, and Russian 2,000 Brazil, Turkey financial Federation expropriation and crisis Iraq renationalization. Rising investor invasion interest in Africa million) 1,500 Asian (US$ financial Mexico crisis banking issued 1,000 MIGA started crisis operations: three projects in mining sector 500 Guarantee 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Fiscal year Volume of guarantee issued 2 per. mov. avg. (volume of guarantee issued) Source: MIGA. Note: ECA = Europe and Central Asia Region; EU = European Union; LAC = Latin America and the Caribbean Region; MIGA = Multilateral Investment Guarantee Agency; PPP = public-private partnership. Coverage for transfer restrictions also tends to be comprised only 11 percent of total MIGA guar- in high demand among MIGA clients investing in antee contracts issued, compared with almost Latin America and the Caribbean, Asia, and lately half in 1996 (figures 2.3 and 2.4). This pattern also in Europe and Central Asia. Investors in Africa, reflects a robust growth of services in global FDI, however, prefer to use the full range of MIGA PRI particularly in the financial sector, as a result of fi- products--transfer restriction, expropriation, nancial liberalization and sector reforms in the and war and civil disturbance coverage are in 1990s. equally high demand, reflecting the perceived risks in the Region. Almost a third of MIGA's Historically, MIGA's financial sector exposure gross exposure in Africa is in the extractive in- has ranged from 30 to 40 percent of its total dustries sector, which is inherently risky because portfolio. Most of this exposure has been moti- of its complexity. vated by compliance of parent international banks with regulatory requirements and their meeting In response to the shift in global investment of country exposure limits. Initially MIGA covered trends, MIGA's portfolio has shifted fromman- the equity investments against currency transfer ufacturing to financial and infrastructure proj- restrictions of parent international banks that ects. Following rapid growth in the 1990s, open a local subsidiary or start bank branch op- guarantee contracts issued to manufacturing sec- erations. In subsequent deals, MIGA provided tor investments have been declining, in part be- coverage (mostly against transfer restriction risks) cause of the decline in new investments in the to the parent banks' shareholder loans to ex- sector and in part because of the perception of pand lending operations. That move allowed the investors that political risk in the sector is lower shareholders to meet regulatory and capital or could be self-insured. As of 2007, these contracts reserve requirements. As these clients com- 2 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Figure 2.3: Distribution of Guarantees Figure 2.4: Distribution of Guarantees by Issued by Sector (gross exposure), Fiscal Sector (number of contracts), Fiscal 1990­2007 1990­2007 Agribusiness Extractive Agribusiness Extractive 2% industries 5% industries Manufacturing, 10% 14% services, and tourism Manufacturing, 18% services, and Financial tourism sector 30% 27% Financial sector 35% Power 19% Other infrastructure 2% Other Telecom Power Water infrastructure 7% 14% Transport 2% Telecom 1% 2% 7% Water Transport 2% 3% Source: MIGA data. Source: MIGA data. menced operations in a host country and be- ments must be AAA" (BIS 2006, paragraph 59). came familiar with MIGA, they tended to seek PRI MIGA benefits from a shadow AAA rating as a coverage--first for their nonbank investments, member of the WBG. MIGA has also supported such as pension funds and leasing operations, and capital market transactions since 2003 in response then for the expansion of these operations in to the demand for PRI coverage for cross-border the same host country, and eventually for their and future flows transactions. bank and nonbank investments in other devel- oping countries. In the infrastructure sector, privatization activ- ities, which dominated the sector in the 1990s, During the last two years, MIGA coverage in the have been replaced byPPPs in recent years. An banking sector consisted of either equity or share- estimated 42 percent of MIGA's gross exposure in holder loans to boost the Tier II capital of bank the infrastructure sector was channeled to proj- subsidiaries to prepare for the adoption of the ects in the Latin America and the Caribbean Re- Basel II requirements, in addition to expanding gion, particularly Brazil. Asia and Africa received loan operations. Shareholder loans benefiting 18 and 17 percent, respectively, of the amount of from transfer and convertibility coverage by mul- guarantees issued to the sector during fiscal tilateral development banks could qualify from 1990­2007. lower capital risk weighting under the Basel II cap- ital adequacy framework, provided the multilat- MIGA's clients in the infrastructure sector either eral development bank has "very high quality have several equity investments (such as inter- long-term issuer ratings; that is, the majority of a national engineering companies and energy/utili- multilateral development bank's external assess- ty firms) or have provided project financing to 2 4 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S power, port, and road projects in one or more the West Bank and Gaza, and Afghanistan-- countries. Several of these investors are repeat which were not eligible countries--or where clients, and although client familiarity has been the risks were considered to be high under the good for MIGA's business, one of the unintended regular guarantee program. These funds were consequences is a trend toward concentration created as new instruments, different from the of its portfolio with a few clients. trust funds mentioned in the Convention, both to provide long-term insurance for eligible small MIGA guarantees were also allocated to in- and medium-size investments in these risky areas vestments for the development of SMEs and and to have a positive signaling effect to foreign investors. Since its inception, MIGA has provided investors. Under these plans, MIGA issues guar- coverage (estimated at $1 billion from fiscal 1997 antees on behalf of, and pays compensation to 2003) for investments to support SMEs. In fis- from, the trust funds. cal 2006, it introduced the Small Investment Pro- gram (SIP) to make PRI more accessible to small The European Union sponsored the $12 million and medium-size investors, especially from Part Investment Guarantee Trust Fund for Bosnia and II countries, who are investing in SMEs. SIP offers Herzegovina, with a special credit line for this standardized terms, with a fixed premium rate by purpose. This fund has been fully utilized to pro- country, and a streamlined underwriting process. vide guarantees for six projects--three in the fi- The program's initial coverage limit of $5 million nancial sector and three in the services and the was increased to $10 million in January 2008.1 manufacturing sectors. Since its implementation in fiscal 2006, 30 percent The Palestinian Authority, through an IDA loan, has of MIGA projects were within SIP and represent contributed to the Investment Guarantee Trust less than 1 percent of MIGA's outstanding port- Fund for the West Bank and Gaza, with additional folio to date. The majority of projects were in the funds from Japan and the European Investment sectors and Regions targeted by the program and Bank. More recently, MIGA created the Invest- were broadly consistent with country priorities or ment Guarantee Fund for Afghanistan with the WBG strategies. support of the Islamic Republic of Afghanistan, IDA, the Asian Development Bank (ADB), and SIP projects, especially in conflict-afflicted coun- the United Kingdom.2 MIGA had been able to tap tries, have potentially significant demonstration into these funds and leverage its resources for four effects. The Independent Evaluation Group (IEG) projects in Afghanistan under its SIP. has been unable to determine the success of the program because the outcomes have not been MIGA has paid three claims since it was estab- evaluated. Of the 15 projects that had coverage lished--anexcellentrecordforaninsurer. MIGA under SIP up to December 2007, 6 were can- has paid $5.2 million (net amount) for those three celled after 1 or 2 years, almost the same cancel- claims. In June 2000, MIGA paid its first claim, for lation rate as for regular projects. Some were an expropriation related to a power plant project cancelled because the project did not go forward. in Indonesia, which it has recovered from the The cancellation rate, especially for SIP projects government. In fiscal 2005, MIGA paid two claims that did not become operational, meant that ex- totaling $0.6 million. One was a war and civil dis- pected development impact was not realized and turbance claim filed in relation to a hydroelectric that MIGA could not fully recover the cost of un- plant in Nepal, and the second was related to a derwriting SIP projects. loan guarantee to a project in Argentina. In these two cases, because the amounts were small, pay- MIGA, together with other donors, created ment was not recovered from the host govern- three investment guarantee trust funds to fa- ments. cilitate foreign investment in conflict-affected countries such as Bosnia and Herzegovina, 2 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Figure 2.5: Reasons Given by Clients for Canceling a Guarantee, Based on Guarantee Contract Cancellations, Fiscal 2001­07 20 18 18 16 14 13 13 (%) 12 10 10 9 responses 8 Total 6 4 3 2 1 1 1 0 Loan repaid Change in Self- Others Project Premium Financial Went to Reassessed corporate insurance sold default difficulty other risk strategy insurer Reasons provided by MIGA clients Source: MIGA data. Between the second quarter of fiscal year 2003 and before the third anniversary of the contract, and fiscal year 2007, MIGA encountered 21 projects a quarter of the cancellations have taken place dur- with claims or preclaims that had been adequately ing the third year.5 Cancellations because of fi- provisioned. More than half of these were infra- nancial difficulty, inability to pay premiums, and structure projects, and about three-quarters of this switching to other PRI providers represent just 5 group were PPPs involving concessions from na- percent of the reasons given (figure 2.5). Thus, tional or local governments. The Overseas Private cancellations take place when the investments Investment Corporation's (OPIC) total claims pay- are still successful from a financial standpoint, out is valued at $964.7 million for 280 claims since and in some cases (self-insurance) the investors' it was created.3 Claims payment by the other re- perception of political risk has improved. These porting members of the Berne Union Investment cancellations are not unusual for the industry Insurance Committee for 2005 amounted to $113 and in some measure reflect MIGA's fulfillment of million ($85.5 million for 2006 and $12.2 million its fundamental mandate to attract investment for 2007), but outstanding claims are higher.4 to developing countries. MIGAPRIhashelped attractinvestments, butin- Effectiveness and additionality vestors tend to cancel when perception of po- MIGA's guarantees fill a gap in the provision of litical risk improves. For MIGA's operational long-term PRI. When the idea for an interna- sustainability, cancelled projects would have to be tional investment insurance agency was concep- offset by a strong pipeline. A MIGA guarantee tualized, the PRI market was dominated by contract has a 67 percent chance of being can- national agencies, and eligibility criteria were celled or terminated before reaching its expiration driven by government mandates and national re- date. Large exposures are more likely to remain quirements. The few private insurers in the busi- active until expiration. Approximately 37 percent ness offered limited PRI products--short-term of MIGA guarantees are cancelled by the investors coverage at high prices. Because of their limited 2 6 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S underwriting capacity, they were reluctant to positive signaling effect of the attractiveness of the cover investments in higher-risk countries unless country as an investment destination. the risk was reduced through reinsurance or co- operation with national insurers. Most of MIGA's dispute-resolution efforts relate to projects for which it has issued guarantee cover- The gaps left by the private and national insurers age. It has also selectively mediated disputes in- in providing long-term PRI to encourage FDI volving non-MIGA investors and host countries, flows to developing countries served as the ra- as in the case of its long-term mediation efforts tionale for a multilateral presence in the PRI mar- in Ethiopia.6 MIGA does not charge fees for me- ket. Then and now, MIGA's most visible and diation efforts that involve a client and the host important value added rests on ensuring the avail- government but seeks reimbursement for out-of- ability of long-term PRI to investments in difficult pocket expenses related to such missions in dis- parts of the world. For the equity holders, MIGA's putes unrelated to its guarantee. IEG evaluations PRI provides access to cheaper financing from of four projects in which MIGA's Legal Affairs and banks, whereas financial institutions can obtain Claims Group intervened indicated that MIGA regulatory relief to free up capacity for their own played a useful role in resolving several disputes internal risk management. related to power purchasing agreements and con- tributed to a successful debt restructuring that en- These additionalities have been recognized by abled the project to continue operating. market players in response to market studies; they also acknowledged that they are aware of MIGA cooperates with and leverages the par- MIGA's function as distinct from those of IFC and ticipation of public and private sector insur- the World Bank. An internal 2007 study reinforced ers. Co- and reinsurance arrangements with other MIGA's strengths in the PRI market, particularly insurers increase MIGA's capacity to support large its advocacy potential and excellent claims history. projects7 and allow it to manage the risk profile of its portfolio. In reinsuring the whole portfolio MIGA is not only a political riskinsurer; medi- or individual projects of other insurers, MIGA al- ation of disputes, based on its special relation- lows them to extend their capacity. This enables ship with governments, is part of MIGA's broad other insurers to participate in projects in more mandate to remove obstaclesto FDIflowsto de- challenging environments and for longer terms veloping member countries. The Convention than they would normally underwrite. Host coun- mandates that MIGA assist in settling investment tries also benefit from an increased interest from disputes; it can act as an independent mediator other insurers. MIGA's partners recognize its su- in cases involving an investor and the host gov- perior loss avoidance and recovery ability. ernment, even when the dispute is unrelated to any guarantee (Shihata 1988, pp. 257­85). Under its syndication program, MIGA has attracted more than $4 billion. Its panel of syndication part- MIGA's direct access to host governments (as its ners consists of private sector insurers, includ- shareholders) strengthens its ability to successfully ing, among others, ACE, AIG, Axis, Chubb, Coface, mediate mutually agreeable resolutions to retain Hannover Re, Lloyd's, Munich Re, Swiss Re, and investments and restore projects to normal op- XL Capital, as well as public insurers such as the eration. As a multilateral entity, it is able to play Export Guarantee Department (United Kingdom), the role of an honest broker, and the deterrence Servizi Assicurativi del Commercio Estero (Italy), effect arising from its mandate is valued greatly by Export Finance and Insurance Corporation (Aus- investors, who understand that host counties tralia), Compañía Española de Seguros de Crédito have more at stake than just individual projects-- a la Exportación (Spain), Garanti-Instituttet for potentially the much larger and more important Eksportkreditt (Norway), Slovene Export and De- relationship with the WBG could be at risk. For velopment Bank, Export Development Canada, host countries, resolution of disputes provides a and the Overseas Private Investment Corporation 2 7 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 (United States). In recent years, there have been An IEG review of the quality of underwriting for concerns that MIGA may be ceding a greater share new projects (fiscal 2005­06) also showed that in of its premium income to other insurers, which more than half the cases, MIGA did not clearly de- prompted a recent management review of MIGA's fine the particular value it brought to the project. net retention policy and methodology. It is unclear in these ex post and ex ante reviews whether those investments would have been MIGA's longer-term vision gives comfort to in- made without MIGA's presence. vestorsenteringintolong-terminvestmentsinde- veloping countries. According to MIGA's client MIGA's impact on projects beyond its role as an survey, a key attribute of MIGA is its long-term insurerhasbeen limited. In keeping with the na- commitment to and relationship with countries. In- ture of its guarantee product (that is, further re- vestors indicated that when there is a crisis in the moved from project implementation than a country, "MIGA stays," because of its development financier or equity investor), MIGA cannot be ex- objectives, when private insurers might withdraw pected to have the capacity to influence project because of their profit orientation. MIGA's addi- design. However, it can have a proactive influence tionality, as indicated through client surveys and ex in raising standards through its environmental post evaluations, enables investors to access fund- and social safeguard policies and environmental ing at a lower borrowing cost; seek improved terms guidelines. Until recently, MIGA did not proactively and conditions for private investments, lowering offer services or advice to improve project de- their "hurdle rates"; have clarity regarding MIGA's velopment outcomes. In an SIP project, MIGA risk definition because it reflects insurance indus- played an important role by assisting the investor try practice; and have additional protection from in successfully structuring the deal for the proj- the umbrella effect of WBG membership. ect. Contributing to a broader set of issues be- yond its traditional role as an insurer is consistent ExpostevaluationsconfirmMIGA'sadditionality with MIGA's development mandate, but it is im- in a sample of MIGA's projects. In 2006 IEG eval- portant to do so in a way that is also consistent uated a sample of 21 mature MIGA projects with market practices and the more remote po- (IEG­MIGA 2006) to assess MIGA's additionality sition of the typical insurer in relation to the proj- as an insurer, its role in leveraging and comple- ect company. menting partners, and its contribution to its clients. For instance, MIGA brings WBG environmental Potential for use and social safeguards to the design and imple- Political risk remains a considerable threat to mentation of their clients' investment projects. global business and has spurred the rapid growth in PRI volumes since 2005. Despite the The results indicate that in the majority of projects variety of risk-mitigation products available to (18), MIGA made important contributions. The investors, the demand for PRI products contin- high ratings reflect the perception of investors ues to be robust, as reflected in the rapid growth that, in many cases, MIGA's involvement was crit- of new business by Berne Union investment in- ical for the investments to proceed or provided surance members from 2005 through 2007 (fig- comfort to clients entering new markets or new ure 2.6). In 2001 total new business by members sectors (especially private provision of public in- amounted to $16 billion; in 2007 new business frastructure). MIGA's insurance was particularly reached $62 billion as investors prepared for the crucial for investors in post-conflict countries. In resurgence of traditional political risks, espe- one evaluated project, MIGA provided a tailored cially expropriation risk, which had been largely product that was not available from private in- discounted until the early 2000s. The outlook surers for the long term required by the investors. for the next five years is similarly strong, ac- In six of the ex post evaluated cases, MIGA coverage cording to a June 2007 global survey of 602 ex- was a condition for lenders to provide funds. ecutives by the Economist Intelligence Unit (EIU 2007). 2 8 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S Figure 2.6: Snapshot of the PRI Market: Total New PRI Business Volumes, Calendar Years 2001­07 (Berne Union investment insurers only) 70,000 . . . then Ecuador and Russian 60,000 Federation Bolivia, Venezuela, Committee) R. B. de expropriation 50,000 and renationalization million) Aftermath of Sept. 11, Insurance 40,000 Argentina partial default, (US$ and Afghanistan invasion 30,000 Iraq invasion Amount Investment 20,000 Union 10,000 (Berne 0 2001 2002 2003 2004 2005 2006 2007 Calendar year MIGA National agencies Private insurers Source: Berne Union. Note: PRI = political risk insurance. The survey also revealed that although many busi- terms of covering nonequity direct investments nesses are not adequately managing their expec- and other types of investments eligible under the tations of increased political risk, companies with Convention. Since its inception, about 70 per- better political risk assessment capabilities expe- cent of MIGA's gross exposure supported equity rienced fewer cases of expropriation, government investments; the share of direct equity was only payment default, or currency transfer restriction slightly higher than shareholder loans and other than other firms. Recent episodes of expropriations equity-like investments. and nationalization in Bolivia, Ecuador, Russia, and República Bolivariana de Venezuela have Coverage of other investments such as non- underlined the continued need for MIGA's tradi- shareholder loans (with an "equity link") com- tional PRI. MIGA can readily provide these prod- prised almost a third of all investments covered ucts, given its expertise in the PRI market. by MIGA. Investments involving production- sharing and management agreements and simi- MIGA has not taken full advantage of the op- lar investments represent a miniscule share of portunities available under its charter and op- MIGA's guarantees. This reflects sporadic demand erational regulations. It has been operating in but may also point to an unexploited opportunity.8 its comfort zone. Table 1.1 shows the eligible in- vestments, investors, and risks that are allowed MIGA has only covered performance bonds as under the Convention. It also indicates the leeway the underlying investment in two projects, available to MIGA in its scope of operation. production-sharing agreements in one project and management agreement in two projects. This There is still room for MIGA to grow by taking ad- represents an untapped opportunity, especially vantage of the broad scope of the eligibility re- with the growth of PPP and extractive industries quirements and the flexibility allowed in its charter. activities. Coverage to nonprofit organizations is MIGA has yet to push the envelope, especially in another unexploited opportunity. 2 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 MIGA has not thoroughly explored other av- MIGA needs to market itself to new investors enues that represent an untapped growth op- from member countries, especially those that portunity that requires a serious, innovative have become overseasinvestors. With the recent business development push beyond MIGA's boom in global capital markets and FDI to de- comfort zone. The Convention broadly defined veloping countries, a new and growing segment the scope of MIGA's operations while giving MIGA's has emerged: South-South investors. Capital flows Board of Directors the authority to amend the Op- among developing countries, particularly FDI, erational Regulations as the need arises. That al- are now growing more rapidly than investments lows MIGA's guarantee program to be responsive from the developed world. The WBG estimates to the changing needs of member countries and that investment from developing countries now investors. Article 13(c) of the Convention also al- accounts for a third of all FDI going to develop- lows the Board, through a special majority, to ex- ing countries. These emerging investors represent tend eligibility to a national of the host country new pools of financing in their own right, and they or to a company that is incorporated in that coun- are potential new clients for PRI providers. South- try or whose capital is chiefly owned by its na- South investments represent only 16 percent of tionals, such as those living abroad with large the number of contracts issued and 14 percent of offshore funds.9 MIGA's gross exposure. Further, the Convention and Operational Regu- Continuing involvement in PPP structures to lations allow the Board, by special majority, to support infrastructure development and im- add any other noncommercial risks--including prove services requires access to long-termPRI. acts of terrorism or kidnapping specifically di- MIGA's support for PPP transactions as an alter- rected against the guarantee holder--to be eligi- native to privatization has remained strong since ble for cover, except for currency depreciation or the late 1990s. The Organisation for Economic Co- devaluation and events that occurred before the operation and Development estimates that $1.8 contract was signed.10 There is also some leeway trillion annually is required across five infra- to provide guarantees in the host country's cur- structure subsectors (electricity, water, roads, rail, rency (Article 3(e) of the Convention and Para- and telecoms) alone. The PPP is a mechanism in- graph 1.09 of the Operational Regulations) if it is volving private sector supply of infrastructure as- a considered a "freely usable currency." sets and services traditionally provided by the government. Adequate risk transfer from a gov- There is room to improve client retention. Re- ernment to the private sector is essential for such peat clients represent 22 percent of MIGA's cus- transactions to generate private investor inter- tomer base but account for 56 percent of its gross est. exposure. With improved communication and client aftercare, this retention rate could increase. MIGA is well placed to continue its support for One of the recurring complaints identified in a se- these projects because of its experience, and the ries of client studies commissioned by MIGA (in long-term nature of its insurance matches the 1996, 1998, 2005, and 2007) was the lack of com- long-term duration of PPP agreements. Most of the munication and follow-up with the clients after infrastructure projects it has covered since 2000 contract issuance. Perhaps because little impor- were structured as PPPs. Other insurers, national tance was given to client aftercare, there was a lack and private, are still building their expertise. of monitoring, or there was very high staff turnover. To a large extent, these impediments also More systematic support of innovative projects affect client diversification, because client satis- offers growthopportunities. MIGA has provided faction could be the most cost-effective market- coverage to several innovative projects, but most ing that MIGA could undertake to diversify its of these are one-offs. Recent examples include pro- client base and solve the persistent challenge of viding coverage against expropriation and breach rectifying client concentration. of contract risks to a landfill project in El Salvador 3 0 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S Table 2.1: MIGA Pricing of Guarantees (2007) Elements of pricing Expected loss Risk load (unexpected loss) Expenses Cancellations Includes exposure type, Identifies risk load required Expenses allocated to help Historical experience, transi- expected frequency, to earn target return on risk ensure cost recovery tion matrixes, and cash-flow severity, and recovery capital (RAROC) timing incorporated Expected results Financial sustainability Consistency Transparency Objectivity Cost recovery to ensure Pricing consistency from one Client rates based on an Objectivity in pricing sought financial sustainability client to the next to attain evaluation of the country and through predictable and over the long term pricing equity project risks methodical pricing Source: MIGA and IEG adaptations. that sought offsets through trading of carbon parency, and objectivity while also ensuring credits. So far, this has been a one-off, but there cost recovery. An in-depth analysis of MIGA's is significant potential for MIGA to offer coverage exposures, costs, and pricing was carried out for projects seeking carbon credit offsets, espe- during 2003. This work produced a compre- cially those involving renewable energy, because hensive pricing framework that was adopted in other Berne Union investment insurers are still 2004 and refined in 2007. This framework sets building this line of business. premiums for guarantees to cover risk (claims risk and reinsurer nonperformance risk), the Project Finance Yearbook 2006 estimated that the portion of risk capital consumed by a project, ad- annual value of project-based carbon credit trans- ministrative expenses, and expected cancella- actions increased to $2.5 billion in 2005 from $500 tion. MIGA updates its in-house ratings for million in 2003.11 It insured its first private place- country and project risk quarterly as an input for ment in support of a toll-road project in the Do- the pricing model as well as current portfolio minican Republic in fiscal 2005, which allowed composition to capture concentration effects the project to obtain a single notch rating above (see table 2.1). the sovereign risk because of MIGA's PRI and its shadow AAA insurer strength rating.12 It had more Constraints to use success in replicating support for projects involv- The market for traditional PRI products has ing asset and future flows securitization, although been growing, but for the past five years MIGA's with the current massive problems in the financial market share has been eroded by private in- markets, demand may remain low for some time. surers, who have ventured into countries that were traditionally served by national agencies Last year, MIGA issued guarantees for PRI and andMIGA.They have offered a wide range of guar- nonpolitical force majeure events for a port proj- antee and tailor-made financial products, as well ect in Djibouti. This experience shows that it as lower premiums (figure 2.7). Private insurers can make specific adjustments to align its con- have gained market share because of their ability tract of guarantee with Islamic finance struc- to offer tailor-made financial products that MIGA ture, offering significant growth potential for and some national insurers are unable to match MIGA as other PRI players are still learning about because of the limitations in their charters. Islamic financing. This competition is not only hurting MIGA, but MIGA established a pricing frameworkin 2004 several national insurers have been unable to that promotes pricing consistency, trans- book new PRI business. However, the situation is 3 1 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Figure 2.7: Comparing MIGA's Market Share with Shares of Other BU-IIC Members 100 90 80 (BU-IIC) 44 44 51 55 70 59 58 59 60 business 50 new 40 total 48 50 of 30 45 31 26 43 38 20 Share % 10 10 13 9 6 4 2 4 0 2001 2002 2003 2004 2005 2006 2007 Calendar year MIGA National agencies Private insurers Source: Berne Union and MIGA. Note: BU-IIC = Berne Union Investment Insurance Committee; MIGA = Multilateral Investment Guarantee Agency. Numbers may not total 100 percent because of rounding. more difficult for MIGA, because it is the only in- Another constraint is MIGA's eligible investments. surer among the Berne Union investment insur- MIGA can cover equity or quasi-equity invest- ance members offering a monoline product. ments, shareholder loans, and nonshareholder Competition from private insurers has also pushed loans, and only if these are equity linked. Private down the premiums for PRI coverage across the insurers are not constrained by this eligibility re- board, which put MIGA and national agencies at quirement and have been able to grow their busi- a disadvantage. One of the comparative advan- ness volume insuring mostly debt. This trend is tages of MIGA mentioned in 1995 and 1998 client supported by the declining share of equity in- studies was its low premium compared to pri- vestments MIGA covers, while the share of loans vate insurers. The erosion of this advantage pres- with an equity link and quasi-equity has been ents a challenge to MIGA. increasing. In addition, private insurers have been making in- In the longrun, MIGA'sConvention needsto be roads into high-risk countries, especially in Africa. updated to accommodate changesin the global In the late 1990s MIGA assumed larger exposures investmentenvironment.Notwithstanding MIGA's in medium- and high-risk countries, but since strengths in the PRI market, the limitations im- 2002, exposure in lower-risk countries has been posed by its Convention pose a large challenge in rising as private insurers have gained an increas- the current global investment environment. The ing foothold in the high-risk territories. In 2005, most notable constraint is the inability to insure private insurers accounted for 50 percent of new stand-alone debt. The requirement to have an PRI business in Africa among all Berne Union in- "equity link" has proved to be a constraint to vestment insurance members; in 2007, their over- growth at a time when debt has been the main all share increased to almost 70 percent. source of financing. MIGA is also unable to par- ticipate in other current growth areas such as 3 2 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S comprehensive (that is, covering both commercial Brokers and other political risk insurers noted that and political risk) nonpayment coverage; existing MIGA is unable to deviate from standard contract equity and portfolio investment/acquisition cov- language and that its requirements, mainly on el- erage unless the investment is also linked to ex- igibility and price, are set throughout the contract pansion and modernization of the investment duration and cannot be altered, even if circum- project; and local currency loan financing. stances change. However, there are potential changes to MIGA's Second, the underwriting process is too long and products that would not require amendments cumbersome--largely because of MIGA's devel- to the Convention. Given the tedious proce- opment mandate as part of the WBG--compared dures in amending the Convention, MIGA can with private sector insurers. Although thorough, take advantage of the flexibility of interpretation the underwriting process takes longer (usually of the Convention. MIGA can, of course, suggest around 139 days) than the private insurer norm changes to its Operational Regulations that require of one week. only approval of the Board. An example cited by MIGA staff during interviews with IEG was the re- And third, clients report inconsistent experiences vision of MIGA's strict application of the eligibil- when working with MIGA staff, including per- ity requirement relating to the timing of the ceived indecisiveness about covering a project, investment. This strict regulation of "new invest- lack of client after care, and lack of continuity in ments" comes directly from the explanation pro- client relationships because of high staff turnover. vided in the Operational Regulations that the MIGA's business development is also hampered application for guarantees must have been filed by the lack of an integrated approach for selling before the implementation of the investment the guarantee product and the lack of clear def- project. Project implementation is assumed to inition of the roles and responsibilities for busi- have begun either when resources of the invest- ness development (IEG­MIGA, forthcoming). ment project have been transferred to the proj- ect enterprise or when such resources have been MIGA's processing time varies by sector and irrevocably committed. Staff have commented Regionbuthasbeenincreasingoverthepastfew that these elements are overly constraining, par- years. MIGA's Operational Regulations state that ticularly for projects in the infrastructure and ex- all guarantees should be completed, to issuance tractive industry sectors. of the guarantee, within 120 days of receipt of a Definitive Application (MIGA 2007b, §14). But on In addition, MIGA staff have brought up several average, it has taken 270 days13 to process a guar- policy issues that may need Board approval but antee--from the application, through the un- would not require changes to MIGA's Operational derwriting process, obtaining Board and host Regulations or Convention. However, some of country approvals, negotiating the guarantee con- the suggested changes, especially those identified tract, to signing the contract. in the latest client study, may still require changes in the two key documents. Processing times vary considerably by sector. Proj- ects in the financial sector have taken an average MIGAfacesseveral internal constraints, ranging of 182 days; projects in the oil, gas, and chemicals from inflexibility of terms, to cumbersome sectors have taken an average of 401 days be- processes, to lack of continuity in client rela- cause of the significantly more extensive envi- tionships with staff. According to a recent mar- ronmental and social due diligence requirements ket study, and confirmed by IEG during staff in these sectors. interviews, MIGA is constrained by several weak- nesses. First, the terms and provisions under the There are also sharp differences in processing contract of guarantee are considered inflexible. times among Regions. Projects in the Middle East 3 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 and North Africa take 430 days, compared with 193 projects. Members consider the analysis of the in- days in Europe and Central Asia. Of particular vestment project's commercial viability important concern is the increase in processing times over in (1) preventing claims, (2) ensuring that the the years, caused partly by several recent large and guarantee holder can pay its premium, (3) in- risky projects (along with coverage of two large creasing the chances of recovery, (4) identifying cross-border investments) and the greater em- sources of weaknesses in the project structure, phasis on a project's expected development (5) facilitating the adoption of policy language in contribution. the guarantee contract to address these issues, and (6) preventing fraud.16 In all but a few national In the interviews for this study and in IEG­MIGA's insurers, the investment insurers require finan- fiscal 2007 Annual Report (IEG­MIGA 2007), cial statements, a copy of a feasibility study or bus- MIGA staff expressed concern about the loss of iness plan, shareholder agreement, licenses, market share to the private sector as the Agency environment-related reports, sales or distri- aligns its requirements, policies, and procedures bution agreements, licenses, legal opinions, and more closely with those of the WBG. Although so on--a list much like MIGA's informational this alignment promotes MIGA's development requirements. role and the deterrence effect of its guarantee products, it also increases transaction costs for There is roomto increase MIGA's underwriting both the client and MIGA and undermines MIGA's capacity, butbusinessvolume and coststructure competitiveness. need to improve to tap market potential. As of fiscal 2007, MIGA's statutory underwriting capac- Achieving efficiency in its due diligence ity17 was valued at $10.6 billion (MIGA 2007a), process--ratherthanrelaxingitsrequirements-- but its outstanding exposure for the same pe- posesachallenge to MIGA. The key challenge for riod is half this amount ($5.3 billion). Actual MIGA is to become efficient in conducting its due amounts issued for the same period totaled only diligence process without affecting the quality of $1.4 billion. the projects it insures. There is a longer lag time between the submission of the definitive appli- MIGA has room to grow its business, but it must cation and the actual underwriting for both reg- also address its cost structure. Average admin- ular projects and SIP, because it takes the investors istrative cost as a proportion of average net pre- a long time to submit the requested documenta- mium income increased from 55 percent during tion.14 Host country approval, a requirement be- fiscal 1995­2000 to 80 percent during fiscal fore contract issuance, also takes longer than the 2005­07. To increase business solely by lowering 30-day period stipulated in the Operational Reg- premium rates to keep pace with the pricing of ulations.15 Listing the required documentation in private insurers may not be a sustainable option the definitive application or introducing some because MIGA has cost structures that private in- conditionality in the guarantee contract that can surers do not have, such as expenses related to help ensure that development impact could align being part of the WBG. In fiscal 2007, MIGA in- the due diligence process with the rest of the in- creased its country and individual project limits dustry. under its regular guarantee program to enhance its ability to make bigger deals. It also expanded In practice, most Berne Union investment in- the individual coverage limit under SIP to attract surers analyze the project's commercial viability, more investors. MIGA has also been trying to off- in addition to political risk, but the depth of the load certain corporate costs to cut its expenses, analysis varies according to the type and size of but this has to be accompanied by growth in busi- the investment and the size of the investor. In- ness volume that goes beyond its business as frastructure and project finance activities are sub- usual. Simplification of business processes also jected to greater scrutiny than other types of needs to be addressed. 3 4 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S Figure 2.8: Total World Bank Net Guarantee, by Country Risk Classification, Fiscal 1990­2007 450 400 350 million) 300 (US$ 250 200 150 commitment Net 100 50 0 1990 1994 1995 1996 1997 1998 1999 2000 2001 2003 2004 2005 2007 Fiscal year Low risk Medium risk High risk Source: World Bank, Institutional Investor Country Risk ratings. World Bank Guarantees financial instruments that reach the private sector other than the Financial Intermediary Loans that Patterns of use reach private commercial banks. Within PRGs, Two distinct World Bank guarantee products there are distinctions between PRGs originating have aimed to enhance sovereign access to from IBRD and those originating from IDA, in- commercial markets as well as to catalyze pri- cluding eligibility, pricing, and country allocations. vate investment projects bymitigating political The Bank also offers IBRD Enclave Guarantees, risk.As discussed in chapter 1, the World Bank has which are PRGs for foreign exchange­generating two distinct guarantee products: PCGs and PRGs. commercial projects in IDA-only countries. The PCGs guarantee government repayment of a com- Bank has also supported establishment of several mercial loan or bond for public sector invest- guarantee facilities to wholesale PRGs through ment projects. PBGs are a type of PCG that support national or regional guarantee agencies. government borrowing that is not tied to a par- ticular investment project. Both PCGs and PBGs The objectives of World Bank guarantees, as iden- are thus issued to support public sector borrow- tified in policy documents since 1994, have been ing. Along with IFC Subnational Finance Guar- to enhance sovereign access to commercial fi- antees, they are the only WBG guarantee nancing, catalyze private investment flows, instruments to do so. broaden the sector and geographic destinations of private capital flows, expand the sources of fi- PRGs support private sector investment projects nancing for WBG projects (mainly for IBRD PCGs), by mitigating political risk. PRGs are thus issued influence sector policies and the regulatory en- to support private sector borrowers and overlap vironment for private sector participation in a with MIGA and IFC PRM products (discussed in sector, and enhance local capital markets and ac- chapter 3; see figure 2.8). PRGs are the only Bank cess to international markets. 3 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Figure 2.9: Distribution of IBRD PCGs by Figure 2.10: Distribution of IBRD PCGs by Region (net commitment), Fiscal 1990­2007 Sector (net commitment), Fiscal 1990­2007 Middle East and North Africa All 9% infrastructure 13% Policy reform Telecom 26% 3% Latin America and the East Asia Caribbean and Pacific 26% 52% Power 58% Europe and Central Asia 13% Source: World Bank data. Source: World Bank data. Note: IBRD = International Bank for Reconstruction and Development; PCG = partial credit Note: IBRD = International Bank for Reconstruction and Development; PCG = partial credit guarantee. guarantee. A limited number of PCGs have been issued, All the PCGs were in IBRD countries except mostlysupporting large public power projects. three in China, which was a blend country at the Since 1990, the Bank has issued 10 PCGs in 8 time of approval. The three projects in China sup- countries, including 2 policy-based guarantees. ported commercial bank lending to the gov- The 10 PCGs, worth $1.6 billion, supported proj- ernment of China for on-lending to state utilities, ects worth $12 billion, with a leverage ratio of 7:1. and all the other operations involved access to By Region, 5 were in East Asia and the Pacific (3 commercial funds from capital markets. All the in China, 1 in the Philippines, and 1 in Thailand); PCGs were issued between 1990 and 2001, and 2 were in the Middle East and North Africa (Jor- no PCGs have been issued since 2001. None of dan and Lebanon); and 1 was in Europe and Cen- the PCG operations involved MIGA or IFC. One tral Asia (Hungary). The two PBGs were in Latin PCG has been called--the $250 million Argentina America and the Caribbean (Argentina and Colom- PBG (in 2001), discussed below. bia; see figure 2.9). Seven IBRD PRGs have been issued, for the All the project-based PCGs supported access to most part in the power sector. Since 1994, the commercial finance by state-owned utility com- Bank has issued seven IBRD project-specific PRGs panies for large infrastructure projects, with an av- worth $838 million, including one IBRD Enclave erage project size of nearly $1.3 billion. Six of the Guarantee in an IDA country. The IBRD PRGs eight project-based PCGs were in the power sec- supported projects worth $5.7 billion, with a tor, one was in telecommunications, and the proj- leverage ratio of 6.8:1 and an average project size ect in Hungary supported public investment in of $812 million. Two PRGs were in South Asia several sectors (see figure 2.10). The two PBGs (both in Pakistan), two in Europe and Central supported economic reform programs in Ar- Asia (Romania, Russia /Ukraine), two in the Mid- gentina and Colombia. dle East and North Africa (Jordan and Morocco), 3 6 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S Figure 2.11: Distribution of IBRD/IDA PRGs Figure 2.12: Distribution of IBRD/IDA by Region (net commitment), Fiscal 1990­2007 PRGs by Sector (net commitment), Fiscal 1990­2007 Sub-Saharan Africa 21% South Asia 30% Manufacturing, services, and tourism 16% East Asia and Pacific 9% Power 78% Extractive industries 6% Middle East Europe and and North Africa Central Asia 18% 22% Source World Bank data. Source World Bank data. Note: IBRD = International Bank for Reconstruction and Development; IDA = International Note: IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; PRG = Partial Risk Guarantee. Development Association; PRG = Partial Risk Guarantee. and the IDA enclave was in the Sub-Saharan Africa Althoughlimited in use, IDAPRGshave reached Region (Mozambique). Five of the IBRD PRGs high-risk, low-income countries. Six IDA PRGs were in the power sector, including four projects have been issued since 1999, worth $378 million. that supported independent power producers The IDA PRGs supported large PPPs, with an av- and one that supported privatization of electric- erage project size of $620.4 million, compared to ity distribution utilities in Romania. The Russia/ an average of $108 million for infrastructure proj- Ukraine project supported development of a pri- ects supported by normal IDA lending. The IDA vate commercial satellite launch service, using PRGs have had a high mobilization rate of 9.7:1. Russian/Ukrainian rocket technology and equip- ment that operated from a mobile platform in the Three IDA PRGs were in Africa (Côte D'Ivoire, Pacific Ocean. The enclave guarantee supported Uganda, and four West African countries), two private sector development of a natural gas were in East Asia and the Pacific (Vietnam and Lao pipeline between Mozambique and South Africa. People's Democratic Republic), and one was in South Asia (Bangladesh). All of these were low- Clients have been mostly large multinational cor- income, high-risk countries at the time the IDA porations such as AES, Boeing, and Sasol Lim- PRGs were approved (see figure 2.11). ited (South Africa). Three of the seven IBRD PRGs were in countries that were high risk at the time Five of the IDA PRGs were in the power sector, of approval (Pakistan, Ukraine/Russia, and Mozam- in support of independent power-producer bique). Three projects involved IFC or MIGA: IFC arrangements. The other project supported was involved in the Pakistan Uch Project and the private development of a natural gas pipeline Mozambique gas pipeline, and MIGA was involved from Nigeria to Ghana, Togo, and Benin (see fig- in the Mozambique project. No IBRD PRG has ure 2.12). been called to date. 3 7 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 As with IBRD PRGs, clients for IDA PRGs have been fiscal 2005 West African Economic and Monetary large multinational corporations, including AES, Union (WAEMU) Capital Market Development Chevron-Texaco, and Sithe Global Power (United Project used the Banque Ouest-Africaine de States). In two projects, the European Develop- Développement (BOAD) to market IDA and MIGA ment Finance Institutions, a group of 16 European guarantees for small and medium-size infra- bilateral institutions, were among the investors. structure projects. The fiscal 2005 Peru Facility of- Four of the six IDA PRGs involved IFC or MIGA, fered PRI through the government's investment- and two projects involved IFC, the Bank, and promotion agency on loans of up to 15 years to MIGA (discussed below). An additional two IDA catalyze investment in infrastructure PPPs. PRGs, both in Sub-Saharan Africa, have been ap- proved but are not yet effective: a Senegal power Effectiveness and additionality sector project and a PRG to support private op- The PCGs helped public agencies tap commer- eration of the Uganda-Kenya railway link. None of cial markets for better lending terms than they the IDA PRGs have been called to date. would have received without guarantees (table 2.2). The tenures of the loans, in particular, were Several efforts have been made to wholesale considerably longer than would have been possible Bankguaranteesthroughguaranteefacilities.Be- without the credit enhancement. In the Philippines tween 1995 and 2005, the Bank made eight efforts and China, for example, although the PCGs did not to wholesale PRI through local guarantee agencies, significantly lower the cost of borrowing, the 15- including five in the Europe and Central Asia Re- year loan tenures obtained were almost double gion, one in the Latin America and the Caribbean what the market was offering. In higher-risk coun- Region (Peru), and two multilateral efforts in Sub- tries such as Lebanon, Jordan, and Thailand Saharan Africa. The efforts took several forms, (shortly after the 1997 financial crisis), both the including a standby loan that would disburse only cost and terms of the loans were significantly im- if a guarantee was called; direct loans to member proved. In Thailand, in particular, the guaranteed countries that provided for the funds to be de- bond issued was rated three to four notches above posited into an escrow account with an agent Thailand's long-term foreign currency rating and bank when guarantees were issued; normal IDA generated significant investor interest, at a time loans, under which guarantee agencies were es- when the appetite for bonds issued by crisis-af- tablished; and direct Bank issue of PRGs identi- fected countries was low. fied by the local guarantee agency. In the case of the two PBGs, the guarantees ex- The effort originated in the Europe and Central erted considerable leverage--generating financ- Asia Region in the mid-1990s, where the facilities ing of 4.7 times the value of the PBG in Argentina supported short- to medium-term trade finance and 6.3 times its value in Colombia. The Colom- operations. These included facilities in Albania, bia operation achieved investment grade status Bosnia-Herzegovina, Moldova, and Ukraine that and enabled Colombia to reestablish access to U.S. were approved between fiscal 1995 and 1998. capital markets at a time when investor interest Subsequent facilities aimed to provide PRI for was minimal. In Argentina, although the country longer-term investment projects. The fiscal 2001 was able to access non-U.S. capital markets at Russia Coal and Forestry Facility aimed to en- similar terms, the PBG enabled it to issue a sig- courage new private investment in these industries nificantly larger bond ($1.2 billion) than would oth- by guaranteeing loans against political risk for erwise have been possible at the time. terms of 5­10 years. The fiscal 2001 African Re- gional Trade Facilitation Project created a dedicated PCGsalso helped introduce orreintroduce bor- multilateral guarantee agency, the African Trade In- rowers to commercial markets. A key objective surance Agency (ATI). This agency was owned by of the PCGs was to introduce the borrowing agen- 10 African states and empowered to issue guar- cies to commercial markets so they could borrow antees and insurance backed by IDA funds. The in the future without needing credit enhance- 3 8 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S Table 2.2: PCGs Enabled Public Agencies to Access Commercial Finance at More Favorable Terms Spread All-in-cost Rating Market over of at spread PCG borrowing Diff. Term Term Diff. $ Millions time of Finance (% over (% over PCG fee (% over with without with with $ Millions financing Leverage Country issue type UST) UST) (%) UST) PCG PCG PCG PCG PCG generated ratio Jordan B+ Capital 3.00 1.1 0.50 1.60 1.40 2 7 5 50 50 1 market Philippines BB­ Capital 3.00 2.5 0.50 3.00 0 7 15 8 100 100 1 market Lebanon BB­ Capital 3.00 1.0 0.25 1.25 1.75 5 10 5 100 100 1 market Colombia BB Capital 6.50 5.0 0.50 5.50 1.00 5 10 5 159 1,000 6.3 market Argentina BB Capital 5.84 5.0 0.75 5.75 0.09 5 5 0 250 1,165 4.7 market Thailand BBB­ Capital 8.50 2.9 0.50 3.40 5.10 0 10 10 300 300 1 market China BBB+ Loan 0.75 0.6 0.20 0.80 ­0.05 7 15 8 150 150 1 Source: World Bank data. Note: PCG fees are approximate, as different fees were charged during the noncallable period; they do not include commitment fees; and some fees were paid up front and others in installments. PCG = Partial Credit Guarantee; UST = U.S. Treasury. ment. All the public agencies that accessed capi- small volumes. In Thailand, the PCG was issued on tal markets under the PCGs, with the exception the heels of the financial crisis, when the country's of the Lebanese electricity utility, subsequently ac- access to commercial markets had been closed. cessed commercial markets again, without guar- Studies also suggest that sovereign bond issues antees. such as the one in Thailand have had a positive ef- fect on corporate bond markets by providing In Jordan, the PCG helped the telecom utility be- benchmarks and stimulating production of infor- come the first Middle Eastern corporation to tap mation, thereby lowering corporate bond yields the Eurobond market. It helped Jordan establish and trading spreads (Dittmar and Yuan 2008). a track record and subsequently reaccess the Eu- robond market, without the Bank's support. The However, PCGs can remove the Bank's leverage Jordan operation also involved the participation in advancing sector reforms, and their timing of the local capital market, facilitating mobilization canbeinappropriate.Three PCG projects had un- of domestic foreign exchange deposits. In the satisfactory outcomes--the two power projects in Philippines, the national power utility supported the Philippines and Lebanon and the Argentine by the PCG has also continued to borrow from in- PBG that supported its 1999 structural reform pro- ternational markets. gram. All three projects were accompanied by Bank loans, however, and the unsatisfactory outcomes re- The PCGs have also reintroduced borrowers to flected broader weaknesses in the programs rather commercial markets following financial crises. In than the guarantee instrument itself. Argentina and Colombia, for example, although both countries had previously accessed interna- In the Philippines, inadequate implementation of tional capital markets, the PBGs effectively rein- policy reforms and tariff adjustments prevented the troduced their large bond issues to international power utility from attaining financial viability. Fol- markets at a time when they were either closed to lowing the collapse of the Argentinean financial sys- emerging market economies or constrained to tem, the country's adjustment program went off 3 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 track, and reforms that were intended to be sup- ample, it was found that investors had given no ported by a Bank adjustment loan as well as the PBG value to the deterrent value against default implied financing were not achieved. In Lebanon, progress by the Bank's preferred creditor status (IMF 2003). on sector reforms was limited and the Bank's self- evaluation of the project found that use of the In this context, in 2000 the Bank adopted a very PCG effectively reduced the Bank's leverage to cautious approach to future transactions using promote agreed reforms by front-loading finan- the RRG structure. Then in 2002, the PBG in Ar- cial support for the project (World Bank 2002a). gentina was called when Argentina failed to service the outstanding bond. Rather than enforce the 60- In Colombia, when the PBG was issued, the gov- day period in which Argentina had to repay the ernment had been implementing a broad reform Bank for the guarantee to roll over, the Bank program supported by a Bank Financial Sector Ad- rescheduled the loan, causing the guarantee to justment Loan. The Bank's evaluation of the proj- lapse. The market immediately downgraded the ect observed that in this context, hybrid policy issue and also downgraded the RRGs in Thailand loan/guarantee operations might provide more and Colombia. policy leverage and better sequencing than stand- alone policy guarantee operations. Unlike direct Since then, the market has considered the Bank's Bank loans, where disbursement is based on ex- RRG a "dead" product, and in 2003 the Bank can- penditures, moreover, the timing of PCG-backed celed a proposed $180 million PCG to support the financing can be inappropriate. The Bank's self- Bolivia-Brazil Gas Pipeline Project because of lack evaluation of the Lebanon project found that of investor interest. Shortly thereafter, the Bank issue of the bond in Lebanon required the utility reaffirmed that "the RRG is not a viable structure to undertake a large foreign debt long before the from the Bank's risk management perspective," funding was needed: five years after the bond as it might affect the Bank's commercial borrow- was floated, only 50 percent of the funds raised ing terms.18 At the same time, however, IFC con- had been utilized. tinues to use the RRG instrument for its corporate bond PCGs. In retrospect, there appears to have The Bank'sdecision to extend repayment terms never been a clear consensus within the WBG as on the called PBG in Argentinaeffectivelyended to the appropriateness of the RRG. its abilityto use rolling, reinstatable PCGs. The Rolling Reinstatable Guarantee (RRG) mecha- Based on the Bank's relationship with govern- nism for PCGs was introduced on a pilot basis in ments, IBRDPRGshave helpedprojectsinhigh- 1999, and three PCGs were issued using it between risksectorswithuntestedregulatoryframeworks 1999 and 2001--in Thailand, Argentina, and reach financial closure. When two PRGs were is- Colombia. The benefit of RRGs is their ability to sued in Pakistan in the mid-1990s, the power sec- enhance credit terms by guaranteeing interest tor had only recently been opened to private payments on bonds on a rolling basis, thereby cov- participation, the country was still a high-risk in- ering the life of the bond if payments continued vestor destination, and Pakistan lacked a track to be made. record in long-term commercial finance. In this context, Bank engagement in the Hub (the first pri- Even at the time of the pilot, however, there was vate investment in power in the country) and Uch some dissension within the Bank about the value Power Projects played an important role in mobi- of the instrument as well as its potential risks to lizing commercial finance with 12- to 15-year ma- the Bank. Given difficulties in modeling and valu- turities to finance the projects. Although the ing the credit enhancement, RRGs were seen as guarantee projects in themselves were successful, being penalized by the market. This in turn was the government's broader PPP strategy in the seen as affecting the value placed on direct Bank power sector created an overcapacity in power gen- bond issues, thereby potentially raising the cost eration that eventually placed significant fiscal of borrowing for the Bank. In Colombia, for ex- burdens on the government. 4 0 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S In Romania, the fiscal 2005 power-distribution surers participated in the project, direct Bank en- PRG supported the first successful divestiture of gagement through the PRG was seen Romania's electricity distribution utilities. Prior ef- as critical for project implementation because forts to privatize the utilities had been unsuc- of the Bank's substantive ongoing dialogue with cessful partly because of uncertainties regarding the government. the policy framework and government commit- ment to the process. With the success of the PRG- Sponsors of the Ukraine/Russia Sea Launch backed privatization, several further distribution Project indicated that they would not finance the companies were privatized. project without the Bank PRG to deter the governments from actions that would prevent The 2003 Mozambique Gas Pipeline Project (in- the production and export of launch equipment. volving both IFC and MIGA) also represented a The Sea Launch PRG also introduced a dispute- large private investment in a sector in its initial resolution mechanism that proved effective in stages of development in a high-risk country. Leg- resolving problems between the government and islation in the gas sector had only recently been the sponsor. The mechanism also supported established; government agencies had limited ex- good governance, in that it provided investors perience in dealing with private sector projects; with an avenue to resolve problems other than and political uncertainties in the country per- bribing government officials. sisted. In this context, the sponsor found term fi- nancing difficult to secure and approached the IDA PRGs have supported the introduction of Bank to help complete the financing package. complex PPPs in high-risk countries using lim- ited IDA resources. Although just 10 percent of As a country's regulatory environment for PPPs im- total PPIs in low-income countries and MICs were proves, however, the added value of a PRG can be in Sub-Saharan Africa, 53 percent of IDA PRGs more questionable, given the greater likelihood were in the Region. The six IDA PRGs covered an that private providers will be in the market. Al- average of 10 percent of total project costs, which though the Bank had a long engagement in the is considerably lower than initially expected.19 power sector in Jordan, by the time the PPP was approved, the regulatory environment for pri- All six projects involved introduction of the private vate participation in the sector had been well es- sector into complex PPPs in high-risk countries. tablished. In this context, the distinction between The Vietnam Phu My Power Project, for example, the benefits of the PRG compared with PRI pro- was the first major competitively bid private in- vided by MIGA or other providers is less clear. frastructure project in the country, and it was pre- pared at a time when the legal and regulatory The additionality of IBRD PRGs has been its framework in the sector was still new; govern- long-standing policydialogue and the potential ment agencies had limited experience in dealing for dispute resolution. In all the power sector with complex contractual arrangements; and per- PRGs (Romania, Morocco, Pakistan, and Jordan), ceived country risks were high. The multicountry the Bank had had a long history of engagement West African Gas Pipeline Project involved devel- in the sector. This placed the Bank in a position opment of legal and technical agreements be- to help guide the development of a policy frame- tween four governments and the project sponsors work for engaging the private sector in power, ad- to support the transport of natural gas from Nige- vise on and help structure specific transactions, ria to Benin, Ghana, and Togo through an under- and be in a position to mediate disputes during sea route along the coast. The Hydropower Power implementation. In Romania, for example, the Project in Lao PDR was the world's largest private sponsor saw the Bank and a PRG as the preferred sector cross-border power project, and the Bank form of risk mitigation, because of the Bank's played an important role in the technical, eco- long-standing policy dialogue in the sector. In nomic, financial, environmental, and social ap- Mozambique, although several political risk in- praisal of the project. 4 1 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 In several large PPPs, IDA's engagement and rica Pipeline Project, the Bank built on its prior en- relationship with governments were critical to gagement in the energy sectors in Benin, Ghana, securing adequate financing for the project. In and Nigeria and advanced the reform process the context of limited access to commercial fi- through dialogue with the governments and key nance, complex undertakings in untested regu- stakeholders. In Lao PDR, through engagement latory environments, and the low-income, in the PRG, the Bank helped advance institutional high-risk status of the countries, IDA's PRGs played and governance reforms that were critical to im- an important role in securing adequate finance for provements in public revenue management. In the the projects. In Côte D'Ivoire, for example, IDA Bangladesh power project, the Bank was able to involvement was sought by the project sponsor help the sponsor renegotiate tariff agreements to to fill a financing gap after other sources of fi- ensure the financial viability of the project. nancing, including a possible increase in IFC's B- loan, had been exhausted. In Vietnam, the PRG Undersome concession contractssupported by helped secure commercial funding with a 16-year PRGs, governmentshave assumedabroadrange tenor at a time when the country had a non- of risks that have not yet been tested byevents. investment-grade sovereign credit rating and re- Whereas PRGs were intended to cover political stricted access to commercial markets. The risk, the line between political and commercial risk sponsors of the West Africa pipeline insisted on can blur, depending on the contractual obligations appropriate risk mitigation that eventually in- of the government that are being guaranteed. volved a combination of an IDA guarantee, a MIGA guarantee, and Zurich/OPIC insurance. The ex- The Bank's approach has been for the PRG to cover ceptionally large power project in Lao PDR would the minimum risks necessary to make the opera- also have been unviable for private lenders and tion viable, but as such has included a range of risks insurers without the Bank's engagement through beyond traditional PRI. For example, in cases where the PRG. According to Bank staff reports, the in- a PPP contract includes assured government pay- ternational lenders, Thai commercial banks, and ments to the private entity (such as a power pur- project sponsors all saw the role of the WBG as chase agreement), a guarantee that supports this essential in enabling the project to be realized. obligation effectively assumes commercial risks, because the guarantee could be called if the gov- Bank engagement through the IDA PRGs has ernment fails to make payment for any reason. provided a platform to further the policy envi- ronment for PPPs. The PRGs have provided an Moreover, several PRGs included coverage of nat- avenue for the Bank to be fully engaged in help- ural force majeure events that were not otherwise ing develop PPPs, even without Bank lending on insurable on the grounds that such coverage was the public sector side, and to further the overall essential to enable the project to move forward. In investment environment. Each of the IDA PRGs effect, however, this has supported a government was deployed in the context of a long IDA policy guarantee of events beyond its control. As cur- dialogue, aimed at establishing appropriate reg- rently structured, these risks are not priced into the ulatory frameworks for private sector participation premiums charged to the sponsors and, in any in the respective sectors. event, the government does not collect a share of the premiums. No claims have been made yet, but Engagement through the IDA PRGs provided the a natural disaster that triggers a call on the Bank's Bank with a platform to further the reform guarantee might prompt questions as to why the process, monitor implementation of specific trans- Bank supported government guarantees of events actions, and help address emerging issues and that were beyond the government's control. problems. In Vietnam, the electricity law passed in 2005 opened the sector for private participa- The PRG facilities have suffered from very low tion. Since then, three additional private energy utilization. Only three of the eight guarantee fa- projects have been established. In the West Af- cilities (see table 2.3) issued any guarantees. The 4 2 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S Table 2.3: Guarantee Facilities Supported by the Bank Approved Number of Project amount Product Fiscal guarantees Project ID Country name ($ millions) type year issued/value P038614 Moldova Pre-Export Guarantee Facility 30 Contingent loan 1995 None P045820 Bosnia and Emergency Industrial Restart Project 10 Loan 1997 26/40 million Herzegovina Deutschmarks P043434 Ukraine Pre-Export Guarantee Facility 120 PRG 1997 None P051602 Albania Private Industry Recovery Project 10 Loan 1998 24/$8.7 million P057893 Russian Federation Coal & Forestry Guarantee Facility 200 PRG 2001 None P063683 Africa Regional Trade Facilitation Project 128.7 Loan 2001 47/$54 milliona (10 countries) P089120 West Africa WAEMU Capital Market Development 70 RPG 2004 None (8 countries) P088923 Peru Guarantee Facility 200 PRG 2005 None Source: World Bank data. a. IDA account as of September 2007. facility in Bosnia issued 26 guarantees worth DEM ects have drawn lessons along the lines that prepa- 40 million (against a target of 75 million German ration of such projects should involve an in-depth deutschmarks). In Albania, 24 guarantees were is- market analysis prior to commitment, apparently sued worth $8.7 million. In both cases, demand none has been successful in accurately doing so. for the guarantees dropped sharply after a few Although surveys have been conducted at ap- years, and the facilities were closed. In Moldova, praisal, actual demand has not turned out to be the facility was closed after 22 months without is- as strong as anticipated, indicating some weak- suing any guarantees. Lack of demand was at- nesses in the survey methodologies. Although tributed to the high pricing of guarantees, as well the private sector has generally indicated that as competition from comprehensive guarantees political risk is a constraint to investment, it has provided directly by the government. been less clear whether this was the "binding" con- straint. Thus, despite the availability of PRI through The Ukraine Pre-Export Facility was cancelled in the facilities, multiple other factors emerged as 1999 without becoming effective after the Ukrain- greater constraints to investment. ian Parliament failed to ratify the project. The $200 million fiscal 2001 Russia Coal and Forestry The BOAD project reflected a promising model to Guarantee Facility was closed in 2005 without is- market a range of political mitigation instruments suing any guarantees because of a lack of de- for small and medium-size infrastructure proj- mand for the noncommercial risk guarantees. ects--including MIGA PRI, Bank PRGs, and African The WAEMU Facility, implemented through BOAD, Development Bank (AfDB) PCGs--under one also failed to market any guarantees and has been roof. According to staff interviews, however, lim- cancelled. No guarantees have been approved ited institutional capacity in BOAD undermined its under the fiscal 2005 Peru Facility to date. success. Some lessons from the experience were that a $50 million project limit was too small for Inaccurate estimates of demand and lack of IDA PRGs and for several potential infrastructure readinessforimplementation have undermined projects; the requirement of a counter-guarantee the facilities. Although evaluations of these proj- proved to be an impediment, especially for the 4 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 small and medium-size infrastructure projects vestment and MIGA's PRI have been raised, be- contemplated for the facility; and MIGA PRI proved cause MIGA is able to issue the same guarantees to be the most flexible instrument, with two proj- on an individual basis. Bank facilities have been ects identified and applications filed. justified in that they help stimulate overall investor interest in a particular country by advertising the In the case of the Peru Facility, a Quality Assurance government's commitment to the rule of law, Group panel found several weaknesses in the ap- rather than an individual investment supported praisal of the project. Subsequent changes in the by MIGA's PRI. To date, however, the facilities government's PPP strategy as well as a recent in- have not proved particularly effective terpretation by the Ministry of Finance that the use in stimulating investor interest. The trade en- of the facility would imply an expenditure in the hancement facilities supported by IFC might also national budget once the guarantee is issued have be substituted for the short-term guarantee fa- also undermined use of the facility. cilities established by the Bank. At the same time, the Bank facilities are designed to fill a gap in ATI suffered froma range of initial design defi- coverage of PRI for long-term local debt that ciencies. The IDA-funded ATI was established in MIGA is not able to cover. 2001 as a self-standing multilateral insurance agency. It was owned by participating African Nevertheless, it is apparent that a more coordi- states and empowered to issue its own insurance nated WBG approach to the creation of guaran- products (and not IDA PRGs) to facilitate trade and tee facilities should be established. The Quality investment in Africa. ATI experienced poor initial Assurance Group report on the Peru Facility, for performance, however. It suffered from a range example, noted that a more proactive approach of design flaws, erroneous demand assumptions, could have been made to engage MIGA/IFC ex- and weaknesses in institutional capacity. By 2005, pertise in the design of the operation, given that it had issued just $110 million in insurance poli- MIGA had generally negative past experiences cies, compared with an initial target of $1,280 with guarantee facilities and that IFC had limited million by 2011. expectations of demand based on its experience in the infrastructure sector in the Region. Problems included lower demand than expected from private investors for long-term projects and Potential for use more demand for short-term, comprehensive trade The use of Bankguarantees has fallen far short insurance from public/parastatal corporations, of expectations, yet some Bank guarantee in- which it was prohibited from covering. Other prob- struments have good potential to be deployed lems were the capacity constraints arising from under certain circumstances. Initial expectations capital reserve requirements to back long-term were that the Bank would issue some $1­2 billion policies and limited membership that precluded worth of guarantees a year, but it has issued only coverage of investors from most of the continent's $2.8 billion in guarantees during the last 18 years. major economies. In 2006 ATI was restructured to Total project costs (as a proxy for investment improve its issuing capacity and broaden its prod- flows) supported by Bank PRGs averaged $500 uct range and potential market. In 2007 it issued million a year during the period, compared with 17 new policies, exceeded its premium targets, a total flow of $130 billion a year in FDI to devel- and paid its first claim (an essential element for an oping countries during these years. insurer). To date, demand has been driven mainly by public/parastatal enterprises. Although the share of investment catalyzed by the Bank is small, as discussed below, the objective of There has also been some question about the Bank instruments is to catalyze additional invest- extent to whichthe facilities overlap withMIGA ment to that already flowing, which represents a and IFC. Concerns about the relationship be- largely unquantifiable market. The limited use of tween the guarantee facilities for long-term in- 4 4 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S guarantees is driven by a range of both internal and assistance in appraisal of complex PPPs. Although external factors, which are discussed below. some of the facilities were set up to support short- term trade finance operations, such facilities have The scope forIBRD PRGsappearsto be limited. been more conducive to private sector opera- Of the seven IBRD PRGs, two were in a blend tions. Within the WBG, IFC has also developed fa- country (Pakistan) and one was an enclave guar- cilities to meet such demand. antee in an IDA country. Thus, just four PRGs have been issued in IBRD-only countries since the Continued demand is likely to remain for IDA product was introduced in 1994. Although inter- PRGsinhigh-riskcountries.In some transactions, nal factors discussed below have constrained the an IDA PRG is likely to remain the instrument of use of the instrument, to some extent this also re- choice. Situations in which an IDA PRG has been flects the lower risk perception of doing busi- sought include those where the Bank's prior en- ness in MICs; their greater access to commercial gagement in the sector or relationship with the gov- finance; and their more established and tested reg- ernment placed it in a position to both establish ulatory frameworks. These factors have created and help maintain a favorable policy environment environments where private PRI providers or spe- for the private sector; either country or sector risk cialized public providers such as MIGA are able to levels were sufficiently high to prevent adequate meet the demand for PRI. commercial financing (including that from IFC) from being made available at adequate maturities Seventy-five percent of MIGA's business to date has or reasonable cost; and the amount of financing re- been in MICs. There is less need for close Bank quiring PRI has been such that MIGA and other pri- engagement in sector policy reform, appraisal of vate providers exceed their exposure limits. PPP projects, and continued monitoring and en- gagement in the implementation of a project the Experience has shown that PRGs also offer a PRG affords. Use of the IBRD enclave PRG has also potentially effective tool to support regional been very limited. Whereas the Board author- integration in Africa. The West Africa Gas Pipeline ized up to $300 million a year for enclave guar- Project, for example, supported a 20-year pro- antees, just one has been approved, for a total of gram for regional integration of the power energy $30 million. Some projects, such as the Lao PDR systems by helping develop a comprehensive Nam Theun 2 Project, which may have qualified commercial, legal, and regulatory structure as an IBRD enclave operation, were instead sup- through a treaty between states and a detailed in- ported with IDA PRGs. ternational project agreement between the states and the project sponsor. An important role for IDA There is also unlikely to be significant scope PRGs is likely to remain in such circumstances. forguarantee facilities.Given the ineffectiveness of Bank guarantee facilities to date, as well as their The potentialalso existsforPCGsto help coun- potential overlap with MIGA and IFC products, the tries regain access to capital markets during approach to Bank guarantee facilities should be marketdownturns, aswellasto introduce well- cautious. Experience has revealed limited use of performing countries to capital markets. PCGs these facilities for a range of reasons. One of the have proved useful in helping countries regain main objectives of the facilities was to improve the access to markets as well as in introducing new bor- overall investment environment by advertising a rowers to commercial markets. When capital mar- government's commitment to the rule of law; kets are liquid there is likely to be less demand however, there is no evidence that any of the fa- for PCGs, as has been the case until recently. At cilities have been effective in doing so. times of financial crisis, however, demand for PCGs has risen. Bank engagement through PCGs There is also some question about the value of offers the potential to help channel the flow of wholesaling a product whose main attributes are commercial funds into underserved sectors. close Bank policy dialogue in sector reforms and 4 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 A further market for PCGs exists among well- for World Bank guarantees were overly optimistic performing IDA countries. IDA countries have not based on the rapid growth of the mid-1990s, been eligible for PCGs on the grounds that they lack which was not sustained. Since 2005, however, the creditworthiness necessary to access interna- there has been renewed growth in the number of tional markets. However, some IDA countries, such PPPs, including investments by local and regional as Ghana and Sri Lanka, have recently accessed in- investors in infrastructure projects. ternational markets without the benefit of credit enhancement. It would seem that the PCG could ConcentrationofBankguaranteesonthepower be an effective tool to assist higher-income, well- sector reflects some difficulties in developing performing IDA countries that are on the verge of commercially viable PPP projects in other sec- accessing markets directly and in which IDA assis- tors. The Bank's guarantee policy framework has tance is constrained by IDA allocations in obtain- consistently emphasized the potential applica- ing more favorable market credit terms. tion of PRGs to sectors outside power, including water, transport, and telecommunications. Nev- PCGs also offer the potential to help mobilize ertheless, the Bank's guarantees have been heav- local capital markets for public investment needs. ily concentrated in the power sector (see figure In Jordan, the bond issue supported by the PCG 2.13). was able to mobilize domestic foreign currency savings in the form of deposits held by expatriate To some extent, this is because the power sector, or returning Jordanians working overseas. A con- particularly generation, has been more conducive tinued need exists for mobilization of local cur- to PPPs. According to the PPP database, 40 per- rency capital markets. In countries and regions cent of all PPPs have been in power. Limited com- with more developed capital markets, potential ex- mercially viable prospects in the water and ists for PCGs to enhance local currency financing transport sectors have constrained the use of of public infrastructure projects. guarantees in these sectors. Constraints in high- ways, for example, have included political sensi- Constraints to use tivity to public resistance to tolling; high-profile The PRG'slastresortnature andthe lower-than- private toll-road project failures, which have made expected level of commercial private invest- investors and governments cautious; a limited mentininfrastructure have narrowed the scope number of potential roads, particularly new roads for BankPRGs. Bank PRGs have aimed to fill the in developing countries, with adequate project gap left by various PRI providers. Political risks economics to attract private financing without can be mitigated by a variety of means, including substantial government contributions; and a com- the ability of large international investors to absorb plex legal and policy framework required for a con- country risk on their own balance sheets, public cession that has proved difficult to achieve (World insurance agencies in large developed countries, Bank 1996). Similar difficulties have been en- bilateral investment treaties between countries, and countered in the water sector. multinational providers such as MIGA. Some 15 Bank PRGs in nonpower sectors were Within the overall market, the Bank's niche has planned and subsequently dropped. In Jordan, for been in large PPPs in infrastructure. This market example, a planned PRG to support the proposed has declined since a peak in the mid-1990s. Al- Disi-Amman water pipeline that had been in though there was a surge in investments in the preparation for a decade failed to materialize mid-1990s, particularly in the power-generation when the bid tariff was much higher than antici- sector, following the 1997 financial crises, the col- pated and the government cancelled the tender. lapse of Enron in 2001, and a series of difficulties A planned PRG in Croatia to support develop- in existing private participation in infrastructure, ment of a highway was also dropped when the there was a sharp drop in the volume of transac- Bank's economic and financial appraisal found tions taking place. To some extent the expectations the project to be unfeasible. Just 2 percent of 4 6 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S Figure 2.13: PPI Projects in Developing Countries by Sector, 1990­2006 400 350 300 250 200 Projects 150 100 50 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Fiscal year Energy Telecoms Transport Water and sewage Total Source: World Bank and Public-Private Infrastructure Advisory Facility, PPI project database. Note: PPI = private participation in infrastructure. MIGA's business has been in the transport sector the commercial orientation or legal framework and 1.5 percent in water. for borrowing on capital markets. Among com- mercially operated utilities, a further constraint The value ofPCGs has been undermined byliq- remains a potential mismatch between local cur- uid capital markets since 2002. Several external rency revenues and foreign currency debt raised factors (as well as internal factors, discussed by the PCG that exposes the state-owned enter- below) account for the absence of PCGs since prise to foreign exchange risks. These constraints 2001. In particular, liquid capital markets have are dynamic, however, and the recent reduction enabled IBRD countries to access markets di- in the liquidity in international markets can in- rectly on favorable terms. In this context, the crease the demand for PCGs. credit-enhancing value of the PCG is low, and the guarantee fee plus the underlying loan spread The pricing of Bankguarantees has not been a makes obtaining a PCG cost-ineffective. constraint. Unlike MIGA, the Bank does not price PRGs for sector or country risk, because given the By their nature, PCGs are also largely restricted to mandatory government counter-guarantee, the a relatively narrow band of countries that are Bank passes the risk of default on the commer- close to accessing markets directly. If countries lack cial loan to the government. The risk to the Bank creditworthiness, then provision of a PCG for is that the government will not repay the loan extended-term maturities is unlikely to make a under the counter-guarantee. This risk thus rep- difference. But if countries have full access to resents the same one as a normal lending oper- markets, then they have limited need for the ation. credit enhancement provided by the PCG. PRGs are therefore offered across countries and Within countries, many utilities that are potential sectors according to a set schedule of charges PCG clients still operate with soft budgets and lack based on loan equivalent pricing (table 2.4). This 4 7 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Table 2.4: IBRD/IDA Pricing of Guarantees (2008) IBRD IDA Fee type PRGs PCGs PRGs Upfront charges Front end fee 25 bp 25 bp n.a. Initiation fee (1) 15 bp on the guaranteed n.a. 15 bp on the guaranteed amount or amount or $100,000 $100,000 (whichever is higher) (whichever is higher) Processing fee, (2) Up to 50 bp of the n.a. Up to 50 bp of the guaranteed amount guaranteed amount Recurring charges Guarantee fee 30 bp per annum 30 bp per annum 75 bp per annum (on the maximum (on the present value of (on the maximum aggregate disbursed aggregate disbursed and the guarantee exposure) and outstanding guaranteed debt) outstanding guaranteed debt) Standby fee (3) n.a. n.a. 10 bp per annum (on the maximum guaranteed debt committed but undisbursed) Source: World Bank. Note: For all private sector borrowers, that is, only applicable to PRGs. Determined on a case-by-case basis. Exceptional projects can be charged more than 50 bps of the guaranteed amount. Data for guarantees approved in fiscal 2008. bp = basis point; IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; n.a. = not applicable; PCG = Partial Credit Guarantee; PRG = Partial Risk Guarantee. has resulted in pricing that is generally competi- As discussed below, because the price for PRGs is tive in the market and has not been a constraint offered to private sector clients, there is some con- to the use of Bank guarantees. According to a cern that the Bank's pricing structure, based on survey, just 10 percent of Bank staff interviewed the loan equivalency, can distort the market for identified the cost of a guarantee as a factor in a PRI. Moreover, although some efforts have been project being dropped (compared with 50 percent made to share guarantee fees with the host gov- in MIGA and 80 percent in IFC). In three of the ernments, governments are currently not pro- four joint Bank-MIGA projects in which both the vided with any portion of the guarantee fee. The Bank and MIGA issued guarantees, the Bank's option to compensate governments might be fur- PRGs were priced lower than MIGA's, despite the ther explored, however, because it could help fact that they offer similar or higher-risk coverages offset costs associated with the contingent liabil- (see below). ity of the counter-guarantee; compensate gov- ernments for the risks they take with respect to The pricing structure has led to some anom- factors out of their control, such as force ma- alies, however, and at current IBRD rates, IBRD jeure; and reduce disincentives for government PRGs are cheaper than IDA PRGs. The Bank's support for PRGs. loan equivalent pricing for PRI products can place below, at, or above the PRI market. In the As most guarantee projects have been in power fourth joint project with MIGA--the IBRD en- generation, Bankguarantees have involved ex- clave PRG in Mozambique, which was issued at tensive due diligence requirements. Significant a time of relatively high IBRD lending rates--the social and environmental assessments have been Bank's PRG was priced significantly higher than associated with all the Bank's guarantee proj- MIGA's PRI. ects. This has been necessitated by the nature of 4 8 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S the projects and the sectors of the PRGs--com- value additions, rather than the guarantee mech- plex, large infrastructure projects, mostly in power anism itself. generation, some cross-border operations, and so on. In the Lao PDR project, for example, ex- As noted, nearly all the Bank's PRGs have sup- tensive analysis and mitigation of the social and ported large power generation projects, with con- environmental impacts included resettlement of sequent due diligence implications. The Hub residents, livelihood restoration, wildlife man- Power Project, for example, which took more agement programs, protection of natural water- than 10 years to prepare and absorbed some 25 sheds, and mitigation of downstream impacts. person-years of Bank resources, was the first com- mercially financed project in a sub-investment- Although these efforts represent a key added grade developing country and involved extensive value of Bank engagement, they have also added policy dialogue, sector work, and coordination costs to guarantee operations that are not re- with some 40 international financiers, the spon- lated to the use of the guarantees themselves. In sors, and the government (World Bank 1997a). some cases, it is apparent that application of the The Lao PDR Nam Theun 2 Power Project, which same standards required of public investment cost more than $5 million in Bank administrative projects might be excessive in private sector proj- costs, involved considerable work on policy re- ects. For example, in some power projects, due form, technical studies, and consultation, as well diligence was conducted on "linked" projects, as extensive social and economic appraisal. such as construction of roads, transmission sys- tems, and pipelines that were beyond the control As discussed above, although the extent of due of the project sponsor. As discussed below, this diligence conducted in Bank PRGs has been has led to exceptionally long preparation times driven by potential reputational risks to the Bank for some PRGs. as well as the need to ensure viability of the proj- ect in the public interest, it also provides a strong Such efforts, though they enhance the prospects disincentive for the private sector to work with that adverse social and environmental impacts the Bank.20 will be minimized, have added to the costs and time required for processing a transaction. That Additional internal processing steps have pro- provides some disincentives for private sector vided disincentives to pursuing guarantee op- clients to work with the Bank. According to staff erations. The long processing times, extensive interviews, these factors have led some clients to procedures, and uncertainties involved in deal- indicate that they would never work with the ing with private companies remain disincentives Bank again. to Bank staff in pursuing guarantee operations. All PRGs have been required to undergo an Op- Some of the Bank's PRGs have taken an ex- erations Committee review, which has added ceptionallylong time to prepare, although this additional effort to their processing. The inabil- has reflected the nature of the underlying proj- ity of the Bank to issue a term sheet that outlines ects.According to a survey, 45 percent of Bank staff the structure of a PRG for a particular project indicated that the long process- without senior management approval (and con- ing time was a factor behind a guarantee project sequent internal clearance procedures), for ex- being dropped. The average processing time for ample, has constrained timely response to an IBRD PRG was 18.8 months, at an average ad- demand. ministrative cost of $416,000 per project; IDA PRGs have averaged 20 months to prepare, at an PCGs have an additional internal constraint in average cost of $800,000 (excluding the Lao PDR that the Bank is required to inform the Board Nam Theun 2 Project outlier). The high process- prior to initiating discussions with government ing costs have been largely caused by the com- clients. The caution exercised in the approval of plexity of the underlying projects and the Bank's guarantees initially reflected the pilot nature of the 4 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 operations. Now, however, although the number circumstances, the last resort principle is of limited of guarantees has been limited, the Bank has had relevance. As discussed below, rationalizing of pric- 18 years of experience with the products. The ing of PRM products across the WBG could help complexity of the projects has caused even further reduce the need for the hierarchy principle. scrutiny. In the Lao PDR Nam Theun 2 Project, for example, the Board requested a semiannual The requirement for acounter-guarantee for all progress report on the project's implementation, IDA PRGs is an advantage in some situations, and two guarantee projects have been subject to but a constraint in others. The IBRD Articles of inspection panel investigations. Agreement require IBRD guarantees to be backed by a sovereign counter-guarantee, but the IDA Ar- According to staff, Bank management has also ticles do not. Instead, requirement of a counter- not sent clear signals as to whether guarantee guarantee for all IDA guarantees is a Board operations should be pursued, with some re- decision. Although in some circumstances there gions actively seeking opportunities and others are clear benefits to requiring a counter-guaran- not. It is clear that more exposure and discussion tee--such as the need to fully engage the Min- of the potential use of guarantees in the Bank is istry of Finance in a project--in other cases the warranted. counter-guarantee requirement may be less nec- essary and serves to discourage government en- The use of IDA PRGs was constrained by full gagement because of its contingent liability country allocation requirements. When IDA implications. PRGs were introduced in 1997, issue of a guar- antee absorbed IDA resources equivalent to the Internal incentives favor the use of IBRD lend- full amount of the guarantee. In the first six years ing over PCGs. Because IBRD/IDA does not lend of the IDA program, between 1998 and 2004, only directly to the private sector, IBRD/IDA PRGs can- three IDA PRGs were issued. This partly reflected not be replaced by other IBRD/IDA lending prod- the reluctance in the Bank to use IDA allocations ucts. In contrast, IBRD loans are a direct substitute to support private sector projects rather than for PCGs, in that a public investment project can public investment. In 2004 the IDA allocation for be financed by either Bank funds or commercial a PRG was reduced to 25 percent of the value of bank funds supported by a PCG. According to the guarantee, and in the three years since, an ad- staff, internal incentives strongly favor direct IBRD ditional seven IDA PRGs have been approved (al- lending over use of a PCG. In some circumstances, though three are not yet active). moreover, direct Bank lending will offer the Bank greater leverage over design and implementa- The last resort principle has some limitations. tion of the project. According to some Bank staff, the designation of Bank guarantees as "last resort" instruments has PCGs offer the Bank an opportunity to support also unduly constrained the use of Bank guaran- the same project with fewer Bank financial re- tees. In some cases, it is argued that whereas sources. By establishing the link between com- MIGA can cover PRI, the Bank's PRGs would give mercial financing and public investment projects the Bank an opportunity to build on its policy di- in member countries, PCGs also offer a long-term alogue and achieve broader reforms for private sec- contribution. It is apparent that there is no con- tor development, and that the PRG would be a sensus in the Bank about the relative merits of better fit for the project. The last resort principle, PCGs versus direct Bank lending in different cir- however, requires deployment of MIGA PRI first cumstances. The issue warrants greater discussion and effectively prevents the Bank's engagement. within the Bank and clearer guidance. In addition, there are some cases where clients have The Bankhas limited expertise in the financial a preference for a certain instrument or institution structuring component of guarantees. In the based on the circumstances of the situation. In such Bank, guarantee projects have involved teams 5 0 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S similar to those of regular loans that have worked the technical and commercial success of the proj- on the underlying project, as well as "guarantee ect. Indeed, the rationale for engaging the private specialists" who worked on the financial struc- sector is that there are built-in incentives for the turing component of projects. Expertise on the private sector itself to conduct adequate appraisal financial structuring of guarantees has been con- and ensure project success. Although this route fined to a limited number of staff, however. In the would likely increase the use of PRGs, it would also past, expertise was concentrated in about 10 staff bring the Bank's PRG closer to instruments offered in a central unit within the infrastructure vice by MIGA and private PRI providers, as well as presidency (supplemented by some Regional staff eliminate some of the main value added of Bank experienced in the use of guarantees) who worked engagement. with Regional sector staff as co-task team leaders of PRG projects. Over time, however, the staff of At the same time, however, it is apparent that the the central unit moved to Regions or retired and Bank's public sector approach in appraising proj- were not replaced. ects or requiring private sponsors to mitigate impacts that go beyond the project might be ex- In 2006 a decision was taken to "mainstream" cessive. According to staff interviews, some pri- guarantees, with Regional staff expected to iden- vate sector clients have indicated that they would tify, develop, and process guarantee operations. never work with a Bank PRG again. It therefore The central unit in the Sustainable Development seems that a better balance is needed between Network anchor now retains only a few staff who the PRG as an investor-friendly instrument that provide financial expertise in support of PRG prod- enhances the flow of investment through PRM ucts as well regular Bank lending products. The and the PRG as an instrument to help broaden Bank's Treasury vice president is now expected to development objectives. undertake several functions related to guaran- tees: act as the repository of knowledge on pol- IFC Guarantees icies, pricing, legal, systems and accounting; undertake outreach and training activities; and Patterns of use support development and integration of new prod- IFC's guarantees have awide range ofpotential ucts. The credit risk assessment of guarantees is applications. IFC's Financial Instruments Guide- undertaken by the Credit Risk Department. lines state that IFC's guarantees may enable pri- vate sector clients to-- The Bank'sPRGshave evolved into instruments withobjectiveswellbeyondthe provisionofPRI. · Access long-term local currency financing. As reviewed above, PRGs have been used to en- · Increase exposure to sectors deemed strate- gage the Bank in structuring complex PPPs; en- gically important for development, such as suring the technical, financial, and economic SMEs, residential mortgages, student loans, viability of the projects; advancing policy reforms; and trade finance. and ensuring close compliance with social and en- · Access international investors for the first time vironmental guidelines and continued engage- or at a time when investors may otherwise be ment in the project during implementation. The reluctant to extend credit to corporations within instrument has thus adopted a broad range of de- the client's sovereign jurisdiction. velopment objectives, rather than strict provi- · Introduce new instruments (such as bonds, sion of PRI to catalyze the flow of private commercial paper, note issuance facilities, and investment. swaps) in local or international financial mar- kets, thus diversifying financing sources. At its most limited, a PRG would only be con- · Access local or foreign currency funding cerned with the extent to which the government on terms and with maturities otherwise not adheres to obligations under a project, and less available. with furthering policy reforms or helping ensure 5 1 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Figure 2.14: IFC's Total Net Commitment (excluding GTFP), by Country Risk Classification, Fiscal 1990­2007 350 300 250 million) 200 (US$ 150 guarantee 100 Net 50 0 1990 1991 1992 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Fiscal year Low Medium High Source: IFC data. Note: Data for 1993 not available. GTFP = Global Trade Facility Program; IFC = International Finance Corporation. · Expand a client's activities in a country or mar- Guarantees have been growing rapidlybut still ket segment where it may be close to its ex- account for only a small portion of IFC's total posure limits by taking on some of the client's financing (figure 2.14). Between fiscal years 1990 portfolio risk. and 2007, IFC committed 196 guarantee opera- tions; 53 of these were under its GTFP. Guarantees accounted for 5.7 percent of IFC's total commit- Figure 2.15: Distribution of IFC Guarantees by ted portfolio as of June 30, 2007. Sector (net commitment), Fiscal 1990­2007 The average growth of the guarantee portfolio Extractive industries over 1990­2007 was 31 percent annually from a 4% Health and very low base versus 10.5 percent for IFC as a Financial sector education Power 53% 3% whole.21 The rapid growth in recent years has 1% been driven by GTFP, whereas the level of de- Telecom ployment of IFC's traditional guarantee products Infrastructure 8% has been stagnant. Guarantees have been con- 16% Water centrated in the financial sector (see figure 2.15), 6% which accounted for 71 percent of all guarantees Transport 1% (including GTFP, whose guarantees are all in the Agriculture financial sector). 8% Manufacturing, services, and There has been limited use of guarantees in the tourism 16% infrastructure sector, however. Although poten- tial exists for the application of guarantees in sit- Source: IFC data. Note: This excludes GTFP. IFC = International Finance Corporation; GTFP = Global Trade Facility Program. uations of local-currency-earning infrastructure projects, IFC has done only 10 guarantee trans- 5 2 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S Table 2.5: IFC's Local Currency Financing Fiscal year $ millions 2004 2005 2006 2007 IFC loans 3,732 4,541 4,968 5,642 IFC total local currency financing 481 820 1,320 1,650 IFC local currency guarantees 46 55 233 137 Growth in loans (%) 22 9 14 Growth in local currency financing (%) 70 61 25 Share of local currency in loans (%) 13 18 27 29 Share of guarantee in local currency financing (%) 10 7 18 8 Source: IFC data. actions in the infrastructure sector (8 percent of swap-based derivative transactions. IFC's Treasury the volume of guarantees, compared with an 18 introduced a series of initiatives in 2007 that pro- percent share of infrastructure in IFC's overall vide incentives for local currency financing through portfolio). Two of these were in the transportation loans or swaps by reducing the pricing of such local sector, five with water utilities, two in power, and currency loans by 30 basis points. No similar in- one in gas distribution. centives have been allowed for the provision of local currency financing through guarantees. IFC's guarantees have mainly supported local currency financing. In three-quarters of IFC's IFC guarantee instruments have focused on guarantee investments, the underlying asset was Africa, reachedSMEs, andembodiedinnovation. denominated in local currency (excluding GTFP A third of all IFC guarantee projects were in Africa, transactions). IFC's guarantees have helped clients including 54 percent of GTFP guarantees. Ex- that are earning local currency--particularly in cluding GTFP, guarantees were distributed broadly sectors such as infrastructure, housing, educa- (see figure 2.16) into East Asia and Pacific (27 tion, or nonexport-oriented SMEs--obtain fi- percent), Europe and Central Asia (18 percent), nancing without incurring currency mismatches. Latin America and the Caribbean (18 percent), and In countries where derivative markets are nonex- South Asia and Africa (14 percent each). Guar- istent or issuing IFC local currency bonds is not fea- antees have also helped indirectly reach micro, sible, guarantees are the only instrument available small, and medium-size enterprises (MSMEs) by for IFC to provide local currency financing. Of the facilitating access to long-term local currency fi- seven countries in which IFC guarantees enabled nancing by leasing, microfinance, and consumer local currency financing in 2007, long-term local finance companies. This had enabled them to ex- currency swap markets were not available in five. pand their MSME business. With limited use, however, guarantees have made In about two-thirds of all guarantee transactions, only a marginal contribution to the rapid growth the ultimate clients were MSMEs, and about half in IFC's local currency financing. Although IFC of these operations were in Africa. In about a local currency financing reached almost 30 percent quarter of the cases, SMEs were direct clients of of total loans in 2007, guarantees accounted for less IFC's guarantees, mainly through the Africa En- than 10 percent of local currency finance (see terprise Fund.22 A third of IFC's guarantee trans- table 2.5). The fast growth in IFC's local currency actions have also had important innovation financing has been largely driven by growth in aspects, either by introducing a new instrument 5 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Guaranteestend to combine the use ofadvisory Figure 2.16: Distribution of IFC Guarantees services, but mobilization of advisory service by Region (net commitment), Fiscal funding has often added to transaction costs. 1990­2007 One-fifth of IFC's guarantee operations involve the use of advisory services. Part of the reason for this Sub-Saharan South Asia relatively high level of occurrence is that guaran- Africa 14% 14% tees often introduce investors to new, unfamiliar sectors where risks are high. In such circum- Middle East and stances, advisory services can be a powerful tool North Africa in building capacities and mitigating risks. Guar- 9% antee operations in the education sector in Africa for energy efficiency, microfinance, leasing, and SME programs tend to be combined with large East Asia Latin America components of technical assistance. Advisory ser- and Pacific and the 27% vices have also been used by IFC to implement re- Caribbean forms that can be supported with guarantee 18% operations, as in the case of Mexico's PPP program. At the same time, the mobilization of advisory ser- Europe and vices to complement IFC guarantees has often Central Asia been a time-consuming process, which tends to 18% add significantly to transaction costs. Source: IFC data. Note: This excludes GTFP. IFC = International Finance Corporation; GTFP = Global Trade Facility Program. IFC's guarantee operations have involved part- nerships. About a quarter of all guarantee oper- ations involved partnering with other members to the market or introducing new clients or in- of the WBG or other national or bilateral devel- vestors to companies (Meddin 2005). Each of opment institutions. A number of corporate bond the Bank Group's six strategic themes is repre- or municipal bond guarantees were provided in sented in the IFC's universe of guarantee trans- risk-sharing arrangements with national devel- actions, with a particularly strong focus on Africa opment finance institutions, as in the case of and innovation. IFC's bond enhancements in Saudi Arabia and South Africa. Partnerships helped combine com- Guarantees have been deployed by IFC in cri- plementary strengths and allowed for risk sharing. sisand postcrisissituations. A quarter of all guar- At the same time, a large number of players have antee transactions were in crisis or postcrisis tended to fragment financing, adding to com- situations. The instrument has been used to re- plexity and increasing overall transaction costs, es- place IFC dollar-denominated loans with local pecially in the case of smaller projects. currency loans as part of balance sheet restruc- turings. Six projects in Pakistan were restructured Guarantees have complemented other IFC in- following the 1996 crisis. Similar restructurings struments.About half of IFC guarantee operations were done in Turkey and in Thailand following were with repeat clients, suggesting comple- crises in these countries, although on a more lim- mentarity over time between the deployment of ited, case-by-case basis. In Lebanon, IFC used IFC's nonguarantee instruments and the use of guarantees to support Lebanese banks in their IFC's guarantees. IFC's ability to take risks that oth- efforts to continue to play their regional role ers are unwilling to take on their own, as implied during the recent crisis there. Guarantees have by the nature of the guarantee instrument, has also been useful instruments for companies in often been based on IFC's comparative advantage distress to mobilize working capital in the wake in knowing the client through prior investment of economic crises. and advisory relationships. In the case of GTFP, for 5 4 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S example, IFC is leveraging its extensive global riod. Prior to 2000, guarantee operations tended network of relationships with financial institu- to have both lower incomes and lower adminis- tions in emerging markets. trative expenses than loans, as a percentage of av- erage outstanding balances. Since 2001, although Guarantees also often come packaged with other guarantee fees tended to exceed loan spreads, IFC instruments that have enhanced the value of guarantees became more expensive than loans to the guarantee. In 17 percent of the cases, guar- process. The higher administrative expenses for antees were part of a larger financial package, guarantees reflect greater complexity, smaller IFC which included IFC loans and/or equity. investment size, and a higher percentage of can- celled projects than in the case of loans. Effectiveness and additionality IFC's guarantees have helped beneficiaries ex- In terms of development outcomes, guarantee pand beyond current client or country exposure operations do not show statistically significant limits, particularly in trade finance. The demand differences with IFC's averages, according to De- for enhancement may reflect how close a bank is velopment Outcomes Tracking System results. In to country or client limits and may have only an terms of portfolio quality, the guarantee projects indirect relationship with country or client risks. had the same average credit risk rating in 2007 Banks often demand enhancement for clients as loans. they are familiar and comfortable with on the basis of long-term relationships. This can be seen IFC'sguaranteeshave helped extend maturities, in the case of the trade facilities, where IFC has butthe impacton all-in-cost23 offinancing isless provided enhancement services to banks that clear.IFC's guarantees have typically extended ma- have been working together and are familiar with turities beyond those available in local markets. each other. For example, the loans extended under one of the African school facilities have maturities of 3­5 IFC's guarantees have successfully introduced years, compared with the average 6- to 12-month new financial instruments to clients. In Mo- financing offered prior to IFC's involvement. rocco, IFC guaranteed for the first time a long-term subordinated loan, with similar characteristics as However, the impact on all-in-cost of financing is Tier II capital for a nonprofit microfinance insti- less clear. IFC characterizes the product as al- tution. IFC facilitated the first-ever bond issuance lowing the borrower--in most cases--to achieve by a nonprofit institution in Peru and the first a lower all-in cost, but it is very rare that project corporate bond in Saudi Arabia. In South Africa, documents attempt a comparison of the all-in an IFC guarantee also supported the first long- cost to the client with the guarantee against the dated municipal bond, issued by the city of Jo- client without the guarantee. Anecdotal evidence hannesburg. In Russia and the Bal-tics, for the first is sometimes presented that shows that banks time IFC guarantees helped introduce mortgage- often do not price the full impact of IFC guaran- backed securities. The IFC-enhanced Johannes- tees and pass the entire guarantee fee on to the burg municipal bond and Saudi Arabia's first client. In many instances, particularly in IFC's di- corporate bond were followed by others within rect investments in Africa, IFC has been putting a fairly short period of time, suggesting a demon- pressure on the banks to fully price the effect of stration effect. IFC's credit enhancement. IFC's policies ask staff to look beyond the guarantee fee in the case of On average, guarantees have been less prof- dollar financing to ensure that IFC's risk is priced itable thanloans, buttheirdevelopmentsuccess correctly in case IFC needs to access the local mar- rate has been similar to that of loans. A com- ket. This requirement may be extended more parative profitability analysis between loans and broadly, as all-in-cost pricing affects the demand guarantees shows that guarantees tended to be and utilization of facilities and also has develop- less profitable than loans over the 1990­2007 pe- 5 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 mental benefits in terms of introducing sound constraints and slow deployment of the associ- practices of risk pricing. ated technical assistance program. The energy ef- ficiency programs in Central and Eastern Europe It is often the case with structured guarantees that and in China have shown good utilization, and a transaction will not happen without the guar- these facilities have seen no claims so far.25 antee. In such instances, looking at the impact of the guarantee on all-in-cost may not be appro- An important new area of cooperation between priate. When comparisons are possible, the price IFC and the Bank is the MSME program in Africa, of the structured guarantee tends to have little im- which combines IFC's PCGs and IDA funding. pact on the all-in-cost to the client, as the IFC guar- Two joint projects have been committed in Mada- antee typically covers a small portion (often less gascar in the finance sector for a total of $12.5 mil- than 10 percent) of the securitization amount. lion; of this, $10 million is IFC's own account and $2.5 million is from the government of Madagas- At a more basic level, plausible assumptions about car, as partial risk coverage for up to $25 million the rationality of guarantee clients and the vol- of new local currency MSME loans. As of Sep- untary nature of the transactions tend to ensure tember 2007, the two participating banks had that guarantees--when they happen--are supe- disbursed about two-thirds of the $25 million rior to available alternatives. Nevertheless, atten- total facility. Currently there are five similar proj- tion on the impact of guarantees on all-in-cost of ects under preparation--two in Mali, two Sene- financing may be justified when the sophistication gal, and one in Ghana. and bargaining position of market participants is vastly asymmetric, as the case may be with some A number of other RSFs have not performed microfinance and SME clients. well. A student loan facility in Asia was cancelled because of low utilization and a delinquency level Providing single-credit guarantees to small in- above 10 percent, which has triggered the ramp- vestments in Africa has not been successful. up termination threshold. A risk-sharing facility in Over the 1990­2007 period, IFC committed 142 Asia targeting middle-size enterprises was not uti- guarantees (excluding GTPF). Of these, 21 have lized because of a change in strategy by partici- been called. Three-quarters of all called guaran- pating banks, which decided to focus on consumer tees have been for small projects in Africa. Of 19 finance and retail banking. Small Enterprise Fund and African Enterprise Fund guarantee projects, 15 have been called. A In the case of a risk-sharing facility in Europe, total of $8.7 million has been paid out of the $1.8 some utilization took place, but difficult approval, billion committed during the period. The African reporting, and managing processes--which led to Enterprise Fund has been discontinued, and di- IFC being involved in each individual subproject-- rect lending to SMEs has been largely replaced by have limited deployment. An RSF in east Asia was a wholesaling approach to SME lending. not utilized because pricing was thin and the first loss provision was very high, at 20 percent. The The RSFs have shown mixed success. The self- joint IDA­IFC operation was also not utilized. standing trade enhancement RSFs have shown Common reasons for lack of utilization include disappointing results and have been replaced by changes in market conditions; thin margins, which a new approach--GTFP--which has shown a re- make it more difficult for banks to give up a por- markable growth.24The school facilities in Africa tion of the net interest margin; unappealing risk have been moderately successful. The first school sharing agreements for partnering banks; and facility in Africa became fully utilized in less than difficult approval processes, whereby IFC was ap- two years, and a second facility has been ap- praising every transaction. proved. The approach was extended to another African country, where ramp-up has been slower Models fordeploying guarantees have been de- than expected because of the sponsor's liquidity veloped, but limited scaling up has taken place. 5 6 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S The first school facility in Africa has been replaced rency assets generally managed by local pension by another in Africa. The energy efficiency program funds that must adhere to strict local investment started in Hungary and was then replicated in guidelines and by IFC's increasing involvement in Central and Eastern Europe and in China. There local and international capital market activities have been bond enhancements for education in- on behalf of its clients. stitutions, municipalities, leasing companies, and banks, but always on a case-by-case basis. The External experts have also identified the poten- GTFP has been the only success in scaling up. tial for greater use of guarantee instruments by Progress has been made in standardizing and multilateral financing institutions, including IFC simplifying structuring, but there has been limited (World Economic Forum 2006). A comparison progress in scaling up and replicating. IFC with other international financing institutions in- continues to follow a largely opportunistic ap- dicates similar or lower levels and similar trends proach to the deployment of the instrument. in deployment of the instrument as in IFC, sug- gesting that there are common factors at work. At Potential for use the same time, use of guarantee and insurance in- Severalfactors, bothexternalandinternalto IFC, struments by private providers has shown growth, suggest greater potential for the deployment of although at different rates, depending on the IFC's guarantee products than current levels type of instrument. suggest. Trends in the external environment sug- gest significant potential demand in areas where IFC's strategic priorities and focus on addi- the instrument is being deployed today. New de- tionalityalso indicate greater potential. IFC has velopment challenges such as climate change are identified the imperative to work with and through emerging that require new approaches, including others (IFC 2008) for larger impact, including developing markets for carbon trading.26 through programmatic approaches and whole- saling. Guarantee instruments are particularly Although there is abundant liquidity in emerging well suited for the application of these approaches. markets, the capacity to channel this liquidity ef- They also have special properties from the per- ficiently into productive investments needs to be spective of additionality, because they are more developed. This involves a focus on deepening the likely to crowd in rather than crowd out private local capital markets, including local bond markets. flows than IFC's funded products. Other things IFC sees a growing demand for local currency fi- being equal, IFC's focus on additionality could nancing. MSMEs are increasingly viewed as the translate into more emphasis on the use of guar- main engine for sustainable and equitable growth. antee instruments. IFC's own experience with The need to develop market solutions to envi- trade facilities and the GTFP suggests that al- ronmental problems, to deepen financial mar- though significant demand may exist, constraints kets in developing countries in the context of on the supply side could limit IFC's ability to re- abundant liquidity, the growing demand for local spond to such demand (see box 2.1). currency financing, and the need to expand access to financing to underserved segments of the econ- IFC is exploring different ways to respond to omy--all these trends create a large and growing these opportunities, including through guar- potential demand for the use of guarantee-type antee instruments. The organization is placing a instruments. major focus on developing capacity in local cur- rency financing. It has issued more IFC bonds in Potential opportunities for IFC to add value and local currencies, supported development of de- to facilitate funding of private sector develop- rivative markets, and experimented with pro- ment through tailored guarantees are likely to grams such as MATCH. It has also experimented continue to increase, given the expanding range with different ways to work with MSMEs and to of opportunities for private investment in devel- facilitate trade and short-term finance. oping countries, as well as by the increasing cur- 5 7 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Box 2.1: GTFP: A New Product in Response to Existing Demand Following the poor performance of its trade finance facilities, one-third are South-South transactions. PRM is a very important in 2003 IFC piloted a new approach based on a successful Eu- factor for the GTFP. Most of the transactions are in difficult envi- ropean Bank for Reconstruction and Development model. From ronments, where the confirming banks have exposure constraints. fiscal 1998 to 2003, IFC committed 21 trade finance facilities for a total of $542 million. Of the 21 facilities, 11 were never used, and The GTFP approach is fundamentally different from the traditional of the 10 that were used, the average utilization rate was just 27 IFC approach to trade facilities. The traditional approach is bi- percent. lateral, whereas GTFP is an open, multilateral network architec- ture. The nature of the instrument is also different: in the past IFC Although there was clearly a demand for short-term trade fi- guaranteed 25­50 percent of the confirming bank exposures and nance guarantees, IFC was not able to respond to this demand used a number of covenants on the participating banks on top of with its products. In 2003 an IFC review of this experience found serious reporting and other requirements, but the instrument that among the main reasons for poor utilization were delays in used in GTFP is a 100 percent unconditional demand guarantee. negotiating the necessary framework agreements with each For country and client exposure limits, every dollar that IFC com- participating bank, caused by imposition of stringent financial re- mits under the GTFP is counted at 50 percent for headroom pur- porting requirements that were not standard market practice for poses. GTFP had to develop a customized booking system, trade-related transactions; the conclusion by participating banks because IFC's standard systems for disbursements could not that IFC's guarantee was not a firm guarantee and would not help handle high-frequency short-term trade finance transactions. reduce capital requirements, a conclusion caused by various representations they had to make according to IFC's covenants; Other success factors include limited competition with other parts and the imposition of high capital charges that were not in line of IFC's business; commitment at the top; a bold initial approach, with the lower-risk profile of trade transactions, which reduced with $500 million approved for the pilot program; delegated authority, the profitability of the facilities. The review also observed the Eu- special systems, and dedicated staff that could rely on support from ropean Bank for Reconstruction and Development's (EBRD) suc- a large department; moving close to established market prac- cessful approach to trade finance and a decision was made to tices; and limited competition in the Sub-Saharan Africa Region. implement a similar approach in IFC. EBRD's experience gave IFC the confidence to start out in a decisive manner. GTFP has promising new applications and growth potential. IFC is continuing to develop the program. Efforts are under way The program has seen rapid growth, particularly in Africa. The to further simplify documentation and bring structures even program started in 2005 and has grown rapidly. In fiscal 2007, total closer to market practices. Opportunities include wholesaling to commitments had reached $767 million with the issue of 564 guar- address single exposure limits to preapproved clients and using antees, of which $377 million was in Sub-Saharan Africa. The pro- the platform to expand the range of products to include short-term gram doubled its coverage in fiscal 2007 to include 96 banks financing, swaps, carbon delivery guarantees, and other prod- across 51 countries. More than two-thirds of all GTFP transactions ucts that involve bank-to-bank interactions. Unlike the old ap- have been in frontier markets. In a number of countries GTFP has proach, the GTFP has the potential to enable IFC to respond been the first IFC deal--or simply the first deal--in years. About quickly to tightening market liquidity. Source: IFC. The imperative to experiment in this and other scale has not occurred, except in the case of the areas remains. The outcome of these experiments GTFP. What have the constraints been? will affect the demand for guarantee instruments.27 In a number of areas, experimentation and learn- Constraints to use ing have converged on the development and im- Guarantees have been a fringe instrument for provement of guarantee-type instruments. Models IFC. They have been used in situations where have been developed, but replication on a large deployment of IFC's direct instruments has been 5 8 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S impractical. In this sense, guarantees have been crease transaction costs relative to alternatives, an instrument of necessity rather than choice. when and if available. They have been used in areas where IFC's ca- pacity to invest directly is limited: local currency, Conditions are more favorable for traditional MSMEs, and short-term finance. partialcreditguaranteesin the case ofcrediten- hancements to local financial institutions that This pattern of use determines both the chal- operate in high net interest rate margin envi- lenges and the opportunities of the instrument in ronmentsand to issuersofcorporate bonds. Mi- supporting IFC's strategic priorities. The chal- crofinance, consumer finance, and leasing lenges relate to IFC's tendency to apply its tradi- companies in developing countries often operate tional project financier's approach to guarantees, in environments that allow lending at high net in- given the fringe nature of the instrument. The op- terest rate margins. Interest rate margins of 20, 30, portunities reside in the areas of use for this prod- and even above 50 percent are not unusual in uct, which are areas where IFC wants to develop micro and consumer finance in countries such as capacities, grow, and expand its presence. Brazil, Mexico, and Indonesia. In environments of interest rate margins of such magnitudes, clients PCGs for single credits in local currency face may not be too sensitive to what is a relatively small some inherent limitations. For investment staff, increase in their all-in-cost of funding. traditional single-credit PCGs are easier, because they are very similar to straight debt financing in Opportunities for traditional PCGs also appear to terms of process and documentation. However, be significant for enhancement of local corpo- they face some inherent limitations. To make rate bond issuances. Pricing for corporate bonds them economical for IFC, PCGs for single credits tends to be more rational than for loans, and this in local currency have to be for relatively large in- allows the IFC credit rating to be more accurately vestments. Large investments are normally with reflected in the market. Also, for corporate bonds, relatively large clients that have existing relation- enhancement often makes the difference be- ships with local banks. In such circumstances, tween a success and a failure. However, bond en- local banks often do not see a role for IFC that hancement is highly sensitive to market conditions would justify lowering their spread enough to and requires the capacity to react quickly. Hence, fully reflect IFC's AAA rating. a traditional project finance approach is not always the most appropriate. Also, in practice IFC's AAA rating has been better recognized and priced internationally than lo- IFC'sapproachto RSFshasconstrainedtheirde- cally, where it is often hard to do better than the ploymentandutilization.It is hard to make a case sovereign. As a result, it is typically more expen- against excessive prudence, particularly in the sive for the client to use the guarantee instrument midst of an unfolding crisis. Still, IFC's overly con- than a direct loan from IFC or from a local finan- servative stance toward the structuring of RSFs has cial institution. IFC guarantee transactions have constrained their use and that there is room for higher drop rates than IFC loans. According to re- a more flexible approach consistent with the prin- sponses to the staff survey conducted for this ciples of efficient risk sharing. study, in 81 percent of cases, a high guarantee fee was the main reason for the droppage (see ap- IFC has taken an inflexible approach against pendix B). Clients are often willing to pay a pre- sharing in first-loss positions. The approach has mium for IFC's name. However, IFC has a more been to look for third-party--often donor-- remote relationship with the client in a guaran- money to fund a first-loss cushion. In the case of tee than in a direct investment. As a result, these the energy-efficiency programs, the source of benefits are less tangible for the client. Moreover, such money has been the Global Environment Fa- adding a third party--the guarantor-- tends to in- cility. In the case of some of the SME facilities in Africa, IDA has provided resources. A foundation 5 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 has funded a first-loss position in a student loan there are no adequate hedging instruments for program. currency and interest rate risk. Although this limit is not a binding constraint at the moment, it in- Mobilizing funding for first loss from a third party creases transaction costs and reflects a formalis- has been a difficult and time-consuming process. tic rather than a pragmatic approach to risk. At times, it has taken IFC two to three years to mo- bilize trust funds for this purpose. That has added An argument can be made that in a situation of to transaction costs. In addition, IFC often de- claim and subrogation, the main risk is a credit mands high first-loss levels, ranging from 5 to 30 event, not a currency event. If a currency mismatch percent in some cases. Strict eligibility criteria is present in a project, the currency risk in the for booking assets under the facilities have also transaction cannot be eliminated by subrogation constrained use at times. in a hard currency. And if there is additional risk by allowing subrogation in the local currency, the Combined with the use of circuit breakers when risk is not likely to be too high. Some multilateral losses approach first-loss limits, the above fea- financing institutions have recently allowed for tures have tended to make risk-sharing facili- greater flexibility as far as subrogation in the local ties a misnomer. IFC started doing RSFs by sharing currency is concerned and have reportedly seen risks pari passu with the beneficiary and without an increase in demand for guarantee products. first-loss provisions. A more conservative ap- proach has been introduced over time. IFC has ac- The traditional project financier's approach to cumulated the data and the experience to give it guarantee-type instruments has constrained the confidence to take bigger risks, simplify the use of the product. IFC tends to apply a uni- processes, and give the flexibility to partners to form approach to all projects, irrespective of the use their strengths. For example, in Eastern Eu- levels and types of risks. The same comprehen- rope, hundreds of small projects have been ap- sive approach to risk is applied across the board, proved in energy efficiency, and none of them has instead of using a segmented approach focusing gone bad. In the IFC-IDA SME facilities, although on the key risks that matter the most in the par- some claims have been paid, numbers and ticular circumstance, based on the lessons of ex- amounts have been so small that the first-loss re- perience and project analysis. serves have not been eroded--on the contrary, they have increased as a result of accumulated In GTFP agreements, for instance, although some interest. accommodations have been made, IFC still has negative covenants, ratios, and other restrictions In all these cases, origination has been decoupled that limit the ability of clients to use the guaran- from first loss via the third-party funding, yet the tees for capital, exposure, and provisioning relief experience has been positive. IFC needs to revisit purposes. That limits the attractiveness of the its approach to structuring RSFs. product. Charging commitment fees over the un- used portion of RSFs has irritated partner banks. Subrogation in foreign currency is still a con- It is not clear whether the same rationale exists straint. The 1997 revision of IFC's guarantee pol- for applying a commitment fee in RSFs, as in the icy introduced greater flexibility in the deployment case of funded commitment. of the instrument by allowing subrogation in the local currency in the event of a call on a local IFC's processes and systems are not well suited currency guarantee. However, IFC introduced to handle some types ofguarantee transactions. limits to the aggregate notional volume of such IFC carries the legacy of a project finance insti- guarantees. There are currently various limits on tution. This is evident in a project finance culture the notional committed value of all guarantees of that relies on heavy documentation and a back of- financial instruments where IFC's post-call claim fice system that cannot handle high-frequency is denominated in a local currency for which transactions. For example, the RSFs have problems 6 0 R E V I E W O F T H E W B G ' S E X P E R I E N C E W I T H G U A R A N T E E S handling claims. Each legitimate claim is treated The incentivesstructure tendsto discourage the as a loan disbursement and is handled accordingly. use of the instrument. Incentives at IFC favor This creates delays and is a further discouragement booking large transactions. Deals involving guar- to the use of the instrument. GTFP, for instance, antees for SMEs, micro, and leasing companies had to establish its own back office to handle tend to be small. Incentives do not favor working high-frequency transactions. on complicated small projects using products with which staff are not familiar, such as those in- IFC needs to differentiate processes according volving guarantees. to types of instruments, following--to the extent possible--practices the market is used to. IFC's With the emphasis on development impact, atti- average processing time has been declining in tudes are beginning to change. There are other recent years but is still close to 10 months. Guar- aspects of the incentives system that tend, al- antee operations are not significantly different though inadvertently, to discourage guarantees. than the IFC's average in processing time. Some For instance, Treasury allows a 30 basis point re- guarantee transactions have taken a long time to duction in pricing for local currency financing develop, given high complexity and the partici- through the derivative market, but not through pation of more players. There is a clear learning guarantees. The incentive is based on the his- element, however, as replications have tended toric difference between the London Interbank to be significantly less expensive. This pattern is offered rate and IFC's real funding costs, but it clearly visible, for example, in the case of IFC's creates a bias against guarantees and in favor of energy-efficiency facilities. According to survey direct local currency funding by IFC where the op- results, high processing costs have been the main portunity exists. An argument can be made to reason that about 43 percent of clients drop IFC's extend the same incentives to guarantees for local guarantee transactions (see appendix B). currency loans.28 A rigid approach to structuring RSFs limits IFC has not institutionalized innovation. Inno- IFC's abilityto exploit fullythe benefits of part- vation in IFC is a slow, highly decentralized process nering with local institutions. By their nature, of product mutation. IFC has a new product group guarantees involve working with and through that deals with risk aspects but does not have others, often local financial institutions. These in- groups focusing on the development of new prod- stitutions have the advantages of local knowl- ucts. IFC's Treasury has been instrumental in in- edge and information, and on this basis they can novating and spreading innovation. However, supervise certain clients more effectively than mainstreaming can be led by the industry and IFC. regional departments. Innovation does not nec- essarily imply coming up with more complex Once incentives are aligned and IFC has gained products. It often involves simplification and en- comfort with the abilities of the local financial in- suring consistency with established market prac- stitution, IFC could allow a degree of flexibility to tices. A more systematic approach to product the local institution in managing the utilization. innovation would also greatly facilitate broader Instead, even with existing and familiar partners, deployment. it tends to impose onerous reporting and eligibility requirements and at times wants to appraise every IFC would also need to move away from small in- project. Working with and through local institu- novations that tend to get lost and do not lead to tions would necessitate greater openness and cumulative improvements, to a focus on innova- willingness to allow the use of legal practices and tive efforts for a few key themes that can lead to precedents that have been developed and ap- quantum jumps in business. This is consistent plied locally. This rigidity and the high transaction with the efforts to move toward a programmatic costs have discouraged use of the instrument. approach and away from one-off transactions. 6 1 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Guarantee instruments are well suited to be part It has not adapted to more current market prac- of this approach. tices with its increasingly cumbersome internal processes. Complexity and lack of familiarity with some products tend to discourage scaling up. Some Internal constraints to the deployment of Bank guarantee structures can be replicated on a larger PRGs include the application of standards de- scale. But products that are more complex and signed for public sector operations to private sec- require a high degree of familiarity to become an tor projects, more onerous internal processing effective business development tool are slowing requirements than are involved in the deploy- deployment down. People on the front line need ment of traditional Bank instruments, and inter- to understand the product so they can match it nal incentives that favor the use of Bank lending with the situation. Decentralization brings risks over Bank guarantees. In addition, the inflexible in representing IFC's products. Treasury is de- use of counter-guarantees for IDA PRGs, although centralizing, putting key staff in the field. useful in some circumstances, has tended at times to diminish the attractiveness of the product. Conclusion A range of factors contribute to the limited use of IFC has tended to apply a traditional lender's the instrument, but factors under the control of approach to guarantee-type instruments. It has the three WBG institutions play a significant role. taken a conservative stance with RSFs, which has constrained their deployment and utilization. MIGA's Convention and Operational Regulations Though IFC has made significant progress in present many restrictions and hinder MIGA's innovation and in standardizing structures, lim- adaptability to new market trends. However, MIGA ited replication and scaling up that has taken has not been sufficiently aggressive in innovating place to date have been a result of remaining within the flexibility allowed by current policies. complexities, poor familiarity with the products, and the lack of a systematic approach to inno- vation. In general, guarantees have tended to be used as last resort instruments by IFC and the Bank. 6 2 Chapter 3 The main street in the university city of Irbid, Jordan, has more Internet cafes per mile than any other street in the world. The World Bank guaranteed one of the first private investment telecom projects in Jordan. Photo © Frédérique Harmsze. The Delivery of PRM Products across the WBG C hapter 1 looked at the design of WBG guarantee instruments and chap- ter 2 at their deployment in practice. This chapter examines the in- teractions among the three WBG institutions in the delivery of guarantees. The focus is on PRM products because this is the area where over- laps exist in the product offerings of MIGA, IFC, and the Bank. Concerns have been raised that the organ- The space of PRM products is different from the ization of the provision of PRM products space of guarantee products: it includes the part within the WBG is not optimal and that of guarantees that covers political risk as well as overlap between the three institutions may nonguarantee products that provide PRM. In this confuse clients and reduce efficiency. There chapter we look at the PRM products of the WBG, have been several initiatives to address this issue. the organization of their delivery, the issues of In fiscal 1997 the Board reviewed the WBG's guar- overlaps and competition, and the coordination antee activities and examined proposals to im- and cooperation mechanisms in place. prove the operational synergy in the provision of risk-mitigating products (World Bank 1997b). A The Market for PRM policy document established the principles for de- The objective of WBG PRM products is to cat- ployment of a Bank PRG over MIGA PRI (World alyze investment that is not flowing because Bank 2000). In fiscal 2005 there was an attempt political risk is perceived to be too high. The to develop a more coordinated approach be- World Bank estimates average FDI flows and ex- tween the Bank's Project Finance and Guarantees ternal private loans to developing countries at Group and MIGA to increase the use of existing $210 billion a year in the period 1990­2005 (World guarantee instruments, especially in the infra- Bank 2007a). The total project cost supported structure sector. More recently, a task force was by all WBG guarantee instruments was $5.3 billion established to explore options for optimizing the annually. Assuming total project cost to be a proxy delivery of guarantees within the WBG. for investment flows, WBG guarantee instruments thus supported the flow of approximately 2.5 All the WBG's policy documents on risk-mitigation percent of total investment flows. products have consistently emphasized the need for close interaction among the three institutions Similarly, of the new investment flows covered by to ensure complementarity and to minimize du- guarantees, a recent internal study estimated MIGA's plication of services. The issues of overlap and share to be between 2 percent and 4 percent of the competition among the three institutions exist total market. From the WBG's perspective, how- only with respect to PRM, as this is the only area ever, although the shares are not large, the objec- covered by the products of all three institutions. tive is not to increase its PRM coverage of existing 6 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 investment flows but instead to catalyze additional example, derives minimal business from U.S. or investments that are not taking place because of Japanese investors, in large part because of the ex- high perceptions of political risk. istence of these national guarantee agencies. Sev- eral private sector providers, mostly members of This potential market is largely unquantifiable. It the Berne Union, also provide PRI products, and ranges from small family businesses consider- many opt for self-insurance. The circumstances will ing opening plants in high-risk developing coun- also dictate the extent of coverage sought, rang- tries, to large conglomerates scouring the world ing from a single to multiple risks or from ex- for investment opportunities, to commercial plicit insurance to an implicit understanding that banks seeking to manage their exposure risks the third-party agency will act to protect the in- around the world. terests of the investor. The varied market for PRM can be met by a WBG Political Risk-Mitigation Products broad range of formal and informal prod- As conceived, the WBG's guarantee products ucts. The perception of political risk remains were designed to complement each other. high among potential investors in developing Several WBG guarantee products offer PRM in- countries (EIU 2007). A range of options exists to cluding MIGA PRI, the Bank's PRG, and IFC's mitigate this political risk, with demand for both PCG, which offers comprehensive political and the type and extent of coverage varying accord- commercial risk coverage. ing to the particular circumstances of the investor and the potential project. The circumstances As originally designed, these products were in- might include the degree of investor familiarity tended to complement--not compete with-- with the country and/or sector, the overall per- one another. IFC's first guarantee policy, adopted ception of political risk in the country, the degree in 1988, the same year that MIGA was established, of sector- or project-specific riskiness, the in- had a whole section on the division of labor be- vestor's exposure relative to prudential norms tween IFC and MIGA. It emphasized that IFC's and and limits on a country and/or sector, and com- MIGA's guarantee operations were different and mercial bank country exposures vis-à-vis pru- that "both programs are potentially more com- dential limits and regulatory requirements. plementary than competitive." The particular circumstances can also dictate the At the same time, however, the policy recognized type of coverage opted for by the investor. Ac- that the coverages offered by IFC and MIGA might cording to the internal market study, self-insurance overlap. IFC indicates that, unlike MIGA, IFC as a (or no formal third-party guarantee) is chosen matter of policy does not offer coverage for only some 65 percent of the time. In other cases, loans one or a few risk elements--for example, transfer from large commercial banks that have estab- risk alone--because doing so may potentially jeop- lished relationships with host governments, or ardize its de facto preferred creditor status as the implied PRM provided by multilateral agen- well as impinge on MIGA's role. Bank policy doc- cies lending to private firms (such as IFC loans, uments have also emphasized the differences be- as discussed below), can provide a sufficient mea- tween the PRG and MIGA and IFC products. Unlike sure of comfort. Bilateral investment treaties be- PRGs, MIGA's PRI does not require a sovereign tween countries that use arbitration to provide counter-guarantee and MIGA could also guarantee reciprocal protection for investors against politi- equity, which the PRGs could not (see table 3.1). cal risks might also provide some comfort. Flexibility of policies has blurred original Some investors have the option of large national product boundaries. All the WBG's guarantee guarantee agencies for overseas investments, policies contain some degree of flexibility in in- such as OPIC in the United States or Nippon Ex- terpretation. IFC and Bank guarantee policies, for port and Investment Insurance in Japan. MIGA, for example, allow a significant degree of freedom in 6 6 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G Table 3.1: The WBG's Political Risk-Mitigation Products Guarantee products that provide PRM Guarantee products excluded as not providing for long-term private investment products PRM for long-term private investment projects IDA/IBRD PRGs IBRD PCGs--provide support for public investment projects MIGA PRI IFC PCGs IFC GTFP--provides support for trade finance, largely short term Nonguarantee products that provide PRM for long-term private investment products Approximately 10 percent of normal IFC investment operations IFC B-Loans IFC CELT IFC GOLF Source: IEG. Note: CELT = credit-enhanced lending transaction; GOLF = Global Offshore Liquidity Facility; GTFP = Global Trade Finance Program; IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency; PCG = Partial Credit Guarantee; PRG = Partial Risk Guarantee; PRI = political risk insurance; PRM = political risk mitigation; WBG = World Bank Group. terms of types of risks covered and extent of cov- Under its regular guarantee program, MIGA has erage. Although MIGA's policies are the most re- also been developing this line of business and has strictive in terms of eligibility requirements and risk provided coverage against transfer restrictions coverage, they also allow significant room for and convertibility risks to residential mortgage- learning and experimentation. For example, MIGA backed securitizations in Latvia and Kazakhstan guarantees were expected to focus on foreign eq- and to accounts receivables securitization in Brazil. uity holders, which allows MIGA to insure non- IFC has also introduced credit-linked guarantees, shareholder loans if it also insures that an equity a flexible partial risk-guarantee product that can holder in the same project has led to guarantees cover various bundles of risks.1 Although both for nonshareholder loans (such arrangements ac- of these products have seen limited use to date, count for about a third of MIGA's guarantees). As they represent a departure from the traditional ap- another example, Bank PRGs are only mandated proach of offering full-risk guarantees. to insure loans. In one case, however (the West African gas pipeline), innovative structuring of the In another example, in the mid-1990s, the Bank guarantee effectively enabled the PRG to guaran- introduced the guarantee facility structure, initially tee the underlying equity investment. for short-term trade finance in Eastern Europe but subsequently for wholesale guarantees to smaller Innovation in the guarantee product space long-term infrastructure investments (see table 3.2 has increased the range of WBG political for comparisons of instrument deployment). risk-mitigation products. IFC has introduced several new partial risk-guarantee products that IFC's innovations in the nonguarantee prod- carry some political risk coverage. The GOLF ef- uct arena have introduced products that fectively provides coverage against transfer and embed coverage for political risk. An example convertibility risk (see page 77 for a more de- is IFC's Credit-Enhanced Lending Transaction tailed description). In its first and only application (CELT). IFC introduced CELT in the late 1990s, so far, GOLF encourages mortgage-backed capi- adapting a similar product used by the European tal market transactions by guaranteeing an off- Investment Bank and EBRD, to help address chal- shore liquidity facility that can be drawn on if a lenges arising from internal and regulatory coun- restriction on currency transfer or convertibility try exposure limits that large international banks is imposed. face in moving into emerging markets. 6 7 68 THE Table 3.2: Comparisons of the Deployment of WBG Risk-Mitigation Instruments WORLD Regional Low- High deployment income country risk Sector Project BANK Total value to date country deployment deployment size Risk of underlying No. of Committed (top three) deployment to date to date deploy- coverage projects projects amount (%) to date (%) (%) (%) ment (5)b GROUP IFC PCG Comprehensive $8.6 billion 142 $1.8 billion AFR: 35 Low: 55 High: 46 Finance: 51 Small: 31 EAP: 18 Med/Low: 54 MST: 14 Medium: 39 Infrastructure: 13 Large: 23 GUARANTEE IBRD/IDA PRG Transfer/convertibility, $9.4 billion 13 $1.2 billion AFR: 31 Low: 75 High: 70 Infrastructure: 92 Small: 0 expropriation, war, breach SAR: 23 Med/Low: 30 MST: 8 Medium: 6 of contract, some Large: 94 commercial risk, some INSTRUMENTS natural events IFC loans or equity Transfer/convertibility, $28.8 billion 404 $5 billion ECA: 29 Low: 34.5 High: 35 Finance: 31 Small: 22 investments with PRI expropriation, war, breach AFR: 26 Med/Low: 65 Infrastructure: 27 Medium: 44 as main reasona of contract Large: 31 1990­2007 MIGA PRI Transfer/convertibility, war, $78 billion 566 $16.6 billion LAC: 31 Low: 32.5 High: 45 Finance: 33 Small: 23 breach of contract ECA: 31 Med/Low: 48 MST: 31 Medium: 49 Infrastructure: 24 Large: 28 IFC B-loans Transfer/convertibility only $112 billion 629 $14.5 billion LAC: 41 Low: 22.6 High: 31 MST: 38 Small: 3 ECA: 20 Med/Low: 69 Infrastructure: 23 Medium: 51 Finance: 17 Large: 43 IFC GOLF Transfer/convertibility only $100 million 1 $19 million ECA: 100 Low: 0 High: 0 Finance: 100 Small: 0 Med/Low: 100 Medium: 0 Large: 100 IFC CELT Transfer/convertibility, war, $0.51 million 4 $0.49 million ECA: 100 Low: 0 High: 0 Finance: 100 Small: 0 breach of contract Med/Low: 100 Medium: 50 Large: 50 Source: IEG, based on World Bank, MIGA, and IFC data. Note: All percentages are by number of projects, not volume. CELT = Credit-Enhanced Lending Transaction; GOLF = Global Offshore Liquidity Facility; IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency; MST = multisector task; PCG = Partial Credit Guarantee; PRG = Partial Risk Guarantee; PRI = political risk insurance; WBG = World Bank Group. Regions: AFR = Sub-Saharan Africa; EAP = East Asia and Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; SAR = South Asia. a. Estimated for fiscal 1990­2007 based on the actual data for fiscal 2005­07. b. Small = less than $10 million; medium = $10­100 million; large = more than $100 million. T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G With CELT, a parent international bank will guar- Institutional Organization antee repayment of an IFC loan to its subsidiary of the Delivery of PRM Products under all circumstances except if nonpayment is The WBG is the only multilateral to have caused by political events. In addition to the credit three distinct institutions providing PRI, PRI risk of the parent, IFC thus bears the political with counter-guarantees, and comprehen- risk on these projects, including the traditional po- sive credit guarantees. The ADB, AfDB, and the litical risks covered by MIGA and the Bank's PRGs. Inter-American Development Bank, for example, Four CELT transactions have been completed to offer PRI (with or without sovereign counter guar- date for a total of about $500 million, all in the Eu- antees) and comprehensive credit guarantees to rope and Central Asia Region. private sector clients. In each institution, guaran- tees are provided by private sector departments Traditional IFC products also embed im- within the main institution that can offer any of the plicit political risk cover. IFC's normal lend- three guarantee products. The same institutions, ing and equity investments also carry a degree of moreover, are also able to offer direct investments implicit political risk cover (IPRC). As a multilateral to the private sector, thereby offering a full range organization and member of the WBG, IFC faces of financing and risk-mitigation products. lower country risks from political and economic conditions than a private investor. Through its A recent independent panel report in the AfDB participation in a project, IFC can transfer that risk considered separating the private sector de- reduction to other investors, "furnishing comfort velopment unit from the main organization but without issuing a formal guarantee" (Haralz 2007). recommended that the AfDB retain a unified structure to be better able to mobilize all its re- According to IFC's client surveys, perceived risk sources toward a common private sector devel- reduction or political risk cover is among the opment objective. The report also recommended main reasons that clients use IFC as a source of creating a single point of entry for all private sec- finance. An IEG review of IFC projects between tor transactions. fiscal 2005 and 2007 conducted for this evaluation found that in approximately 10 percent of in- However, multilaterals with unified struc- vestment projects, IPRC was indicated as a primary tures have not seen a larger share of guar- reason for IFC's involvement in the project and antees in their businesses. Each of the major its main additionality. multilateral institutions has also seen limited growth in the provision of guarantees for medium- to long- IFC's B-loan program is another instrument that term investment. ADB has issued just 27 guaran- provides participating banks with implicit cover tees (both PCGs and PRGs) in its history, and in against transfer and convertibility risk. The 1997 2006 guarantee commitments accounted for just internal review of WBG guarantees (World Bank 2 percent of its loan commitments. EBRD has is- 1997b) identified the B-loan as IFC's main polit- sued three guarantees in infrastructure and more ical risk-mitigation product. Participation in IFC's than 60 in the financial sector, for a total volume B-loan program provides an implied rather than of 3 percent of its total investment operations. explicit cover and can mitigate transfer and con- Guarantees account for 3.6 percent of the European vertibility risk through IFC's preferred creditor Investment Bank's investment portfolio. In AfDB, status. guarantee commitments accounted for just 1.5 percent of its commitments in 2006. IFC's guarantee These traditional products have long been part of operations consist of 6.6 percent of its loan com- IFC's arsenal. What has changed in recent years mitments, and IBRD's guarantee operations rep- is that IFC's business development capacity has resent 1.6 percent of its total lending operations. strengthened, and that has been accompanied by a more aggressive marketing of the implicit po- A review of the organization of the provision of litical risk cover as a substitute for PRI. guarantees in several of these multilaterals found 6 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 that a range of internal policy and structural con- cisions to the Board for concurrence. A Project Re- straints have inhibited a more widespread use of view Committee consisting of MIGA senior man- guarantees. Factors undermining greater use were agement meets on new guarantee proposals early similar to those found in the WBG in this evalu- in the underwriting process to provide guidance ation and included the complex nature of guar- to underwriting teams. antee products, an internal lack of information and knowledge, the de facto preference for loans, In the Bank, although many lending decisions and the lack of a dedicated guarantee unit with have been delegated to Regions, a complex ap- sufficient authority to coordinate the provision of proval procedure is still followed for guarantees, guarantees. with all guarantee operations requiring senior management approval. Projects are initiated by There are significant differences in the or- Regional departments and typically involve ob- ganizational structures of the three WBG taining support from the central guarantees unit, institutions. The literature on organizational obtaining senior management approval of an in- design distinguishes among five broad organiza- dicative PRG term sheet, obtaining inputs from Re- tional structures: functional, product, market, ge- gional staff through the Regional Operations ographical, and process oriented. Committee, and finally obtaining inputs from Bank-wide staff and senior management approval MIGA is de facto a single product institution and through the Operations Committee before being its organizational structure is of the functional presented to the Board. type.2 MIGA remains a Washington-based orga- nization. Buyers of MIGA PRI tend to come from In IFC, the system involves a high degree of co- headquarters of the sponsors, usually located in ordination between the industry departments in major business centers in the United States, Eu- headquarters, industry department staff based in rope, or Japan, rather than from regional offices. the field, and Regional departments in the field. Senior management acts as arbiter in the case IFC combines geographical and market (industry) of disagreement but is still heavily involved in structures with an increasing focus on geo- project-by-project decisions. Increasingly, the ten- graphical structures integrating product and dency is for IFC decision-making authority to mi- process components. Until recently, it did not grate to the field under delegated decision-making have product-based organizational units. Recently, authority. however, it established an equity department, and with the reorganization of the financial mar- Distinct repositories of expertise have ket departments, it has introduced departments evolved. In IFC, the Treasury is the custodian of based on product or product lines such as the product knowledge and expertise regarding guar- short-term finance department. IFC's advisory antees, but industry and regional departments services have been recently organized along busi- identify and develop prospects that might use ness lines. risk-mitigation instruments. In MIGA, product ex- pertise is separated into key functional units: op- The Bank is similar: the geographic structure is erations (underwriters); legal, economics, and becoming dominant, but market (industry) and policy (country risk analysis); and finance (pric- process elements continue to be important. ing). In the Bank, regional units now prepare Currently, the Bank and IFC are going through a guarantee projects with some support from a similar process of decentralization and are de- small central unit, if needed. The Bank's Treasury centralized to a similar extent. Unit retains the responsibility of maintaining knowledge and developing new products. Organizational structures determine the loci of decision making. In MIGA top man- Table 3.3 identifies some key areas of exper- agement in headquarters presents all project de- tise in delivering PRM in the WBG. MIGA has two 7 0 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G Table 3.3: Potential Synergies of Expertise across the WBG Area of advantage Institution Potential synergies Evaluating, rating, and pricing MIGA MIGA has two departments that evaluate, rate, and price PRI risk. IFC has benefited from PRI risk MIGA's specialization with evaluating and pricing this risk with respect to some of its projects. The potential exists to incorporate MIGA's evaluation and pricing of PRI risk within IFC's all-risk PCGs and IBRD/IDA's PRGs to allow a more uniform approach to pricing PRI, some unbundling of pricing components, and greater transparency in risk pricing. Structuring private sector IFC IFC staff have the most experience in helping clients structure private sector financing for financing projects. With limited capacity in IBRD/IDA, potential outsourcing of the financial structuring component of PRG projects to IFC to complement Bank sector expertise can enable the Bank to expand its capacity for PRGs without needing to build its own capacity in financial structuring of private sector projects. Claims management IFC/MIGA MIGA and IFC both have the most experience managing claims and near-claims situations. The legal expertise that MIGA and IFC have developed in their claims practices could be applied to IBRD/IDA clients and claim situations. MIGA also has the mandate and the experience to mediate disputes between investors and governments unrelated to its own guarantees that provide opportunities for learning. Regulatory framework building, IBRD/IDA IBRD/IDA has the most extensive experience providing guarantees for high-risk projects high-risk, and groundbreaking that break new ground, involve cross-country projects, and/or involve developing extensive projects regulatory frameworks or breach of contract. In the past, both IFC and MIGA have referred projects to the Bank for potential deployment of PRGs in such circumstances. Relationship with governments IBRD/IDA IBRD has the closest relationship with governments and the best potential leverage in the event of a dispute. Both IFC and MIGA, although retaining their own capacity to resolve disputes with governments, have relied on Bank support in the past. To fully enable this process, MIGA has developed mechanisms to ensure that its projects are consistent with the Bank's country strategies. Similar mechanisms might be introduced for IFC products carrying significant PRM to ensure that the Bank's interventions can be fully effective. Marketing IFC Both IFC and MIGA have dedicated marketing staff. IFC has the broadest network of private sector clients and international commercial banks that are key drivers of the demand for PRI. MIGA has a smaller staff but has developed a variety of marketing tools and established relationships with other insurance providers and brokers. Relative strengths of the marketing approach in each institution could be combined to provide stronger and more unified marketing of all WBG PRM products. Evaluating commercial risk IFC IFC has the strongest experience and practice in evaluating and pricing commercial risk. At present, no other WBG institution currently offers commercial risk coverage for private sector firms. However, with breach-of-contract coverage--offered by both Bank PRGs and MIGA PRI--increasingly blurring the distinctions between political and commercial risk, potential engagement of IFC's capacity in assessing commercial risks might be warranted. Source: IEG. Note: IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; MIGA = Multilateral Invest- ment Guarantee Agency; PRI = political risk insurance; PRG = Partial Risk Guarantee; PRM = political risk mitigation; WBG = World Bank Group. 7 1 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 departments that undertake evaluation, rating, need for MIGA to improve its relationship with the and pricing of political risk. IFC is the only insti- client. IEG­MIGA's 2008 Annual Report also noted tution that has the capacity to assess commercial the absence of a business development plan in risk. IFC staff have the most experience helping MIGA's fiscal 2005­08 strategic directions that clients structure private sector financing for proj- would include objectives, responsibilities, and ects; the Bank's capacity in this area remains resources, and would have allowed MIGA to mea- small. Along with its overall leverage, the Bank's sure at the end of each period its effectiveness in strengths in PRM are its sector knowledge, policy business development and origination (IEG­ dialogue, and close relationship with govern- MIGA 2008b). It also noted that assignment and ments that are embedded in its Regional and sec- responsibilities of individual staff for managing spe- tor departments. IFC has the most extensive cific clients and key accounts are unclear. marketing infrastructure for the same broad group of clients. The Bank and IFC have tended to be more engaged in the underlying project, although IFC has the most extensive business de- MIGA's engagement is increasing. Given the velopment infrastructure. IFC is the only possibility that poor project economics enhance WBG institution that has meaningful business pressures to renege on commitments, PRI pro- development capacity with private sector cli- viders have enhanced appraisal of projects. ents. Business development is done by field and Washington-based staff, often jointly. Some 600 Within the WBG, the Bank's PRG is probably the staff of 3,200 (as of February 2008) are actively most engaged PRI product, with Bank staff con- involved in promoting IFC's range of products, ducting extensive technical, social, environmen- including risk-mitigation instruments. Half of tal, and economic assessments of projects. The them are based in the field. Bank then stays engaged in the project through regular monitoring and supervision. The Bank, in contrast, has not undertaken con- sistent marketing and promotion of PRGs, with IFC is also concerned with the commercial viability only occasional updates of product brochures of its borrowers under straight and B-loan oper- and some promotional efforts undertaken. PRGs ations, as well as of clients of its PCGs. As a po- have tended to originate through governments litical risk insurer, MIGA has limited ability to rather than directly from project sponsors. By influence project design and outcomes because and large, the Bank has lacked consistent and its relationship with projects is more removed systematic contact with a range of potential pri- than it would be if it were a lending institution. vate sector clients. However, as part of the underwriting process and consistent with its business model, which was in- MIGA has a dedicated marketing unit that func- troduced in 2004, MIGA now has teams of un- tions mainly as a corporate relations department derwriters, economists, and environmental and but does not undertake systematic business social specialists to conduct due diligence field as- development of its guarantee operations. Oper- sessments for most nonfinancial sector projects. ational staff are primarily responsible for main- taining regular contact with potential clients, Although project monitoring and supervision sys- along with MIGA's executive vice president. The tems in MIGA have been minimal in the past, fol- extent of MIGA's business development, market- lowing several IEG recommendations, MIGA is ing, and outreach efforts involves a group of 26 gradually introducing greater monitoring and underwriters (operational staff), 2 staff responsible evaluation of its projects. This trend is in line for syndications, 8 lawyers, 3 staff responsible for with MIGA's broad development mandate, but it postcontract management, and 24 staff in an ex- needs to achieve this efficiently. Thus, there has ternal outreach department. Yet a series of client been some convergence with respect to the de- studies since 1995 have consistently noted the gree of involvement in project appraisal and su- 7 2 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G pervision by the three WBG institutions in their This flexibility allows clients to purchase exactly guarantee-related operations. what they need, depending on the circumstances of the project. A clear demand exists for this flex- Overlaps, Competition, ibility. More than 40 percent of MIGA guarantee and Market Niches projects to date have been for a single risk, and As comprehensive insurance, IFC's PCGs 65 percent were for either one or two risks. cover both PRI and commercial risk. The main feature of IFC's PCG is its coverage of all Another distinct feature of MIGA's PRI is that it can risks, both commercial and political (see table be appended at a very late stage in the develop- 3.4). The product has an advantage over traditional ment of a project without significantly affecting its political risk guarantees in situations where clear financial structure. It is therefore the least dis- definition and isolation of specific risks are diffi- ruptive project-financing structure, unlike the cult. However, PCGs have been expensive, as the Bank's and IFC's guarantee and PRM products. In "loan equivalence" approach to their pricing has most cases, especially for nonshareholder loans, tended to result in higher client costs than direct PRI coverage is required as a condition of loan ap- lending has. Consistent with these features, IFC's proval or disbursement, which helps the client ob- PCGs use to date has been mostly limited to local tain a longer loan period and better terms. MIGA currency financing needs in the banking sector in has been able to enhance its capacity to take ad- high-risk countries. vantage of this feature despite limitations posed by its Convention that require it to insure in- Unlike MIGA PRI and Bank PRGs, IFC's PCGs vestments prior to the investor committing any have not been used in a significant way in the in- amounts. MIGA's product is also the most stan- frastructure sectors. That indicates limited use dardized among the WBG instruments for PRM, of PCGs to meet demand for specific, clearly iden- which allows for quicker deployment. tifiable political risks, such as breach of contract. Nevertheless, market studies show increasing de- Thus, MIGA's PRI has comparative advantages mand for comprehensive guarantees. over other WBG products in situations where specific political risks can clearly be identified, Following the Argentine crisis in 2001 in partic- isolated, and managed; the project is at an ad- ular, clients that had comprehensive guarantees vanced stage of development and possible fund- were able to recover their investments, whereas ing sources have been identified; and the client those that had PRI only were not able to, be- wants specific, narrowly defined risk mitigation, cause their policies did not cover default caused as in the case of banks seeking relief for regula- by economic difficulties, local currency deval- tory capital. uation, or borrower insolvency (Political Risk Insurance Newsletter 2007). ADB, which offers IFC's B-loan product can substitute for both comprehensive PCGs and PRGs to the pri- MIGA's transfer and convertibility insurance vate sector, has seen stronger demand for its in some situations, although it can com- PCGs than its PRGs. This trend toward compre- plement other types of MIGA political risk hensive coverage has also been observed in the coverage. The implicit political risk coverage of practices of some export credit agencies and pri- IFC's B-loan structure can reduce banks' regula- vate financial institutions. tory capital requirements and address specific provisioning and country exposure limits con- MIGA offers a flexible product and a recog- cerns (Hays, Audino, and Cavanaugh 2001). IFC's nized brand name. MIGA has a well-recognized B-loans have been used predominantly in the brand in the industry. Its clients can buy single risk manufacturing and financial sectors, both sec- coverage or a combination of any or all of the four tors with relatively low political risk intensity. risks, making it the most "unbundled" product Nearly 70 percent of the loans have been in within the WBG. medium- to low-risk countries (mostly in the 7 3 74 THE Table 3.4: Comparison of the Main Attributes of the WBG Political Risk-Mitigation Products WORLD Weaknesses 1: Weaknesses 2: inherent policy- Risk How investor product driven BANK coverage benefits Strengths limitations limitations Opportunities Threats IFC PCG Comprehensive Receives comfort,a Provides both commercial and Expensive; does not cover Unable to separate Strong market demand Substitutable by GROUP improved credit terms political risk in one product; equity investments risks; last resort for Cl; meets local direct IFC loans covers local currency financ- product currency demand ing; applicable where risks are unclear GUARANTEE IBRD/IDA PRG Transfer/convertibility, Receives comfort, Increases government com- Demand mainly limited to Sovereign guarantee Continued demand in Improved regulatory expropriation, war, improved credit terms, mitment to success of proj- PPPs and sectors with heavy required in all cases high-risk low-income environment breach of contract, and is not liable for ect; accompanied by policy government engagement in cumbersome process- countries with untested reduces demand; some commercial risk, loan repayment dialogue and so forth high risk countries; some ing; high transaction regulatory environments; limited commercial some natural events heavy baggage inherent costs; debt only; last OBA PPPs resort product; limited INSTRUMENTS marketing IFC loans or Transfer/convertibility, Receives comfort Cost-effective; can provide No assured compensation to Lower-risk situations Failure to ensure equity invest- expropriation, war, adequate cover in lower risk client increasing compensation in a ments with breach of contract situations and where clients few cases could IPRI as main are more familiar with coun- end demand for reason try and absorb risk on own IPRI 1990­2007 MIGA PRI Transfer/convertibility, Receives comfort, Flexible coverage of all PRI No comprehensive coverage No stand-alone debt; Large available capacity; Declining market expropriation, war, improved credit terms, risks; main product for equity (commercial risk and political equity coverage; capacity for higher risk for traditional PRI breach of contract mediation services and investments; dispute resolu- risk cover); no local currency exclude domestic given excellent claims only; bureaucratic compensation in the tion; MIGA brand name in the loan financing coverage investors; lengthy history; potential to baggage increasing event of loss PRI market; high PRI special- process to change increase demand by to keep up with ization and expertise; possi- Convention limitations removing policy WBG demands ble to add at end of financial limitations structuring; minimal time and processing; fits check IFC B-Loans Transfer/convertibility Receives comfort, Package of IFC services; cost Debt only; no assured Combining with PRI, only improved credit terms effective compensation to client guarantees, securitizations IFC GOLF Transfer/convertibility Receives comfort; No waiting or evaluation peri- Single risk only, complicated None Cross-border only compensation in the ods prior to claim structure, opportunity costs of securitizations event of a loss funding liquidity facility IFC CELT Transfer/convertibility, Carves out political Package solution to various Uncertainty as to eligibility None International banks with expropriation, war, risks funding, risk management, for c-l relief, provisioning subsidiaries in emerging breach of contract and PRI needs of clients markets Source: IEG. Note: CELT = Credit-Enhanced Lending Transaction; GOLF = Global Offshore Liquidity Facility; IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; IPRI = implicit political risk insurance; MIGA = Multilateral Investment Guarantee Agency; OBA = output-based aid; PCG = Partial Credit Guarantee; PPP = public-private partnership; PRG = Partial Risk Guarantee; PRI = political risk insurance; WBG = World Bank Group. a. The WBG will act to resolve disputes as they arise. T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G Europe and Central Asia and Latin America and the Bank's PRGs is quite similar. Both coverages focus Caribbean Regions), suggesting that their risk on infrastructure projects in high-risk countries. mitigation might apply in relatively lower-risk sit- There are some important differences, however, uations or in countries where banks may have in that the Bank PRGs are packaged with unique reached exposure limits. Although B-loans can policy dialogue services. Nearly all Bank PRGs meet the demand of commercial banks for PRM, were deployed in support of large, complex PPPs, unlike MIGA PRI, they do not provide explicit where a proactive government role was important cover.3 to project success. Bank engagement in these cases went well beyond the provision of PRI in it- There might still be a demand for traditional PRI self to provision of a range of value-added services. to accompany B-loan products, and in these cir- However, Bank engagement and requirements cumstances, the products are complementary. also added significant time and financial costs for IFC has worked with private PRI providers to in- the project sponsors. clude options for B-loan participants to purchase certain types of political risk coverage. There have Deploying the Bank PRG has also required pre- also been instances where an IFC B-loan has been existing Bank programs and involvement with combined with MIGA PRI, as in the Manila North the government. To some extent, the product Tollways Corporation project. has been also rationed by supply-side constraints, and there has been a self-selection mechanism at In some circumstances, IFC's implicit PRM work in that only if the demand for PRI is ac- can be sufficient comfort to investors. From companied by the demand for strong dispute- the client's perspective, an IFC straight loan may, resolution potential and proactive government under some circumstances, be a substitute for a engagement will a PRG be sought. commercial loan plus formal PRI. IFC's IPRM may offer some advantages over traditional PRI. For ex- GOLF offers similar coverage to MIGA's trans- ample, investors that "self-insure" might find IFC fer and convertibility cover, although it is not participation itself to be adequate. Moreover, ob- a self-standing product and has only been taining a commercial loan plus PRI adds costs used once. GOLF has several distinctive features and time to closing a transaction, given the in- relative to traditional PRI products: (1) it is not of- volvement of multiple parties who will each con- fered as a stand-alone product but as part of a duct its own due diligence procedures and have package with other financing; (2) as such, there its own documentation and requirements. Such is no additional due diligence required for IFC to transaction costs may be a more important factor offer the product, and additional documentation in smaller projects: 50 percent of IFC's projects is minimal; and (3) it has the characteristics of self- with IPRM were smaller than $50 million.4 insurance and this limits the product to only those countries in which borrowers perceive the In addition, IPRM may be an attractive option in currency transferability and convertibility risks to situations where political risks, although pres- be minimal. IFC's GOLF provides single risk cov- ent, are general and difficult to specify and isolate erage for transfer and convertibility risk. With 40 in advance and are lower (about 65 percent of percent of MIGA's business to date being single IFC's IPRI projects were in medium- or low-risk issue transfer and convertibility risk, GOLF over- countries). At the same time, there will always be laps closely with MIGA's transfer and convertibil- clients who either have no funding needs but ity risk cover. IFC, in fact, markets the product as still want PRI or want explicit insurance that as- a more streamlined version of conventional PRI. sures them of compensation in the event of losses. For investors, the main advantage of the GOLF The Bank's PRG and MIGA's PRI compete in structure is that there are no waiting or evaluation breach of contract coverage. The breakdown periods of the type typically associated with PRI of breach of contract coverage of MIGA and the policies. Traditional transfer and convertibility 7 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 coverage typically require a waiting period of 60 in other situations, the same client bought trans- days to evaluate whether an event has met the pol- fer restriction and expropriation of funds cover- icy conditions, but GOLF is automatically trig- age from MIGA. Thus, investors still choose the gered if the issuer is unable to transfer money to type of risk-mitigation instrument that suits their an offshore account and unable to service inter- specific needs. est payments to bondholders. GOLF is also ad- vertised as economizing on transaction costs There are significant overlaps between the vis-à-vis working with traditional PRI providers. client bases of the three institutions. All the PRM products of the three WBG institutions serve Given that GOLF has been deployed only once, essentially the same broad group of clients. All of it is hard to test this assertion. However, IFC im- MIGA's top 10 guarantee holders are also existing plemented two similar securitization transactions: IFC partners. More than 80 percent of MIGA's one with MIGA PRI and the other with GOLF in- top 50 guarantee holders are IFC partners, and stead of MIGA participation. The first transaction close to 60 percent of existing MIGA guarantee took about three months longer to move from holders are also IFC partners. Similarly, most of concept to commitment stage than the second. the beneficiaries of the Bank's PRGs are also Given that it typically takes PRI providers less clients of IFC and/or MIGA. time to conduct due diligence on projects than it takes credit providers, it is unlikely that MIGA The overlapping clientele highlights what has involvement per se can cause delays in closing been observed about demand for guarantees, such transactions. particularly PRI: many investors seek a variety of PRI coverage types to suit the specific needs and IFC's CELT is an alternative to traditional PRI circumstances of their investments. This has been for banking sector clients. CELT is a funded a consistent finding in the various market stud- transaction in which IFC takes the political risk of ies commissioned by MIGA over the years-- the host country and the credit risk of the parent investors do not purchase all their PRI from a sin- company to the eligible borrowers. It was first in- gle provider (Booz Allen Hamilton 2005; Moran troduced in 1997 with ABN-Amro in Kazakhstan. and West 1998). It has been used since then in transactions with Societe General and Bank Intesa in Ukraine, Rus- There is strong anecdotal evidence that sia, and Serbia. Overall, IFC has completed four overlaps have led to some competition be- CELTs for a total of $475 million. tween agencies and confusion among clients. Interviews with staff and clients indicate that there PRM is an important aspect of the transaction. have been instances where staff of different in- From the client's perspective, the structure sub- stitutions have approached the same client in- stitutes for PRI; IFC has, in fact, marketed the dependently, each institution has worked on the product as a PRI substitute. MIGA could have same project without the knowledge of the other, provided PRI in all CELTs if those were loans from or clients have been in a position to play one in- the parents. It is important to note, however, that stitution against the other. The survey adminis- there are important differences between a CELT tered for this evaluation also produced examples and traditional PRI, including the fact that the of overlaps and competition. World Bank and CELT is a funded transaction and PRI is not. A dis- MIGA documents on the WBG guarantee instru- tinctive advantage of the CELT over a traditional ments have also recognized issues of overlap that PRI is the package solution to various funding, risk have led to competition among the three insti- management, and PRI needs of the client. Two of tutions and confusion among clients as to who is IFC's CELT clients are also important MIGA clients doing what. Some of the various products' dif- and are familiar with MIGA's product. IFC's field ferentiating features, such as counter-guarantee presence has also been a factor in the client's by the host government, are on the supply side choice of CELT over MIGA PRI. At the same time, and are thus invisible or immaterial to clients. 7 6 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G The evaluation did not find strong evidence antee products that embed political risk coverage. that overlaps have led to loss of business op- Thus, complex relationships of both substi- portunities, but competition has often re- tutability and complementarity exist among the sulted in higher costs for clients. In judging WBG PRM instruments. That implies a need for how severe the overlaps are, it is important to coordination and cooperation. note that (1) the evaluation did not find clear ev- idence that overlaps and competition have led to Cooperation on Projects-- loss of business opportunities for the WBG, but Efforts and Results they have tended to impose excessive costs to MIGA and the Bank have both provided PRI clients; (2) overlaps, imperfect coordination, and in five large high-risk projects. Four infra- approaching clients in an uncoordinated fash- structure projects involved both MIGA and a Bank ion are features present to some extent in any PRG (table 3.5). MIGA also participated in the large, complex, and decentralized organization, unsuccessful attempt to wholesale PRI through the including within IFC and the Bank; and (3) MIGA Banque Ouest Africaine de Development in West perceives overlaps as more threatening, as it has Africa. They were large projects, averaging more the most restrictive product space of all the WBG than $900 million. Several were cross-border, institutions. including the Lao PDR Nam Theun 2 project-- where the dam was in Lao PDR and most of the A weakness in the current structure is that off take was in Thailand--and the West Africa products are not offered as a single menu Pipeline, which involved four countries--Nige- of options to prospective private sector ria, Benin, Togo, and Ghana. The large size of the clients. Because clients may perceive that the projects and covered amounts as well as the com- WBG institutions are poorly coordinated and plexity meant that a single agency was unlikely to compete with each other for business, there is a be able to provide full coverage. This led to the reputational risk for the WBG. The WBG needs to participation of multiple PRI providers, including facilitate client choice in a coordinated fashion by the Bank and MIGA. presenting a single menu of options. Cooperation can involve higher transac- One approach is to promote common marketing tions costs for clients. In some cases, cooper- arrangements among the three institutions. IFC ation came at a price, however. The involvement is in a strong position to identify projects for the of various parties tended to fragment financing WBG's full range of risk-mitigation products in that and increase transaction costs. Large projects it has the broadest network of private sector such as Uch Power or Lao PDR Nam Theun 2 in- clients and extensive relationships with interna- volved a large number of financiers and added to tional commercial banks, which are key drivers complexity and transaction costs. In the case of of the demand for PRI.5 Thus, it may be advisable the West Africa gas pipeline, for example, PRI for the Bank, MIGA, and IFC to complement their from IDA, MIGA, OPIC, and a private sector own marketing efforts with common marketing provider (Zurich) required the project sponsor to arrangements that specify incentives for pro- enter into multiple contractual agreements with moting the products of all three institutions. different structures, coverage, and mechanics. Large projects, such as Uch Power or Lao PDR Nam Overlaps imply a need for coordination, Theun 2, involved a large number of financiers, and complementarities offer opportunities and PRI from multiple parties added to com- for cooperation. Although the guarantee prod- plexity and transaction costs. It is unclear what ucts of each institution were originally intended benefits the sponsor of the Lao PDR power proj- to complement rather than compete with each ect gained from having three PRI coverages of $42 other, some of the distinctions have eroded with million each (from IDA, MIGA, and ADB). Bene- changes in the product mix as well as the market. fits to MIGA from these joint projects were risk Potential competition also comes from nonguar- sharing and reduced due diligence costs, given the 7 7 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Table 3.5: Joint Bank-IFC-MIGA Guarantee Projects Fiscal Project MIGA IFC PRG Other PRI Country year name role role role providers Joint Bank-IFC Pakistan 1996 Uch Power Project B-loan of $75 million PRG for Export-Import Bank and A-loan of $75 million PRI for $153 million $40 million Côte d'Ivoire 1999 Azito Partial B-loan of $30 million PRG for Risk Guarantee and A-loan of $35 million $30 million Senegal 2005 Kounoune Power A-loan for PRG for $72. (67.5 megawatts) $20.624 million million, credit of $15.7 million Joint Bank-MIGA Ghana, Nigeria, 2005 West African PRI for $75 million PRG for Steadfast Insur- Benin, Togo Gas Pipeline (IDA $50 million ance Co. PRI for S/UP) $125 million; re- insurance by OPIC Lao PDR 2005 Lao PDR Nam PRI for $90 million PRG for ECA coverage of Theun 2 Power Proj- $42 million $200 million; ADB ect (formerly under PRG for $42 million PE-P004206-LEN) Joint Bank-IFC-MIGA Mozambique 2004 Southern Africa PRI for R90 million of Equity investment of PRG for R140 Of the downstream Regional Gas the upstream and $18.5 million in million of the MIGA guarantee, Project R630 million of the upstream portion upstream and R70 R310 million was downstream million of the reinsured ­ R155 downstream million each by SACE of Italy and EFIC of Australia. Uganda 2007 Private Power PRI for $115 million A-loan of $100 million; PRG for Generation C-loan of $30 million $115 million (Bujagali) Project Joint IFC-MIGA Mozambique 1997 MOZAL PRI for $40 million A-loan of $108 million Angola 1998 AEF Flecol PRI for $2.3 million A-loan of $0.61 million Slovakia 1999 EuroTel. Brati. PRI for $26 million A-loan of $27.49 million Philippines 2001 Manila Tollways PRI for $85 million, A-loan of $45 million plus $22 million equity Dominican Republic 2005 Basic Energy PRI for $11.1 million A-loan of $22.65 million Argentina 2005 Banco Galicia CL PRI for $58.9 million A-loan of $40 million Southern Europe 2006 Mercator Retail PRI for $20.3 million A-loan of $51.23 million Russian Federation 2006 Bema Warrants PRI for $364.8 million A-loan of $39 million Uruguay 2007 Orion PRI for $300 million B-loan of $70 million Peru 2007 Lima JCI Airport PRI for $11.5 million Equity of $20 million (Phase I only) (Phase II only) Central Europe 2005 BalAEF MBS PRI for $10.1 million A-loan of $7 million Source: World Bank, IFC, and MIGA data. Note: EFIC = Export Finance and Insurance Corporation; FY = fiscal year; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency; OPIC = Overseas Pri- vate Investment Corporation; PRG = Partial Risk Guarantee; PRI = political risk insurance. 7 8 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G Bank's full economic, technical, social, and envi- fiscal management along with IFC's financial ronmental appraisal. structuring experience and ability to finance with- out a sovereign guarantee. The PRG has been combined with IFC lend- ing to increase the amount of financing for Several issues arose during the experience, how- a few large high-risk projects. Joint Bank­IFC ever, including the availability of Bank staff, who PRG financing took place in the Côte D'Ivoire tend to be programmed more in advance than the Azito project, the Pakistan Uch Project, the Mozam- IFC, within a timeline imposed by the client's bique (Southern Africa Regional Gas Project) gas budgetary authority to issue the bond; and the pipeline, and the Uganda Bujagali project. MIGA question of how Bank staff would be recognized also provided coverage to the investors of these for their contributions that would not result in a last two projects. Bank product and how their participation in sub- sequent supervision of the project would be In all these countries the Bank had a longstand- funded. Issues identified in other subnational proj- ing policy dialogue and had been involved in de- ects include limited staff familiarity with the prod- veloping the regulatory framework for private ucts and processes of other Bank institutions, the sector participation. The project in Côte D'Ivoire possibility of delays caused by parallel decision was the first PPP in the power sector in Africa, and making in the Bank and IFC, and finding the right the Mozambique project was the first to test balance in the extent of policy/institutional re- the new legislative framework in the gas sector forms that should be included in a capital market and the only IBRD enclave guarantee. In these operation that funded public expenditures. contexts, even with the participation of IFC, ad- equate levels of commercial finance could not IDA lending has also been combined with be raised and the sponsors approached the Bank IFC PCGs to enhance SME access to finance, for a PRG. although difficulties persist. An important new area of cooperation between IFC and the In Pakistan, two successive power projects illus- Bank is the SME program in Africa. This program trate the potential synergies of IFC participation. combines IFC's PCGs and IDA funding. So far, Under the first project (Hub) in which IFC did not two joint projects have begun in Madagascar and participate, the Bank's PRG helped obtain 12- five projects are under preparation. The programs year loan maturities with a guarantee for 23 per- have two participating banks, which have dis- cent of the project cost. In the subsequent Uch bursed about 64 percent of the $25 million total project, in part because of greater investor fa- for SME loans. miliarity with the country but also because of IFC's participation, 15-year maturities were ob- However, several issues have arisen out of the ex- tained with a PRG that covered only 12 percent perience: Different processes and misaligned in- of project costs. centives have made it difficult to mobilize IDA resources for first-loss provisions; the need to fit Subnational finance exploits the compara- these operations into existing IDA loans or op- tive advantages of the Bank and IFC. The joint erations in the country has been a limitation; and IFC­Bank Subnational Finance Department has procurement and conflict-of-interest policies have institutionalized some synergies between IFC added to transaction costs. A lack of continuity of and the Bank in subnational finance. In fiscal staff working on these transactions has limited op- 2006, the Bank and IFC collaborated on an IFC portunities to replicate the program. guarantee operation to support the Chuvash regional government in Russia. The experience Thus, the IDA­IFC SME programs have had high involved dual roles that exploited the Bank's transaction costs and have been difficult to repli- relationship with regional governments and its cate. To some extent, high transaction costs are strengths in public policy, the social sectors, and inherent in the development of new products 7 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 and solutions. Lessons from this first experience A generic solution is needed to avoid miscom- are being reflected in ongoing work and are likely munication with clients and to enhance oppor- to lower the development costs of new projects tunities for the two institutions to cooperate on under preparation. projects. This issue has not been an impediment to MIGA working jointly with other investors; in The pledge of shares and sharing of arbitral cases where the issue has arisen, it has been re- awards issues have limited direct coopera- solved through a claims cooperation agreement. tion between IFC and MIGA. IFC and MIGA MIGA requires that it cover the shares as evi- have jointly been involved in about 10 projects dence of ownership of the investment. Such clear over the last 10 years. Most of these joint projects evidence is critical for recovery from the host have been relatively large transactions in the in- government. The claims cooperation agreement frastructure and financial services sectors, where for the Bujagali hydroelectric power project pro- MIGA has typically provided political risk cover- vides a template to resolve the pledge of shares age on the equity (or junior tranches in the cases and sharing of arbitral awards, but it needs to be of securitization) side of the financing and IFC has institutionalized for joint WBG projects. provided debt financing. In one cross-border high-risk transaction in the manufacturing sector, Cooperation has been driven as much by MIGA's PRI has been critical in providing comfort clients' preferences as by internal institu- to the investor to make a large equity investment. tional reasons. There have been benefits to In the first cross-border securitization of resi- joint WBG projects, but their occurrences have dential mortgages in Eastern Europe, a MIGA been relatively few and cooperation has tended guarantee ensured that the notes pierced the for- to add to costs. In some circumstances, WBG eign currency country "ceiling." products can be direct substitutes rather than complements. In such cases, internal pressures to A few issues have emerged in the limited cases promote joint projects can bring few gains and sig- where the two institutions have worked together. nificant costs for clients. The objective should be In some instances, IFC staff working on joint proj- to enable deployment of the product with the best ects had the impression that MIGA's involvement fit for client needs rather than to promote joint added to processing time. IFC implemented two products for internal institutional reasons. The similar securitization transactions: one had MIGA principle of minimal WBG involvement is em- PRI and the other did not have MIGA's participa- bedded in various WBG policies; basically, WBG tion but had similar enhancement that IFC pro- involvement is expected to be kept at a mini- vided. The first transaction took about three mum to make transactions possible. months longer to move from project data sheet- extended review to commitment than the second. Coordination Mechanisms IFC management has been consistently concerned Given the potential for overlaps, several about the possibility of IFC becoming a benefici- mechanisms to ensure coordination between ary of MIGA coverage under certain circumstances. the WBG have been established, although IFC management has also been reluctant to en- with varying effectiveness. Given both the po- gage MIGA in its B-loan program on the grounds tential synergies in products and the potential that minimal involvement of the WBG would be overlap in clients and the need to ensure that the preferred option when available and it would clients are not confused by the array of WBG risk- involve an unnecessary double use of the WBG's mitigation instruments, the need to ensure proper preferred creditor status. coordination among the three WBG institutions is clear. Efforts to coordinate deployment of guar- The pledge of share issue has also been a matter antees to reduce duplication and ensure com- of contention for some time, and although solu- plementarity have included (1) establishment of tions have been worked out, the issue re-emerges a hierarchy of instruments to govern the deploy- every time IFC and MIGA work on a joint project. ment of the various WBG products; (2) involve- 8 0 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G ment of all three institutions in the Country ditionality) were critical from a risk management Assistance Strategy (CAS) preparation process; and/or market point of view. Thus, Bank PRGs (3) establishment of a Guarantee Review Com- would be deployed when sector reform was in its mittee (GRC), chaired by the IFC executive vice early stages and the operation was larger and president, with representation from all three in- riskier and highly dependent on government sup- stitutions; (4) increased linkages at the sector, port or undertakings. policy, and operational levels among the three in- stitutions; and (5) staff training across the WBG These principles remain relevant, but internal re- to increase familiarity with the all products. How- views have acknowledged that implementation ever, as discussed below, although these efforts has been complicated by the fact that the hierar- have improved coordination to some degree, chy principle has often conflicted with the pref- they have not been fully effective. erences of governments and market participants. Thus, the hierarchy of instruments has been The products were designed to not com- mainly a supply-driven approach to the deploy- pete with each other, but their differences-- ment of instruments rather than a reflection of created by institutional mandates and market demand. Internal reviews have also noted policies--are somewhat artificial and not that the three WBG institutions have not always necessarily client friendly. In any particular adhered to the principle, sometimes pursuing circumstance, if policies are strictly followed, the same projects independently of each other. there is only one WBG institution that can meet a client's needs for a guarantee product for a par- The principles that govern the relationship ticular type of investment. Thus, by design, the between MIGA and IFC products have not guarantee instruments of the three WBG institu- been clear. The hierarchy of instruments prin- tions have distinctive features that make them ciple applied mainly to IBRD/IDA instruments complementary rather than substitutes. How- relative to other instruments, and the principles ever, several key differences among Bank, MIGA, of deploying MIGA PRI and IFC's instruments and IFC PRI products are driven by internal op- have not been articulated. The WAEMU Capital erational mandates and policies. This creates an Market Development Project established more de- artificial supply-driven product differentiation tailed principles of deployment of Bank PRGs, that is not necessarily aligned with patterns of de- MIGA PRI, and IFC-type PCGs. In the WAEMU mand and that forces a client to deal with a com- guarantee facility, IDA's PRGs, MIGA PRI, and pletely different institution based on the type of AfDB's PCGs and comprehensive guarantees investment. This has complicated the imple- were to be marketed by BOAD to prospective mentation of coordination mechanisms. investors/lenders as separate but complemen- tary products and deployed on the basis of proj- The hierarchy of instrument principle has ect profiles. Detailed guidelines for deployment provided some guidance, but its implemen- were prepared, but the facility was not success- tation has been difficult. In 1997 a hierarchy was ful, so the rules and guidelines for deployment established under which the deployment of WBG could not be tested. risk-mitigation instruments would adhere to the principle of market first, MIGA/IFC facilitating the IFC has marketed GOLF and CELT products as su- market second, and the World Bank (with its sov- perior to traditional PRI, creating some tensions ereign counter-guarantee) as a last resort. In 2000 between IFC and MIGA. It also appears that in mar- a further clarification was made on the deployment keting efforts, IFC has at times emphasized sub- of IDA/IBRD PRGs vis-à-vis IFC and MIGA products. stitutability between its implicit political risk The Bank's PRGs would be considered for de- coverage and formal PRI. This leaves potential ployment only when one or several of the fea- clients with the impression that if IFC is involved tures (explicit counter-guarantee, influence of the in a transaction, there is no need for MIGA's Bank, linkage to the Bank's sector dialogue, or con- involvement. 8 1 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Guarantees are sparingly featured in CAS cy- Several further mechanisms to coordinate cles. The initial expectation for IDA guarantees risk-mitigation products at the Regional, was that CASs would clearly establish whether sector, policy, and operational levels have guarantees were justified and the extent to which been established, although gaps remain. such instruments would be used. CASs were ex- MIGA has introduced a range of mechanisms to pected to indicate the sectors and projects where improve coordination with the Bank and IFC at guarantees would be more appropriate than loans. the institutional, policy, strategic, and operational levels. For example, it participates in several sec- A review of 40 recent CASs (20 IDA and 20 non- tor boards, has integrated its technical assistance IDA countries) was carried out to assess the services into the Foreign Investment Advisory extent to which guarantees were taken into Service, and is harmonizing its environmental consideration during the formation of CASs. Ref- and social policy and performance standards erence to potential deployment of a MIGA guar- with those of IFC's. More systematic consultations antee was frequent, but just 13 of the 40 identified between MIGA and Bank country and industry de- a potential use of an IFC or Bank guarantee. In partments have also helped ensure that MIGA- cases where guarantees were referenced, more- supported projects are consistent with the WBG's over, they were general in nature, with almost strategy in a country. For MIGA, this coordination no specifics on the potential application of the at the project level has been both a way to man- instruments. age risks going forward and a necessary condition to be able to use its relationship with the Bank A WBG-wide GRC had limited success in and governments to work out potential prob- harmonizing approaches, and it added to lem situations. transaction costs. In March 1999 a GRC was es- tablished to coordinate the Bank PRGs with the The same degree of consultation on IFC projects activities of IFC and MIGA. The committee was carrying IPRC is not apparent, however. Often chaired by the IFC's executive vice president and IFC's ability to deliver on its IPRC is premised on had some success in improving the tone of and the tacit or active cooperation of the Bank. There relationship between institutions, particularly be- is, however, no mechanism to ensure that the tween IFC and the Bank. Discussion and formal IBRD and IFC cooperate in the context of an IPRC endorsement of a potential PRG by the GRC event. IFC and the Bank have at times had different helped minimize internal disagreements, ensure views on government policies in the context of consistency with the hierarchy of instrument prin- specific IFC-supported projects; that can affect ciple, and enhance the prospects of MIGA and IFC IFC's ability to deliver on its IPRC. Given the participation in the PRG projects. reliance of IFC's IPRC on the WBG's relationship with governments, more coherent views and sys- The GRC was confined to reviewing Bank PRGs, tematic coordination are clearly desirable. however, and did not consider complementarities or overlaps of proposed IFC and MIGA risk- A degree of informal coordination has also mitigation instruments. Some staff perceived it been effective. Staff interviews indicate that a mainly as a mechanism to carve out spaces for IFC degree of coordination takes place informally and MIGA in PRG-supported projects, rather than between staff who share information, market in- to ensure the deployment of the most suitable telligence, and bring information to each other's instrument. The GRC was not successful in pro- attention. IFC's infrastructure department, for moting business, simplifying processes, or har- example, advertises the Bank's PRG and advises monizing approaches. In fact, it created a further clients to take advantage of it when appropri- layer that added to transaction costs. It ceased to ate. Bank PRG staff have from time to time re- exist in 2005, when the previous IFC executive vice ferred projects to MIGA or IFC; on a few occasions president departed and those functions were these referrals have led to MIGA or IFC projects. transferred to the Bank's Operations Committee. MIGA has also referred clients to the Bank's PRG 8 2 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G as a more appropriate instrument (such as the in each of the three institutions indicates that case of Phu My 2 in Vietnam). This coordination there is a high degree of unfamiliarity with the is ad hoc, however, and often depends on indi- products of other institutions. For example, only vidual relationships. 23 percent of IFC staff indicated that they were fa- miliar with IBRD/IDA PRGs, and less than half Limited coordination in the development of were familiar with MIGA PRI. In addition, there are new products is also apparent. The experience no incentives for staff, including those in field of- of IFC's introduction with products that carry fices, from one institution to promote the prod- PRI (CELT and GOLF) and the Bank's introduction ucts of the other. In the case of the municipal of guarantee facilities indicate that there is limited finance operation in Russia, for example, there was coordination in new product development. In- some question as to how Bank staff would re- formation exchange and some sharing of ex- ceive recognition for work that did not result in pertise have taken place, but there has been no a Bank product. systematic review of the implications of new prod- ucts across the WBG. The Quality Assurance There is significant potential for more sys- Group review of the Bank's guarantee facility in tematic links between Bank/IFC advisory Peru, for example, concluded that a more proac- services and the deployment of WBG risk- tive approach could have been made to engage mitigation instruments, particularly in in- MIGA/IFC expertise in the design of the operation frastructure. Synergies between WBG advisory and strengthen complementarity. It was observed services on regulatory reform and structuring that MIGA had had generally negative experience specific PPP transactions have also not been fully with guarantee facilities that could have been in- exploited. At present, IFC investment teams typ- structive and that IFC had significant experience ically get involved in a transaction when the struc- with infrastructure investments in the Latin Amer- ture of the deal has already taken shape. MIGA's ica and the Caribbean Region and had expressed involvement is usually at an even later stage, when reservations about demand that could have been the financing structure is largely in place. Some better explored. of the Bank's PRGs, however, have been intro- duced as part of the bidding documents for PPPs. In October 2006 IFC constituted a New Products Assessment Group to help ensure a coherent Greater upstream engagement of WBG risk- and efficient approach to new product develop- mitigation products might be enhanced with closer ment. The group assesses the expected risks and links to Bank and IFC advisory services that help benefits of new products and recommends establish an appropriate regulatory framework and whether the product should be introduced. develop a specific PPP transaction. IFC's corporate IBRD/IDA has established a similar group, the Fi- advisory services, in particular--that is, helping nance Instrument Subcommittee, which also governments design concession agreements and oversees financial product innovation. Given that privatization--represent a potentially important the products reach the same clients, an oppor- mechanism to more systematically introduce WBG tunity exists for the respective units responsible risk-mitigation products earlier in a project's de- for new product development in the three insti- velopment. As the CAS is involved in the initial tutions to establish links to facilitate mutual ex- stage of project design, it can include options in the change of inputs from their perspectives on new bidding package for the winner to take advantage product development. of Bank PRGs, IFC financing, or MIGA PRI. Stronger links offer the potential to coordinate WBG Lack of staff incentives and familiarity with involvement and offer an optimum product con- the products of the other institutions has figuration from the client's perspective. prevented exploitation of clear synergies in marketing the WBG's products. A survey The Bank applies a completely different ap- of staff who have worked on guarantee products proach to product pricing for private sector 8 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Table 3.6: Comparison of WBG Pricing (guarantee fees) Project Bank MIGA IFC West African gas pipeline project (IDA) .75 1.15 Nam Theun 2 power project, Lao PDR (IDA) .75 1.65 Private power generation project (Bujagali), Uganda .75 1.3 Southern Africa regional gas project, Mozambique (IBRD enclave) 2.0 1.3 IFC GOLF mortgage-backed securities, Russian Federation (IDA) versus 0.5%/ MIGA Raffeisen Leasing, Russian Federation (MIGA) 1.01 LIBOR + 2% Source: WBG data. Note: Guarantee rates are for principal amounts of ID guaranteed loans (IDA), aggregate principal amount (IBRD), and current coverage (MIGA) and do not in- clude processing, front-end, initiation, or standby fees. IFC and MIGA offered the same coverage for their comparison, and the Bank had equivalent or greater coverages than MIGA. IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; LIBOR = London Interbank offered rate; MIGA = Multilateral Investment Guarantee Agency; WBG = World Bank Group. clients. In 2004, MIGA introduced a model that a good system for pricing specific risks; in the few provided pricing guidance for its guarantees based instances where it has covered political risks, as on cost plus risk. In 2007 it refined the pricing in the case of a CELT transaction, it has used model to account for financial, market, and pol- MIGA pricing as a reference. Both IFC and MIGA icy considerations. Current actual pricing con- prices are sensitive to market signals, but Bank tinues to be based on cost plus risk, with some pricing is not. adjustment to account for market factors. A more uniform approach to PRM pricing Application of this model still has some weak- could help eliminate potential market dis- nesses--such as the inability to accurately capture tortions and reduce the need for hierarchi- per project administrative costs--yet it represents cal application of instruments. The Bank's set a pricing approach that reflects the nature of the schedule of PRG pricing is not consistent with the risks being covered. In contrast, Bank PRGs are not pricing approaches MIGA uses for its PRI. The priced for sector, country, or type of risk covered Bank's long but unsatisfactory experience with Fi- but are instead offered across countries and sec- nancial Intermediary Loans--the only other Bank tors according to a set schedule of charges that are instruments that reach the private sector--re- based on loan-equivalent pricing. A comparison of vealed that undermining commercial lending rates the Bank's PRGs and MIGA's PRI in four joint proj- led to market distortions and crowded out the ects indicates that pricing of IDA PRGs was lower private sector (IEG 2006). Table 3.7 presents ben- than MIGA's PRI in three IDA projects; however, efits of a more uniform approach to PRM pricing. the IBRD enclave PRG in Mozambique, priced at IBRD terms at the time, was priced higher than The Bank's Operating Principle 8.30 also requires MIGA's PRI. that in financial intermediary lending, on-lending occur at or near market rates. Should this logic be A comparison of IFC's GOLF and MIGA PRI in a extended to PRGs, which also reach private sec- project in Russia, for which both provided trans- tor clients, then PRG pricing should be revisited fer and convertibility coverage, indicates that and perhaps more closely aligned to MIGA's the cost of IFC GOLF is much higher if coverage model.6 This alignment would reduce the po- is triggered but is half the price if the facility is tential for market distortions as well as reduce the not used (see table 3.6). IFC is piloting and using need for the rationing on the supply side through a new capital pricing and risk approach for in- the hierarchy of instruments principle, in that it ternal risk-management purposes, including would allow the most appropriate WBG product pricing. Nevertheless, IFC has not yet developed to be deployed according to the particular cir- 8 4 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G Table 3.7: Benefits of a More Uniform Approach to PRM Pricing Benefits How benefits would unfold Help eliminate market distortions The WBG guarantee operations are part of a broader market that measures and prices political and commercial risks. By adopting a more uniform pricing approach that more closely reflects actual risk and cost, the WBG can help avoid giving the market confusing signals regarding risk pricing, thus helping enforce market discipline and efficiency in resource allocation. Explicitly recognize pricing incentives By adopting a more uniform pricing approach that more closely reflects provided to clients actual risk and cost, the WBG can more clearly recognize and track the pricing incentives given to various types of clients to pursue strategic development objectives and maintain competitiveness in the market for PRM. Monitor and improve delivery efficiency By adopting a more uniform pricing approach that identifies the cost of administrating and delivering guarantees, the WBG can better track the factors that impede delivery efficiency. Improve the allocation of risk capital Each of the WBG institutions has recourse to a capital base that can help it meet any guarantee obligation that arises. By adopting a more uniform pricing approach that incorporates the cost of this underlying risk capital, the WBG can better recognize and optimize the costs of capital incurred in its guarantee operations. Source: IEG analysis. Note: PRM = political risk mitigation; WBG = World Bank Group. cumstances of the project, rather than according The demand for MIGA PRI is affected by the level to price. of loan spread in the market, but MIGA as a single- product firm does not have the flexibility to sub- Such an approach may at times reduce the price sidize across products.8 These differences present of a PRG, such as the case of the IBRD enclave op- some challenges to the application of a consistent eration in Mozambique. In other cases, higher pricing approach across the three institutions. fees charged to the private sector firm might be transferred to the government to offset costs as- Improving Delivery: Some Organizational sociated with issuing a counter-guarantee. For Realignment Options IFC, an important although difficult question is Given the existence of both overlaps and whether efforts should be made to unbundle the complementarities among the WBG PRM in- implicit political risk coverage embedded in IFC's struments, Board members and manage- traditional products.7 It is important to note that ment have asked about alternative ways to in today's market reality, the major players are typ- organize the delivery of WBG guarantee ically financial conglomerates, which often take a products so that synergies are maximized relationship approach rather than product-by- and redundancies eliminated. In this context, product approach to pricing. Under such an ap- efforts have been made through several internal proach, loans are typically the loss leaders, and notes and presentations to take a fresh look at the market players look at investment banking, equity, way the delivery of guarantee products is orga- and other products for return. Thus, multiprod- nized within the WBG (World Bank 2005, 2007b). uct firms tend to cross-subsidize across products. Various options have been presented, ranging from the status quo with increased coordination As a multiproduct firm, IFC can also be forced by at the level of country/sector strategy, environ- the market to cross-subsidize across products. mental analysis, and joint Board papers to explicit 8 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 proach applied in the case of joint projects could Table 3.8: Simple Comparison among be extended to include MIGA and Bank PRGs. Organizational Options Minimizing Potential benefits. This approach could combine conflicts the relative strengths of IFC and MIGA with respect Efficiency of interest Coordination to working with private sector clients: the mar- Client integrated High Medium Medium keting, business development, and structuring Country integrated Low Low High capacities of IFC, and the processing and book- Product integrated Medium High Low ing processes of MIGA. IFC would be able to offer Source: IEG analysis. private sector clients a richer package of financial products that would now include stand-alone institutionalized incentives and policies for staff col- PRI. The expanded menu of products could cre- laboration, to organizational integration, mean- ate efficiencies through economies of scope in ing jointly administered staff, budget, and targets marketing a variety of financial services to pri- (World Bank 2007c). vate sector clients. This approach would provide a single point of entry and address the coordina- With the objective of making a better use of the tion issues among IFC, MIGA, and the Bank. It has array of guarantee instruments within the WBG the potential to minimize internal competition, to better attain its mission, this evaluation pre- overlaps, and client confusion. sents some options involving organizational re- alignment and briefly analyzes their potential Potential adverse effects. Conflicts of interest benefits and costs. The options examine three may arise, given that IFC would be making the alternative integration perspectives: client, geo- decisions, but consequences could be borne by graphic market, and product (see table 3.8). the balance sheets of MIGA and the Bank. There may also be a conflict of interest in IFC's deter- Private sector­client-integrated approach mination of which product to recommend to a Description and rationale. Under this ap- client. Given IFC's preference for funded solu- proach, all products for private sector clients tions, this approach may result in fewer choices would be offered in an integrated fashion. Of all for private sector clients relative to the status WBG institutions, IFC has the most extensive quo. MIGA's PRI has often been an unfamiliar or relationships with private sector clients, the most misunderstood product by IFC staff and--in most systematic business development capacity, and cases--perceived as a second best option to IFC's presence on the ground. IFC also offers the clients. broadest set of products to private sector clients. Thus, under this approach, the decision making This arrangement could thus add another con- regarding the delivery and deployment of risk- straint for the growth of MIGA's business. An- mitigation products--including political risk-- other potential issue is the incentives to offer would be placed within the IFC. The separate PRGs under this approach. Demand for Bank legal identities and balance sheets of MIGA, IFC, PRG solutions has often come from governments, and the Bank (with respect to the PRGs) would and a private sector­client-focused structure may be preserved. not be able to capture this source of demand ef- fectively. Further, a number of PRG transactions In this scenario, marketing, business develop- involved years of preparation and discussions ment, product choice, and supervision of PRGs with the governments, and it is not clear whether and PRI products could be led by IFC, and MIGA the incentives to maintain the long-term dialogue could do underwriting. For the purpose of align- with the government and sponsors would be pre- ing incentives, IFC's internal system for double served under IFC leadership. Thus, under this booking investments between IFC's regional and option the IDA/IBRD PRG product may see even industry departments and the joint venture ap- less deployment than now. 8 6 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G Country-level integrated approach however, is not an indicator of the likely success Description and rationale. Under this ap- of this approach, as these groups do not involve proach, coordination and integration of guaran- integration at the country level. tee products would take place at the country level. The developing member country is viewed Another potential negative of this approach is as the ultimate client, and all the tools at WBG's that it could add to processing time and transac- disposal would be used to further the country's tion costs, as it may rely more heavily on Bank pro- development priorities. The deployment of WBG cedures and processes in dealing with private guarantee products is thus coordinated at the sector clients. A possible inefficiency would be the country level, in close partnership with the coun- loss of staff underwriting skills if it is carried out try's government, rather than with respect to by the World Bank. Staff knowledge of guarantee private sector client needs. This approach would products and underwriting expertise at the coun- imply that all decisions about the deployment of try level would have to be built. WBG guarantee instruments would be in the purview of a common management arrangement, Product-level integrated approach taking into consideration the full array of WBG Description and rationale. Under this ap- products available. proach, all insurance products would be under one roof. As the WBG institution specializing in Potential benefits. Such an approach would en- PRI, MIGA would consolidate all the PRM and in- hance coordination among the deployment of surance products of the WBG. Thus, IDA/IBRD all WBG products, not just among the guarantee PRGs would be delivered under the decision- instruments. As discussed earlier, substitutability making leadership of MIGA. This option could also and a certain degree of competition exist between entail expanding the scope of MIGA by enriching some guarantee and nonguarantee products; co- its mandate with other insurance products (in- ordination at the country level could address this cluding commercial) not currently offered by the issue by facilitating deployment of all WBG prod- WBG, to keep up with trends in demand and ucts in line with country needs. With respect to market developments. This would not affect the PRM, the advantage of this approach is that it deployment of IFC's existing products. builds on WBG's comparative advantage in miti- gating political risks through close partnership Potential benefits. This approach would allow with the government authorities. greater specialization in the delivery of guarantee products. Structuring and assessing the risks of Potential adverse effects. This approach may guarantees could be quite different than for loans entail conflicts of interest, particularly between IFC and equity investments. Thus, it might be argued investments and Bank policy advice. Conflicts of that clients would be better served technically interest are much less pronounced in the rela- by such specialization. A dedicated product-based tionship between MIGA and the Bank, as MIGA structure would also minimize potential conflicts does not benefit from the commercial success of of interest in the offering of funded versus non- a project. Thus, MIGA does not face a conflict of funded solutions. MIGA would reach a critical interest in its dealing with the Bank country di- mass that would allow it to take more risks, in- rector, where the Bank director may be providing novate on a more systematic basis, and develop advice that may have an impact on the commer- systematic business development capacity. Using cial profitability of a project. MIGA's relatively lenient processing to deliver Bank PRGs would result in efficiency gains. The joint World Bank­IFC operational depart- ments (oil, gas and mining, and communications Potential adverse effects. The approach may and information technology) have developed not address the competition to PRI from tra- procedures to deal with such conflicts of interest. ditional nonguarantee IFC products that embed The experience of the global product groups, implicit political risk coverage, as well as the 8 7 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Box 3.1: The 2005 Proposal to Develop Synergies between the World Bank and MIGA Guarantees Rationale. Similarities between MIGA's breach-of-contract cov- investment." The proposed integration was expected to increase erage and IBRD/IDA's PRG provided the impetus to create a work- the relative strengths of the Bank and MIGA for optimal service ing group comprising the Bank's Project Finance and Guarantee delivery to private sector clients and governments, enhance the Group and MIGA that would institutionalize the collaboration be- coordination of product offerings, close potential product gaps by tween the two "so that synergies for the use of guarantees as an providing a "one-stop shop" for guarantees, and create an en- important tool in mobilizing the private sector for infrastructure hanced platform for sharing knowledge and facilitating product delivery are aligned and potential synergies between the various innovation and structuring of new forms of risk sharing. guarantee products are fully exploited." Demand was growing for these guarantee products, and despite previous joint guaran- The Outcome. The working group made recommendations about tees in several projects,a there was confusion among the private the organizational structure and modalities for integration, such as sector, the Bank's Regional staff, and governments to differenti- linkage to the Infrastructure vice presidency, avoidance of conflicts ate between the breach-of-contract coverage and the PRG. This of interest, and human resources and budget frameworks; how- confusion has inhibited their successful deployment. ever, no progress has been made on the integration since the working group submitted its report in 2005. Since then, the Proj- The Recommendation. The Project Finance and Guarantee Group ect Finance and Guarantee Group unit has been dissolved and its recommended that the management of the Bank's guarantee pro- function had been dispersed to the Bank's country and Regional gram be transferred to the MIGA executive vice president, who offices. Although this arrangement links the decision to offer the would act as an officer of the World Bank with respect to Bank PRG to clients with the policy dialogue, the uptake by the private guarantees; it also recommended that the group report directly sector has been slow since this function was dispersed. Possible to the executive vice president. It also concluded that the merger explanations include that decisions to use guarantees are normally would provide the right institutional set-up for the two units to work made at the investor's headquarters and not at their representa- closely together with the end goal of "positioning guarantees as tive offices and that Bank staff at the country or Regional offices the WBG's primary risk-mitigation instrument in mobilizing private do not have the necessary expertise to market the product. Source: World Bank, MIGA. Note: IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; MIGA = Multilateral Investment Guarantee Agency; PRG = Partial Risk Guarantee; WBG = World Bank Group. a. Such as Nam Theun 2 Hydroelectric Power Project in Lao PDR and Thailand; UMEME Power Distribution Project in Uganda; Bujagali Hydroelectric Project in Uganda; South Africa Regional Gas Project in Mozambique; and the West African Gas Pipeline Project in Ghana. competition coming from product innovation. If structure and modalities for the integration of PRGs continue to be booked on the Bank balance Bank PRG and MIGA under the leadership of sheet, then conflicts of interest would remain. MIGA's executive vice president. However, no progress has been made on the integration since An important issue would be how to preserve the working group submitted its report in 2005 the unique and distinctive features of Bank PRGs, (see box 3.1). which derive from Bank dialogues with the gov- ernment, under such a structure. Applying the in- A flexible and pragmatic approach may not ternal booking system of IFC may be an approach require organizational changes. It is impor- to consider in this regard. Given MIGA's mandate tant to note that organizational changes in them- to encourage FDI flows into developing coun- selves do not automatically produce desired tries, the changes to its convention could make outcomes. For instance, other multilateral devel- sense, but it will be a lengthy and complex process. opment banks have, as a rule, adopted country- integrated approaches to the delivery of their A version of the product-focused approach products; however, their operations differ signif- was attempted in 2005, when a working group icantly, including the relative shares of their pri- made recommendations about the organizational vate and public sector activities. 8 8 T H E D E L I V E RY O F P R M P R O D U C T S A C R O S S T H E W B G The literature on organizational change identifies tional structures have created new challenges incentives, culture, and "commitment at the top" and offered new opportunities to enhance WBG as more important factors for achieving strategic coordination in the delivery of guarantee products. objectives than simply redrawing organizational boundaries (Mintzberg 1994). If these ingredi- This evaluation found that the delivery of the ents are in place, desired results can often be WBG's guarantee instruments has several achieved without organizational changes. Many of strengths, including the following: (1) WBG guar- the benefits identified above can be achieved antee instruments are effective tools to promote within the current structure, provided incentives the WBG's development objectives; (2) they ex- and commitment at the top are in place. pand the range of instruments available to the WBG to best meet client needs; (3) each instru- For instance, the benefits of joint marketing to pri- ment can meet the demand for risk mitigation vate sector clients can be achieved among IFC, under different circumstances; (4) MIGA, a rela- MIGA, and Bank PRG by properly aligning incen- tively small institution of 120 people, has issued tives, including by following models currently $17 billion of guarantees that supported the flow used within IFC--such as shadow booking, dou- of $78 billion in foreign investment in developing ble booking, joint ventures, and so forth. Effi- countries and accounts for about 4 percent of the ciency gains can be achieved by harmonizing global market for PRI; (5) the Bank's PRG instru- among the three institutions the procedures and ment has supported some of the largest and most practices of working with private sector clients complex PPP infrastructure projects in high-risk along the lowest cost common denominator. developing countries that would have been un- MIGA's scope of action can be enlarged gradually, likely without the engagement of the Bank; and first by encouraging innovation within the scope (6) IFC's guarantee instruments have led its pen- of the current operational regulations (as rec- etration of the market for local currency finance, ommended by this evaluation) and over time by and its ability to take commercial risk has placed relaxing constraints in the authorizing environ- it in a unique position of meeting the demand for ment. Such a hybrid approach may provide a commercial and political risk mitigation. more flexible and pragmatic response to changes in internal and external environments. This evaluation therefore recommends that the WBG maintain its broad range of guarantee in- Conclusion struments, as they offer WBG clients a greater WBG guarantee instruments have been designed choice of products to meet their needs. as complementary products. In reality, however, there have been overlaps and competition within At the same time, a range of important the WBG in the delivery of PRM to private sector weaknesses in the WBG's delivery of guar- clients. Overlaps, despite complementarity by de- antee instruments is apparent. Although the sign, have been driven by innovation and flexibility overall demand for WBG guarantees is condi- in interpretation of policies. Competition for the tioned by varying market conditions, the WBG Bank and MIGA PRI has come from new and tra- needs to create an environment in which guar- ditional nonguarantee IFC products that incor- antee products are deployed in a flexible manner porate explicit and implicit political risk coverage. in response to evolving client needs. In this re- spect, the evaluation found a range of internal The WBG has not suffered from any paucity of co- weaknesses that effectively inhibit the deploy- ordination mechanisms in delivering guarantee ment of the WBG guarantee instruments: products, although these mechanisms have been only partially successful and significant gaps exist · Competition among institutions for the same with respect to coordination needs. Cooperation clients and of the kind that often imposes ad- in joint projects has brought benefits as well as ad- ditional transaction costs on clients and adds ditional transaction costs. Changes in organiza- reputation risk for the Bank 8 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 · Weaknesses in the marketing efforts for MIGA · Inconsistent pricing of the Bank PRG instru- and Bank products that limit client awareness ment, which runs the risk of WBG products and choice being differentiated based on price · A range of supply-driven policy and mandate re- · Weak links between advisory service activities strictions that inhibit the deployment of WBG on PPPs in infrastructure and deployment of guarantee instruments in specific situations guarantee instruments that do not take full ad- · Limited internal awareness, skills, or incen- vantage of the opportunities to systematically tives in the Bank and IFC to use guarantee in- introduce a range of WBG risk-mitigation prod- struments in relevant situations ucts earlier in a project's development. 9 0 Chapter 4 MIGA guaranteed the water treatment project in China. Photo by Roger Batstone. Recommendations WBG Management have also reflected artificial supply-driven To overcome the current limitations of the delivery product differentiation that is not client system of WBG guarantees and PRM instruments friendly. For example, restrictions such as and enhance its use and development potential, IFC's constraints of not offering partial IEG recommends the following to WBG senior risk and full credit guarantees and MIGA's management: equity link have been associated with loss of business opportunities or higher trans- 1. Take a strategic approach and make a action costs. A more rational approach decision on whether to maintain the ex- would be to reduce the mandate/policy isting organizational structure while ad- differentiations while emphasizing the dressing some of the important problems, distinct value the products add. Such an or develop and propose an alternative approach would, however, expand the op- organizational structure to the Board. portunities for overlaps and therefore re- quire enhanced coordination. Under any 2. Under any scenario, take action to in- scenario, establishing the mechanisms for troduce greater flexibility in the use of enhanced coordination should precede guarantee instruments in response to the greater flexibility in the product spaces dynamic country and client needs and of individual WBG institutions. market developments by taking the fol- · Ensuring that adequate incentives lowing actions: exist for staff to offer the full array of WBG guarantees and PRM products to · Revising existing policies and regu- private sector clients within a single lations on guarantees to minimize menu of options. The objective should supply-driven product restrictions be to enable the best product fit with the where most needed and to allow prod- particular circumstances of the client. Mar- uct differentiation on the basis of keting staff from each institution need to value added. Current policies and regu- be more familiar with the products of the lations contain supply-driven restrictions other institutions to provide clients with that do not eliminate overlaps yet tend to a full menu of options. IFC's extensive add to transaction costs and reduce the business development infrastructure sug- WBG's flexibility to respond to client gests an important IFC role in the mar- needs. Processes that add value with re- keting of WBG risk-mitigation products. spect to risk mitigation and safeguards It has the broadest relationship with pri- need to be maintained and strengthened. vate sector clients as well as commercial The hierarchy of instrument principle has banks--key drivers of PRI demand. IFC's often conflicted with the preferences of Client Relationship Management System market participants and has been a supply- might provide a channel for coordinating driven approach to the deployment of marketing efforts across the WBG for prod- instruments. Similarly, the product differ- ucts that reach the private sector. For such entiations based on eligibility and man- coordination to be effective, however, it date restrictions in the three institutions has to be based on clear management and 9 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 staff incentives in IFC, in particular, in- pricing might also reduce the need for centives to promote the products of the supply-side rationing of products through Bank and MIGA. the hierarchy of instruments principle and · Establishing more systematic links be- instead allow the most appropriate prod- tween advisory services and the de- uct to be deployed. MIGA has developed ployment of WBG PRM instruments a pricing system that reflects political risks and other products, particularly in in- as well as market conditions; this can serve frastructure, while keeping in mind as a basis for a more uniform pricing of po- the need to manage potential conflicts litical risk across the WBG. of interest.Greater upstream engagement · Strengthening internal awareness of of WBG PRM products might be enhanced the guarantee instruments and the with closer links to Bank/IFC advisory ser- incentives and skills for their use and vices that are engaged in helping govern- reducing transaction costs where pos- ments establish an appropriate regulatory sible, keeping in mind the importance framework and developing specific PPP of maintaining adequate processes transactions. IFC's corporate advisory and regulations for risk management. services, which are helping governments de- Lack of staff incentives, inadequate skills, sign concession agreements and privatiza- and poor familiarity with the products of tions, for example, represent a potentially the other institutions have prevented bet- important mechanism to more systemati- ter exploitation of synergies downstream cally introduce WBG PRM products earlier in marketing WBG products. Safeguards in a project's development, such as during and fiduciary requirements in each of the the bidding stage. IFC's corporate advisory three institutions generally add value and services could, for example, systematically need to be maintained and strengthened, include in bidding packages the option for especially in light of their importance for the winning bidder to choose from a menu mitigating financial, social, environmental, of WBG products such as IDA/IBRD PRG, and ultimately political risks. But where MIGA PRI, and IFC investments. Moreover, feasible, inefficiencies and unnecessarily enhancing client opportunities to choose cumbersome procedures need to be re- from a menu of WBG PRM products needs duced to minimize transaction costs for to be accompanied by a consistent ap- private sector clients. proach to pricing of political risk among the WBG institutions. 3. If a new organizational structure is de- · Following a consistent approach to veloped and proposed, consider at least pricing PRM across its guarantee in- three alternative perspectives for orga- struments to avoid potential distor- nizational realignment (client, country, tions. At present, the Bank has a distinct and product). pricing structure that does not reflect risk or market conditions. The Bank's Oper- · Under the client approach, all products for ating Principle 8.30 requires that in finan- private sector clients, including guaran- cial intermediary lending, on-lending occur tees and PRM instruments, would be of- at or near market rates. The same logic fered through a single window. should be extended to the pricing of IDA · Under the country approach, the deploy- and IBRD PRGs, which also reach private ment of WBG guarantee and PRM products sector clients. A consistent approach to would be made according to country the pricing of comparable risks would re- needs, under a management arrangement duce the potential for distortions and en- common for all three institutions. able WBG private sector clients to make · Under the product approach, the bulk of rational decisions as to which WBG in- guarantee/insurance products would be strument best fits their needs. Consistent managed under one institutional roof. 9 4 R E C O M M E N D AT I O N S · Each of the organizational realignments nesses that reduce efficiency and slow re- may bring higher benefits than the exist- sponsiveness without lowering MIGA's ing structure in terms of lower transac- financial, social, and environmental stan- tion costs, better client focus, and higher dards. These include organizational issues in efficiency, but each also entails significant staffing, performance review, and incentives as uncertainties and risks in terms of loss or well as consideration of matters such as in- marginalization of some products, con- flexibility on guarantee contract terms and flicts of interest, and reduced flexibility to conditions. However, efficiency in its under- respond to changes in the external envi- writing process must not come at the expense ronment. The assessment of alternatives of quality, risk mitigation, safeguards, and de- for organizational realignment should take velopment impacts of the projects it insures. into account efficiency aspects, conflict-of- · Improving its client relationship man- interest implications, effective coordination agement, including aftercare, to enhance within the WBG, and responsiveness to pri- MIGA's value added and increase client vate sector clients' and countries' needs. retention. Improving client relationships could be the most cost-effective marketing 4. If the current organizational structure that MIGA could undertake to increase its is maintained, direct management of business volume. Managing client relation- each individual WBG institution to im- ships requires a focused and coherent business prove the delivery of its own guarantee/ development plan implemented by a staff insurance products. with expertise in the guarantee business and/or financial markets. MIGA Management · Proposing to MIGA's shareholders amend- Bank Management ments to its Convention to remain rel- · Maintaining and promoting the PCG in- evant and meet its market potential. strument as a countercyclical tool to lever- Constraints imposed on MIGA by its 1985 age government access to commercial Convention need to be reconsidered so that funds and extending such access to IDA MIGA can better serve its developing country countries. PCGs remain a potentially impor- members and clients. The most notable con- tant instrument to help well-performing coun- straint is the inability to insure stand-alone tries that do not have full access to commercial debt (with no equity participation), and ex- markets by introducing them to markets and isting assets, and local and foreign investors improving the terms of initial transactions. among others. Although amending the Con- They can also provide countercyclical assis- vention is likely to be a lengthy process in- tance to countries whose access to markets is volving shareholder approval and ratification temporarily restricted. At the same time, PCGs from members, it would be important for can substitute for direct Bank lending and can MIGA to take the necessary steps now to re- reduce the Bank's leverage in some situations. tain its relevance in the future. More discussion and clearer guidance on the · Considering, in the meantime, alleviat- merits of PCGs vis-à-vis direct lending in vari- ing several constraints derived from its ous circumstances should be developed in the operational regulations and policies. Sev- Bank. Given the growing number of IDA coun- eral changes to MIGA's Operational Regula- tries that fit the profile of well-performing tions to add new eligible investments and countries with restricted access to markets, risks can help MIGA develop new products the option of a PCG instrument should be ex- and meet evolving market demands. Several tended to IDA countries. Current incentives in changes in policy and MIGA's Operational the Bank promote direct Bank lending rather Regulations are being discussed internally. than leveraging of private commercial finance, · Increasing its responsiveness to market yet PCGs can have significant development demand by addressing internal weak- impact and additionality in the right circum- 9 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 stances. They should be more thoroughly con- · Assessing the extent to which it can sidered in the CAS process. bring its guarantee products closer to · Taking several measures to enhance the meeting Basel II--and regulatory re- use of Bank PRGs. Although the market for quirements in general--so that the guar- Bank PRGs will remain narrow because of the antee beneficiaries can use IFC products nature of the product, PRGs remain the only more effectively for capital, provisioning, instrument the Bank has to directly support pri- and exposure relief. A great deal of the de- vate investment projects, and Bank lending mand for guarantee-type instruments derives cannot substitute for them. They provide a from financial institutions' needs for relief unique means of PRM because of the Bank's on capital, provisioning, and exposure re- close relationship with governments, sector quirements. In this context, IFC should review knowledge, and policy dialogue on private its guarantee products to assess the extent to sector development, and they can be deployed which they can be tailored to better meet to enable transactions that would otherwise be Basel II and regulatory requirements for the perceived as too risky. Important measures above purposes. for enhancing their use are creating aware- · Revisiting its approach to RSFs to in- ness among Bank staff of the potential use crease flexibility and improve the at- and benefits of PRGs and building necessary tractiveness of the product. The rigid skills; developing a marketing strategy that approach to structuring RSFs has limited IFC's encompasses both governments and the pri- ability to fully exploit the benefits of partner- vate sector to better identify situations in which ing with local and international financial insti- the role of a PRG can make a difference; and tutions. This rigidity and high transaction costs streamlining processing steps to reduce both have discouraged utilization of the instrument. internal disincentives to working on PRGs and More flexible structures should be considered transaction costs for private sector clients while to make the product more attractive to part- ensuring that crucial measures for social and ner financial institutions. IFC has accumulated environmental safeguards and risk manage- the data and the experience to give it the com- ment are maintained and strengthened. fort needed to simplify processes and to give flexibility to partners to use their strengths IFC Management while strengthening those processes intended · Mainstreaming its guarantee products for risk management and social and environ- through its operations departments in mental safeguards. the same manner that its equity and loan · Scaling up successful models in energy ef- products are deployed. IFC's Treasury has ficiency, education, and capital market de- been instrumental in innovating and spread- velopment based on the use of guarantee ing innovation. The Treasury is decentralizing, structures. IFC has developed models based beginning to put senior staff in the field. How- on guarantee structures in the areas of energy ever, mainstreaming needs to be led by the in- efficiency, SME financing, education, and cap- dustry and regional departments, which have ital market development. Limited replication the incentives and resources to scale up the use has taken place so far. IFC needs to assess its of the instruments in response to evolving experience with these products, simplify, stan- client needs. IFC needs to develop the skills dardize, and bring them closer to market prac- and capacity in its operational departments tices to enhance prospects for scaling up in line to offer a broad range of guarantee products. with its programmatic approach. 9 6 Appendixes Complex cross-border projects represent high-risk endeavors that can benefit from WBG guarantees. Photo courtesy of Sasol, a gas pipeline project guaranteed by MIGA and the World Bank. APPENDIX A: EVALUATION METHODOLOGY The evaluation covers World Bank Group (WBG) cluded project information on all guarantee op- guarantee operations from 1990 to 2007. The erations conducted in the WBG since 1990. These evaluation used the methods outlined here to data were then matched against a range of crite- gather evidence and compile the evaluation. ria, including the WBG's lending eligibility criteria, host countries' income level classification, Insti- Literature and Project tutional Investor Country Credit Rating country risk Documentation Review classifications, project size, project mobilization The Independent Evaluation Group (IEG) un- data, processing times and costs, and so forth. dertook a broad review of both internal and ex- ternal literature on guarantees and political risk Data on normal IFC lending operations and Bank mitigation instruments. The IEG team reviewed infrastructure lending were also gathered to en- a number of documents: able some comparison between the instruments. Data on joint projects were collected to assess co- · All WBG policy documents related to the use operation among the three institutions. IEG also of guarantee instruments since the late 1980s reviewed all IFC investment operations during a · WBG analytical reports on guarantee instruments set sample period (fiscal 2005­07) to extract those · Relevant country and sector strategy documents in which PRM was identified as a primary ration- · Guarantee project approval, supervision, and ale for IFC engagement in the project. The review completion/evaluation reports also examined the eligibility criteria, nature of · Administrative guidelines from the World Bank, coverage extended, types of clients, and pricing the International Finance Corporation (IFC), structure used in political mitigation instruments and the Multilateral Investment Guarantee across the WBG. Agency (MIGA) that address the conditions and procedures relevant to the use of guarantees. The evaluation also drew on external data sources. External data included LoanWare, mar- The external literature included major publications ket reports, and Berne Union data. Using data on political risk mitigation (PRM) and credit en- from the LoanWare database, an analysis of the hancement; journal articles; and research con- market for loan guarantees was conducted that ducted by private sector practitioners, academics, covered the period of 1993­2007 (during which and other multilateral and bilateral development period there are 113,661 deals recorded, of which agencies. approximately 8 percent were guaranteed.) A background report with this analysis is available Data Collection and Processing on request. The IEG team also extracted considerable internal data on the use of guarantee instruments from in- Staff Interviews stitutional databases such as IFC online databases, The IEG team interviewed some 50 staff from including the Management Information System MIGA, IFC, and the Bank who were experienced and the Resource Management System. These in- in the use of WBG guarantee instruments. 9 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 IEG conducted a survey of WBG staff who have credit agencies. Among the multilaterals reviewed worked with guarantee instruments. The objec- were the African Development Bank, European tive of the survey was to solicit staff views about Investment Bank/Enhanced Integrated Frame- the use and effectiveness of guarantee instru- work, the European Bank for Reconstruction ments. The survey questionnaire, which was e-mail and Development, the Asian Development Bank, based, was sent to 363 preselected staff on the the Nordic Investment Bank, and in general basis of their current or previous experience with terms Islamic Development Bank/Islamic Cor- guarantees. Of those, 206 staff responded to the poration for the Insurance of Investment and survey; responses included a range of comments Export Credit, Ceylon Electricity Board, and Afri- that were used in the evaluation. The breakdown can Trade Insurance Agency/ACA. Export credit of survey response rates as well as an overview of agencies reviewed included Finnvera, Office key results is presented in appendix B. Nationale du Ducroire (Belgian Export Credit Agency)/Delcredere, and the Commonwealth External Interviews Development Corporation. In addition, some IEG interviewed several external stakeholders information was gathered on guarantee activities located in Washington, DC, London, Paris, Brus- of the Groupe Agence Francaise de Developpe- sels, and Luxembourg. These included private ment and its private sector arm, Proparco. A companies and commercial banks that are clients background report compiled during this review of WBG guarantee instruments, the Berne Union is available on request. (a consortium of other providers of political risk insurance), brokers of political risk insurance, External Reviews other multilateral providers of guarantees, and na- The report benefited from external reviews. Three tional export credit agencies engaged in the pro- external experts provided comments and feedback vision of guarantees. on the evaluation: James Hanson, a former WBG staff and financial sector expert; Professor Marshall Review of Other Multilateral/ Meyer of the Wharton School of Business, who is National Providers of Guarantees an expert in organizational design; and Bob Chest- IEG undertook a desk review of the organi- nutt, a practitioner with long experience in de- zational structure and guarantee operations of veloping infrastructure projects in Africa, including several other providers of guarantees. These in- with the participation of the World Bank, IFC, cluded both multilateral providers and export and MIGA. 1 0 0 APPENDIX B: WBG GUARANTEES STAFF SURVEY As part of the evaluation of WBG guarantee in- two-thirds of IFC staff were very familiar with many struments, IEG conducted a survey of WBG expert of their diversified products--PCGs, Risk-Sharing staff between February 6 and 25, 2008 (see at- Facilities (RSFs), PRGs, and the Global Trade Fi- tachment on page 106). The objective of the sur- nance Program (GTFP), but less than half knew vey was to solicit views about the use and about MIGA PRI. Only one-fifth of IFC staff were effectiveness of guarantee instruments. The sur- familiar with IBRD/IDA products. Compared with vey questionnaire, which was e-mail based, was the Bank and IFC staff, MIGA staff were more fa- sent to 363 preselected staff on the basis of their miliar with the other two institutions' products, es- current or previous experience with guarantees. pecially with Bank guarantees. However, less than The total number of responding staff was 206. The half of MIGA staff reported being familiar with IFC breakdown of staff and their respective response guarantees. As is seen above, Bank and MIGA staff rate is shown in table B.1. are familiar with products of the other institution, but not with those of IFC. In contrast, IFC staff are not familiar with products of either the Bank or Table B.1: Staff Survey Responses MIGA. Total number Response According to the WBG staff, the most critical ben- WBG of staff rate (%) efits of the WBG guarantee instruments were World Bank 61 45.9 enhanced image of financial soundness and IFC 243 52.6 improved financing terms (rates and tenors). More than 85 percent of WBG staff felt that this MIGA 59 84.8 was the case. Total 363 56.4 Source: IEG survey. Note: IFC = International Finance Corporation; MIGA = Multilateral Invest- In addition to these two common benefits, staff also ment Guarantee Agency; WBG = World Bank Group. pointed out several other benefits. For IBRD/IDA PRGs and PCGs, most Bank staff reported that the WBG staff are familiar with their own products WBG's role as an honest broker and IBRD/IDA's as- but not with the guarantee products of other WBG sistance in securing other investors and structur- institutions (see table B.2). Most Bank staff were ing finance were other critical benefits. familiar with their own instruments as Interna- tional Bank for Reconstruction and Development For IFC PRGs and PCGs, about 75 percent of IFC (IBRD)/International Development Association staff reported that IFC's technical and economic (IDA) Partial Risk Guarantees (PRGs), IBRD Partial appraisal of the project and assistance in securing Credit Guarantees (PCGs), and Policy-Based Guar- other investors and structuring finance were also antees. Bank staff were also familiar with MIGA po- important benefits. In MIGA, more than 70 per- litical risk insurance (PRI) as much as their own cent of staff felt that the WBG's role as an honest products; however, except for IFC PCGs (a little less broker and MIGA's assistance in securing other in- than half), less than one-third of staff were famil- vestors and structuring finance were additional iar with IFC guarantees. As reported, more than benefits of their PRI product. Ability to provide 1 0 1 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Table B.2: Staff Familiarity with WBG Guarantee Instruments WorldBank(%) IFC(%) MIGA(%) Familiarity with the guarantee IBRD/IDA PRG (96.4) IFC PCG (87.4) PRI (100) product of own institution IBRD PCG (85.7) IFC RSF (77.2) IBRD PBG (65.4) IFC PRG (67.8) IFC GTFP (64.8) IFC CLG (48.3) IFC GOLF (28.8) Familiarity with the guarantee MIGA PRI (85.2) MIGA PRI (45.8) IBRD/IDA PRG (83) product of another WBG institution IFC PCG (48) IBRD/IDA PRG (22.7) IBRDPCG(58.7) IFC PRG (28) IBRD PCG (12.7) IFCPCG(47.8) IFC GTFP (20) IBRD PBG (7.1) IFCPRG(43.8) IFC RSF (16) IBRDPBG(41.3) IFC CLG (15.4) IFC GOLF (37) IFC GOLF (11.5) IFCRSF(28.9) IFCGTFP(26.1) IFCCLG(22.7) Source: IEG survey. Note: CLG = credit-linked guarantee; GOLF = Global Offshore Liquidity Facility; GTFP = Global Trade Finance Program; IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; MIGA = Multilateral Investment Guarantee Agency; PBG = Policy-Based Guarantee; PCG = Partial Credit Guarantee; PRG = Partial Risk Guarantee; PRI = political risk insurance; RSF = Risk-Sharing Facility; WBG = World Bank Group. assistance in securing other investors and struc- by IBRD/IDA lending, one-fourth by MIGA's PRI, turing finance were seen as additional critical ben- and one-fifth by IBRD PCGs (table B.3). efits of WBG guarantee instruments by a majority · IFC. In IFC, staff suggested substitutability of of staff in all three institutions. several instruments. About 15 percent of staff familiar with this product felt that it had no Staff had varied views on which products could substitute; more than 85 percent reported be substituted by another. The survey results in- that it was substitutable. From those, about 40 dicate that there is no clear consensus on which percent felt that it can be substituted by IFC's products can be substituted by the others across direct investment and about one-fourth by the institutions. RSFs. Though not significant, there were also views suggesting substitutability of IFC PCGs · World Bank. In the Bank, a significant number by IBRD PCGs, IFC CLGs, PRGs, Global Of- of opinions on substitutes were collected only fensive Liquidity Facility, and GTFP, and IBRD/ for the PRG instrument. According to WBG staff IDA lending. As for RSFs, though about 20 who are most familiar with IBRD/IDA PRGs, this percent of staff reported that the product has product can have as many as seven different sub- no substitute, another 20 percent reported stitutes within and outside its originating insti- that it can be substituted by IFC's direct in- tution. Although about 30 percent of WBG staff vestment, and about one-third suggested its responded that the product has no substitute, substitutability by IFC PCGs and PRGs. An in- about 70 percent suggested substitutes. Here significant proportion of staff suggested sub- one-third reported that it can be substituted stitutability of RSFs by GTFP. As for IFC PRGs, 1 0 2 A P P E N D I X B : W B G G U A R A N T E E S S TA F F S U RV E Y Table B.3: Suggested Substitutes of WBG Guarantee Instruments IBRD/IDA IFC IFC IFC IFC MIGA Suggested substitute PRGs (%) PRGs (%) RSF (%) PRG (%) GTFP (%) PRI (%) IBRD/IDA lending 37.5 6.7 MIGA PRI 25 12.5 IBRD PCGs 18.8 4.5 IBRD/IDA PRGs 12.5 6.7 26.0 IBRD PBGs 4.0 IFC PCGs 6.3 33.3 12.5 6.0 IFC RSFs 12.5 22.7 25.0 6.7 4.0 IFC PRGs 12.5 15.9 28.6 12.5 6.7 14.0 IFC direct investment 12.5 36.4 19.0 20.0 4.0 IFC credit-linked guarantees 18.2 25.0 2.0 IFC GOLF 6.8 IFC GTFC 4.8 None of the above 18.8 11.4 19.0 0 53.3 44.0 Source: IEG survey. Note: GOLF = Guaranteed Offshore Liquidity Facility; GTFC = Global Trade Facility Program; IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency; PBG = Policy-Based Guarantee; PCG = Partial Credit Guarantee; PRG = Partial Risk Guarantee; PRI = political risk insurance; RSF = Risk-Sharing Facility; WBG = World Bank Group. about two-thirds of staff reported their was insignificant. In terms of GTFP, about half substitutability by IFC's direct investment. Al- of staff most familiar with this instrument sup- though there were views supporting their ported its nonsubstitutability, whereas substitutability by RSF, CLG, IFC PCG, IBRD/ an insignificant proportion of staff suggested IDA and IFC PRGs, and MIGA PRI, the rsponse that this product could be substituted by Table B.4: Percent of Surveyed Staff Who Reported That the Change Is Important Important changes IBRD/IDA (%) IFC (%) MIGA (%) Improving coordination with other WBG institutions 77.8 49.1 89.1 Improving marketing of guarantees 96.3 81.3 97.9 Clarifying policies and guidelines, explaining when guarantees are appropriate 82.1 82.3 80.9 Offering more staff training 77.8 90.3 71.1 Reducing time and cost to process guarantees 89.3 90.1 89.4 Offering more flexible contract terms 77.8 83.2 89.4 Investing in new product development 77.8 83.0 91.3 Source: IEG survey. Note: IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency; WGB = World Bank Group. 1 0 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Table B.5: Reasons for Droppages of Guarantee Projects Droppage reason IBRD/IDA (%) IFC (%) MIGA (%) Inadequate compliance with environmental or social guidelines 16.7 1.9 5.0 Another WBG agency provided the guarantee 0.0 1.9 10.0 Intermediate commercial banks withdrew from project 11.1 20.8 10.0 Underlying project technically or financially unsound 16.7 15.1 15.0 Government objected to the project 0.0 7.5 17.5 Another multilateral or bilateral agency provided the guarantee 27.8 13.2 20.0 A private firm provided the guarantee 16.7 9.4 27.5 Other 33.3 18.9 35.0 Processing time too long for client 44.4 43.4 40.0 Client proceeded with the project but without any guarantee involved 66.7 41.5 47.5 Cost of guarantee was too high for client 11.1 81.1 50.0 Client dropped the underlying project 22.2 28.3 55.0 Source: IEG survey. Note: IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency; WBG = World Bank Group. IBRD/IDA PRGs and lending and IFC's PRGs, and offering more training to staff on guaran- RSFs, and direct investment. tees were also strongly supported across institu- · MIGA. As for MIGA's PRI, about 40 percent of tions. In addition, MIGA and IBRD/IDA staff staff supported the product's nonsubsti- stressed the importance of improving the coor- tutability. From those staff that felt that it had dination within WBG institutions. Overall results substitutes, one-fourth felt that it can be sub- suggest strong support for these changes in all stituted by IBRD/IDA PRGs, and one-fifth felt three institutions. that it is substitutable by IFC PRGs and the Global Offshore Liquidity Facility. Though there According to WBG staff with experience, clients were other views supporting substitutability proceeding with the project without a guaran- of PRI by IBRD policy-based guarantees, IFC tee and long processing times were the main PCGs, RSFs, CLGs, and direct investment, the reasons for dropped guarantee projects. About significance was low. 65 percent of IBRD/IDA staff, 50 percent of IFC staff, and more than 80 percent of MIGA staff re- Changes Needed to Improve Instruments ported having experience with dropped guar- A high proportion of staff felt that changes are antee projects (table B.5). needed to improve the WBG's guarantees in- struments (table B.4). Overall, most staff felt According to about one-third of IBRD/IDA staff, that reducing time and cost of processing guar- an involvement of another bilateral or multilat- antees and improving marketing are important for eral agency in providing the guarantee was an- improving WBG guarantee instruments. Whereas other decisive factor for the guarantee. About these changes were supported by about 90 per- 40 percent of IFC staff took views in support of cent of overall surveyed staff, investing in new the two reasons mentioned, but about 80 percent product development, offering more flexible con- reported that the droppages occurred because tract terms, clarifying WBG policies and guide- the cost of the guarantee was too high for the lines to explain when guarantees are appropriate, client. Moreover, clients dropping the underlying 1 0 4 A P P E N D I X B : W B G G U A R A N T E E S S TA F F S U RV E Y Figure B.1: Originator of Guarantees Project sponsor Private commercial bank Marketing staff in your institution Other staff in your institution Staff in another Bank institution Host government 0.0 25.0 50.0 75.0 100.0 Percent MIGA IBRD/IDA IFC Source: IEG survey. Note: IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency. project and withdrawal of intermediate com- marketing staff, and private commercial banks mercial banks from the project were reported by originated guarantees. Compared with products IFC staff as contributing reasons in one-fifth of of other institutions, Bank guarantees were also dropped guarantee cases. relatively frequently originated by host govern- ment and staff of another WBG institution. In MIGA more than 40 percent of staff shared views in support of the two common reasons, In IFC, as reported by about 80 percent and 60 per- whereas 50 percent pointed out a too-high cost cent of staff, respectively, its marketing staff and for the client and the client's droppage of the other staff play an important role in originating underlying project as reasons. About 20 percent guarantees. Private commercial banks and proj- of MIGA staff reported that the involvement of a ect sponsors also approach IFC for a guarantee. private firm and a multilateral or bilateral agency As IFC staff reported, host government and staff in provision of guarantees was another reason in another Bank institution are the ones that are for dropped guarantees. least likely to originate its guarantees. Bank and MIGA staff reported that project spon- In MIGA, as reported by about 90 percent of sors/investors most frequently originated the re- staff, project sponsors and private commercial quest of guarantees (figure B.1). In contrast, in banks first approach MIGA for a guarantee. Ac- IFC, marketing staff were the ones to most fre- cording to staff, its marketing staff and other quently originate a guarantee. According to Bank staff also play an important role in originating staff, mostly project sponsors, its other staff, its guarantees. 1 0 5 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 Attachment: Survey Structure and Questions The survey questions were structured to provide multiple choices as well as open-ended answers. All staff were asked the following questions: SURVEY 1. How familiar are you with the following WBG IBRD Partial Credit Guarantees (PCG) guarantee instruments? (Check all that apply.) IBRD/IDA Partial Risk Guarantee (PRG) Answer Options IFC Partial Credit Guarantee (PCG) IFC Risk Sharing Facilities (RSF) IBRD Policy-Based Guarantees (PBG) IFC Partial Risk Guarantees (PRG) IBRD Partial Credit Guarantees (PCG) IFC Credit Linked Guarantees (CLG) IBRD/IDA Partial Risk Guarantee (PRG) IFC Guaranteed Offshore Liquidity Facility (GOLF) IFC Partial Credit Guarantee (PCG) IFC Global Trade Facility Program (GTFP) IFC Risk Sharing Facilities (RSF) MIGA Political Risk Insurance (PRI) IFC Partial Risk Guarantees (PRG) IBRD/IDA lending IFC Credit Linked Guarantees (CLG) IFC direct investment IFC Guaranteed Offshore Liquidity Facility (GOLF) None of the above IFC Global Trade Facility Program (GTFP) MIGA Political Risk Insurance (PRI) 5. How might the delivery of this instrument be Specify the level of familiarity: Not familiar / Barely improved? familiar / Somewhat familiar / Very familiar Answer Options: Open 2. Select the WBG guarantee instrument that you are most familiar with. 6. How important are the following changes for Answer Options improving your institution's guarantee operations? IBRD Policy-Based Guarantees (PBG) Answer Options IBRD Partial Credit Guarantees (PCG) Improving its coordination with other WBG IBRD/IDA Partial Risk Guarantee (PRG) institutions IFC Partial Credit Guarantee (PCG) Improving its marketing of guarantees IFC Risk Sharing Facilities (RSF) Clarifying its policies and guidelines, explaining when IFC Partial Risk Guarantees (PRG) guarantees are appropriate IFC Credit Linked Guarantees (CLG) Offering more training to staff on guarantees IFC Guaranteed Offshore Liquidity Facility (GOLF) Reducing the time and cost to process its guarantees IFC Global Trade Facility Program (GTFP) Offering more flexible contract terms MIGA Political Risk Insurance (PRI) Investing in new product development Specify the level: Extremely important / Somewhat 3. In your experience, how critical are the following important / Not very important / Not at all important / benefits to clients for this guarantee instrument? No opinion Answer Options WBG role as honest broker 7. Have you worked on a guarantee project that was Enhanced image of financial soundness dropped before becoming effective? Compliance with environmental and social standards Improved financing terms (rates and tenors) 8. If you had a project dropped, identify which were Your institution's technical and economic appraisal of the most likely reasons (select up to 5). the project Cost of guarantee was too high for client Your institution's assistance in securing other Client proceeded with project but without any investors and structuring finance guarantee involved Specify the level: Extremely critical / Somewhat Client dropped the underlying project critical / Not very critical / Not at all critical / No Processing time was too long for client opinion Intermediate commercial banks withdrew from project Government objected to the project 4. What other WBG instruments can substitute for Underlying project technically or financially unsound this guarantee instrument? Inadequate compliance with environmental or social Answer Options guidelines IBRD Policy-Based Guarantees (PBG) Another WBG agency provided the guarantee. 1 0 6 A P P E N D I X B : W B G G U A R A N T E E S S TA F F S U RV E Y Another multilateral or bilateral agency provided the Other staff in your institution guarantee. Staff in another World Bank institution A private firm provided the guarantee. Specify frequency level: Frequently / Occasionally / Other reasons (please specify) Infrequently / Never / No opinion 9. Who typically first suggests that your institution's 10. What immediate change would you make to guarantees might be appropriate instruments for a improve the WBG's guarantee program? project? Answer Options: Open Answer Options Host government 11. What risk mitigation needs of clients are not met Project sponsor by your institution's guarantee instruments? Private commercial bank Answer Options: Open Marketing staff in your institution 1 0 7 Factory worker in Indian plant. Photo by Ray Witlin, courtesy of the World Bank Photo Library. ENDNOTES Executive Summary customers, debt holders, and shareholders (Merton 1. This study does not evaluate IFC's Global Trade and Perold 1998). Finance Program, which started in 2005. 5. Referred to as "obsolescing bargain," which is de- fined as the propensity of host country authorities, Chapter 1 mostly successors to signatories to the original in- 1. As pointed out by Berger and Udell (1988), guar- vestment agreements, to tighten the terms and con- antees typically operate like external collateral, but ditions of investment contracts that were originally they do not give control over specific assets. Instead, drawn to reward high early risk and uncertainty, after they represent a generic claim on the entire wealth of risk decreased and the project proved successful. the guarantor, who thus has a large degree of free- 6. MIGA Convention, chapter X, Article 59: A vote dom in using--and possibly neglecting--it. of three-fifths of the governors exercising four-fifths of 2. The New Oxford American Dictionary defines the total voting power is required for amendments in insurance as "a practice or arrangement by which a com- the Convention and its annexes. pany . . . provides a guarantee of compensation for spec- 7. Until fiscal year 1988, IFC issued three equity- ified loss . . . in return for payment of a premium" and, related guarantees: a guarantee that insures a minimum even more simply, as "a thing providing protection return on notes issued in local stock markets, a guar- against a possible eventuality." antee insuring repatriation of equity principal, and a 3. In determining which financial risk transfer services Guaranteed Recovery of Investment Principal (GRIP). are insurance, five characteristics are typically identified: IFC's GRIP program was designed to encourage pri- (1) the insured must have an "insurable risk" (such as vate investors to participate in projects financed by the risk of a financial loss in the case of a disaster, theft, IFC, even though the risk was considered quite high or credit event) with respect to a "fortuitous event" by the private investor. The program gave the inves- (defined as "any occurrence or failure to occur which tor a number of alternatives in participating with is, or is assumed by the parties to be, to a substantial ex- IFC. The private investor considering an equity in- tent beyond the control of either party") that is capa- vestment in an IFC project gave the funds to the IFC ble of financial evaluation; (2) the insured must transfer and received a dollar-denominated certificate (GRIP), its risk of loss to an insurance company under a contract which IFC had to repay in some stipulated period-- that provides the insured with indemnity against the loss; for example, 20 years. IFC then used the funds to (3) the insured must pay a premium to the insurance make the equity investment in its own name. When the company for assuming the insured's insurable risk; (4) debt certificate or GRIP matured, the investor had the insurance company typically assumes the risk as part the options of getting the funds returned with some of a larger program for managing loss by holding a large profit included, buying the shares by cancelling the debt pool of contracts covering similar risks that is large and paying a prearranged premium to the IFC, or enough for actual losses to fall within expected statisti- extending the GRIP for an additional period. IFC cal benchmarks; and (5) before it can collect on an in- stopped using equity-related guarantees in the late surance contract, the insured must demonstrate that its 1980s, in part not to compete with MIGA, which was injury was from an "insurable risk" as the result of an "in- established in 1988 with a special focus on promoting sured event." In other words, the insured must demon- FDI through PRI. strate that it has actually suffered a loss that was covered 8. GOLF is not a self-standing guarantee product and in the contract (Culp 2003). has been used only once. 4. Finally, it is important to note that besides third- 9. Article 14 of the MIGA Convention limits its guar- party guarantors, other potential issuers of asset in- antee operations to investments made in the developing surance to the firm are the firm's stakeholders, including member country. 1 0 9 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 10. Article 12, §§ (a)­(c) of the MIGA Convention positioned to assess and price such risks because of its and MIGA's Operational Regulations set out the eligi- understanding of the players and/or its ability to divest ble investments that qualify for a MIGA guarantee. the risk. (2) Efficiency Reasons: All-risk coverage There was a deliberate effort in developing MIGA's helps avoid the ambiguity inherent in most investment Convention to avoid enumeration of eligible invest- situations, because different types of risks are often in- ments with an exclusive list of types of investments cov- terrelated and difficult to disentangle. Also, all-risk cov- ered, in order to provide flexibility to MIGA's Board of erage is more cost-effective because appraisal and Directors. MIGA's founding members recognized that supervision costs do not vary significantly with the the success of the guarantee program hinges on its abil- number of risks assessed. (3) Preferred Creditor ity to adapt to innovations in the marketplace (Shi- Status and MIGA: Providing PRGs may jeopardize hata 1988, pp. 111­12). IFC's preferred creditor status and impinge on MIGA's 11. Includes portfolio investments, which could be role. These are sound principles, but their validity can minority participations in joint ventures, preferred be limited to certain circumstances. For instance, risk stock, and shares resulting from the conversion of debt sharing in terms of IFC guaranteeing certain types of instruments (per MIGA's Operational Regulations). risks and not others is perfectly in line with the Busi- 12. Includes production and profit-sharing con- ness Principle. In any event, IFC never takes exactly the tracts, management contracts, franchising and licens- same risks as the sponsor, and through the structuring ing agreements, turnkey contracts, operating leasing of transaction risks and rewards are always apportioned agreements and subordinated debentures issued by the in various ways among deal participants. The efficiency project enterprise, and guarantees or other securities reasons can also be subject to limited validity. Certain provided for loans to the project enterprise. Coverage kinds of risks often can be clearly isolated from others. must have terms of at least three years and depend sub- Giving clients the option to trade certain types of risks stantially on the production, revenues, or profits of should enhance efficiency. Assessing one or a few spe- the investment project for repayment (per MIGA Op- cific risks may also require fewer resources than a full erational Regulations, §§ 1.05 and 1.06). appraisal. 13. Per MIGA Operational Regulations, §1.08. Refer 17. For instance, when a full credit guarantee is also to the MIGA Convention, "Commentary on the cost-effective, IFC can perform its due diligence and su- Convention Establishing the Multilateral Investment pervision functions in a cost-effective way, and IFC is Guarantee Agency." These provisions give the MIGA unlikely to fund in the same instrument, currency, and Board the flexibility to extend MIGA coverage to market, particularly in the local markets. medium- or long-term investments, except for loans that are unrelated to a specific investment covered or to be Chapter 2 covered by MIGA. 1. Note that the size of the total investment in the 14. Loans and guarantees of less than three years may project can be above the limit, as long as the MIGA cov- be eligible for cover if the Board so approves, pro- erage does not exceed this ceiling. vided that the investor demonstrates a long-term com- 2. IDA provided a $5 million credit and the ADB a mitment to the project. $5 million concessional loan. The U.K. government 15. Value must be determined in terms of the cur- support materialized as a $1 million grant from the rency in which the guarantee is to be issued (Shihata Department for International Development. 1988, pp. 116­17). 3. Sixty-five of these claims were due to expropria- 16.Three reasons are given for IFC's focus on full- tions, with a total payout of $600 million (OPIC 2007). risk guarantees: (1) Business Principle: Because IFC 4. According to the Berne Union Investment In- acts as a full-risk partner in developing country in- surance Database, total claims outstanding amounted vestments, its guarantee activities cover all types of to $337.1 million, $159.8 million, and $150.2 million for macroeconomic, commercial, and political risks where calendar years 2005, 2006, and 2007, respectively. such guarantees are critical for the provision of addi- 5. Cancellations of contracts are highest for the tional funding from other sources and where IFC is best agribusiness, manufacturing, tourism, and services sec- 1 1 0 E N D N O T E S tors, followed by financial sector projects. Cancella- 14. Interview of underwriters and risk management tions are below average for infrastructure projects, officers for the IEG­MIGA fiscal 2007 Annual Report. which usually have a longer gestation period, as well 15. The reasonable period proviso of MIGA's Oper- as for mining and oil and gas projects. ational Regulations states that MIGA may deem the 6. The mediation efforts were intended to resolve out- host country approval as given if the host country pre- standing claims from the Menghistu government. In fis- sents no objection within a reasonable period, which cal 2004, about 40 claims had moved toward resolution, "shall in no case be less than 30 days from the date of with concrete settlement offers from the government. the request for approval and shall be extended at the 7. In 2007 the Board approved the increase in MIGA's request of the host country." country limits from $420 million to $600 million and the 16. All members conduct an analysis of the investors' individual project limit from $110 million to $180 million. (and, in a few cases, the local partners') bonafides. A 8. The United National Conference on Trade and De- few insurers require that the investor have a minimum velopment reports an increasing internationalization of number of years of experience in the sector where the research and development and a growing share of FDI investment is being made. Some insurers also require based on the transfer of intellectual property rights submission of financial statements after the contract of (UNCTAD 2005). guarantee is issued. 9. The MIGA Convention, Article 13(c), states that 17. MIGA's Council of Governors and Board of Di- "the assets invested are transferred from outside the rectors set the maximum amount of contingent liabil- host country." The investor and the host country must ity that may be assumed by MIGA as 350 percent of the jointly apply for Board approval. sum of its unimpaired subscribed capital and reserves 10. See the MIGA Convention, Article 11(b), and and retained earnings, 90 percent of reinsurance ob- MIGA Operational Regulations, paragraphs 1.53­1.57. tained by MIGA with private insurers, and 100 percent 11. In 2007 Zurich Financial Services Group, a pri- of reinsurance from public insurers. MIGA's maximum vate insurer, launched its global climate initiative. net exposure is therefore determined by the amount Under this initiative, Zurich provides PRI for compa- of available capital. nies that invest in programs to limit greenhouse gas 18. An internal review found that creditors would emissions. perceive an indirect extension of the Bank's preferred 12. FitchRatings uses the rating of the insurance creditor status to the nonguaranteed portion of the company's insurer financial strength as a first step in debt, possibly pressure the Bank to lend in the event the process of rating a structured finance with PRI of financial distress of the borrower and reduce flexi- coverage. For example, if OPIC or MIGA provides the bility for the borrower in the event of debt reschedul- PRI, an AAA insurer financial strength is assumed by ing or restructuring; it also found that the market might Fitch because OPIC is a U.S. government agency and implicitly overprice Bank credit based on its valuation MIGA is a member of the WBG (FitchRatings 2005). Pro- of the guaranteed portion of the debt as well as the un- vision of PRI for structured finance and for capital guaranteed portion, thereby raising the cost of bor- markets is not new. OPIC provided its first capital mar- rowing for the Bank. kets transaction PRI coverage (transfer restriction and 19. Initial expectations were that although IBRD inconvertibility) in 1999 for the placement of $150 guarantees covered, on average, 20 percent of project million of debt obligations of Otosan, the Ford-Koc financing, IDA guarantees would cover a higher pro- Group joint venture automobile manufacturer domi- portion of project financing. ciled in Turkey. Since then and through 2005, there have 20. It should be noted that Nam Theun 2 was the been 30 transactions issued by several providers, in- first hydropower project to be approved in 10 years cluding Zurich U.S. Political Risk, Sovereign, and oth- using the PPP format and also the first to be financed ers, including MIGA. by the Bank after the publication of the findings of the 13. Measured from the time the definitive applica- World Commission on Dams. Thus, the project came tion is submitted to MIGA until the guarantee contract under significant international scrutiny, and it is unlikely is issued. This number excludes outliers. that the private sector would have gone ahead with the 1 1 1 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 project without a PRG. The Lao PDR Nam Theun 2 on the fact that there are liquidity requirements on practices and lessons offer the private sector a new way which Treasury returns are above the cost of funding. of building better dams. 21. Structured finance operations have seen strong Chapter 3 growth. Since 2000, when the Structured Finance unit 1. In its sovereign-linked credit guarantee applica- was established, the compound growth rate of struc- tion by IFC, the instrument actually excludes certain po- tured finance deals booked is 35 percent. litical risks. 22. The African Enterprise Fund has discontinued 2. The functional approach is appropriate for small use of guarantees to enhance single credit for SMEs in companies that offer one product and that need pro- line with an overall shift in IFC away from direct sup- prietary expertise and scale. port to SMEs. 3. IFC's B-loans represent beneficial interests in a 23. All-in-costs include the spread, commission, in- trust or direct participation in the loan only; there- terest payments, and any other fees resulting from the fore, they do not insure immediate payment in the transaction. event of a loss. Moreover, because IFC's loans are not 24. Two IFC Korean Trade Facilities approved in the guaranteed by the host government, there is far more aftermath of the Asian crisis had the best utilization, at risk embedded in its loans and guarantee portfolio de- about 50 percent, and the Brazil 2002­03 trade facility spite its preferred creditor treatment (Standard & was fully utilized. Poor's 2007). Also, IFC's B-loans do not substitute for 25. The China Utility-Based Energy Efficiency Fi- coverage of risks such as expropriation, breach of con- nance Program was started in 2006 with an RSF sup- tract, and war and civil disturbance. porting a total portfolio of $100 million. As of March 4. An ADB evaluation of its PRG program also found 2008, $65 million had been used. A follow-on to the pro- that "if a choice is given to borrowers [within the $50 gram's RSF was approved by the Board to support an- million per project guarantee limit], they will always pre- other portfolio of $335 million. Development of similar fer to receive ADB assistance in the form of a direct loan RSFs has started in the Philippines, Vietnam, and In- as opposed to a more time-consuming PRG-supported donesia, all designed to support energy efficiency, re- loan funded by commercial banks" (ADB 2000). newable energy, and other types of investment whose 5. Based on MIGA estimates, MIGA derives 75 per- implementation will lead to direct reduction of green- cent of its business directly or indirectly from commer- house gas emission. Of the original energy-efficiency cial banks. RSFs in Hungary and Central Europe, about $67 million 6. In cases where the Bank's PRG pricing would be of $87 million approved was used by December 2007. higher than prices prescribed by the current approach, The Commercializing Energy Efficiency Finance off- the difference might be transferred to the host gov- spring project in Hungary is for up to $128 million, of ernment, thus reducing disincentives for governments which less than 10 percent has been used. to support PRGs with counter-guarantees. 26. IFC has approved several carbon delivery guar- 7. The 2002 WBG Private Sector Strategy recom- antees, but the product is still at a development stage. mended explicitly to "unbundle subsidies from IFC 27. For instance, if IFC develops capacity to provide financial products that support private firms and to al- local currency financing, then the need for the traditional locate such subsidies more transparently to purposes single-credit partial-credit guarantee will diminish. How- that merit being supported with subsidies" (World ever, the demand for an RSF will be there until IFC de- Bank 2002b, p. 52). velops the capacity to handle small investments. 8. International banks, which are MIGA's main ben- 28. The argument may not be based on funding eficiaries, look at spreads when deciding whether to pur- costs, but on the unfunded nature of guarantees and chase PRI. 1 1 2 BIBLIOGRAPHY ADB (Asian Development Bank). 2000. Review of Par- Ernst and Young. 2005. "Financial Guarantees: Amend- tial Risk Guarantee of the Asian Development ments to IAS 39 and IFRS 4." http://www.ey.com/ Bank. Manila: ADB. Global/Assets.nsf/Russia_E/IFRS_Financial_ Berger, Allen, and Gregory F. Udell. 1988. "Collateral, Guarantees/$file/IAS_39_IFRS_4%20Fin_Guarantee_ Loan Quality, and Bank Risk." Finance and Eco- amendments.pdf. nomics Discussion Series 51, Board of Governors of Fitch Ratings. 2005. "Political Risk Insurance and Struc- the Federal Reserve System, Washington, DC. tured Finance." International Criteria Report. New BIS (Bank for International Settlements). 2006. Basel York: Fitch Ratings. II: International Convergence of Capital Meas- Haralz, Jonas. 2007. The IFC. http://www.ifc.org. urement and Capital Standards: A Revised Frame- Hays, Larry, Diane Audino, and Marie Cavanaugh. 2001. work. http://www.bis.org/publ/bcbs128.htm. " `B' Loans and Political Risk Insurance: Different Booz Allen Hamilton. 2005. "Market Perceptions of Roads to the Same Destination." http://www. MIGA Political Risk Insurance: Survey of MIGA's standardandpoors.com, under ResourceCenter- Clients and Political Clients." McLean, VA: Booz Allen RatingsCriteria-Sovereigns. Hamilton. Heitfield, Erik, and N. Barger. 2003. "Treatment of Culp, Christopher. 2003. "Credit Risk Management Les- Double-Default and Double-Recovery Effects for sons from Enron." In Corporate Aftershock: The Hedged Exposures under Pillar I of the Proposed Public Policy Lessons from the Collapse of Enron New Basel Capital Accord." White paper, Board of and Other Major Corporations, ed. Christopher L. Governors of the Federal Reserve System, Wash- Culp and William A. Niskanen, 211­35. New York: ington, DC. Wiley & Sons. IEG (Independent Evaluation Group). 2006. IEG Review Delmon, Jeff. 2007. "Mobilizing Private Finance with of World Bank Assistance for Financial Sector IBRD/IDA Guarantees to Bridge the Infrastructure Reform. IEG Study Series. Washington, DC: World Funding Gap." http://siteresources.worldbank.org/ Bank. INTSDNETWORK/Resources/2007_6_Risk_ IEG­MIGA (Multilateral Investment Guarantee Agency). Mitigation_Final.pdf. 2008a. IEG-MIGA Annual Report. Washington, DC: Demirgüç-Kunt, Asli, and Enrica Detragiache. 2005. World Bank. "Cross-Country Empirical Studies of Systemic Bank ------. 2008b. Evaluating MIGA's FY05­08 Strategic Distress: A Survey." National Institute Economic Re- Directions. Washington, DC: World Bank. view 192 (April): 68­83. ------. 2007. IEG-MIGA Annual Report. Washington, Dittmar, Robert F., and Kathy Yuan. 2008. "Do Sover- DC: World Bank. eign Bonds Benefit Corporate Bonds in Emerging ------. 2006. IEG-MIGA Annual Report. Washington, Markets?" The Review of Financial Studies, 21(5): DC: World Bank. 1983­2014. IFC (International Finance Corporation). 2008. "IFC EIU (Economist Intelligence Unit and Columbia Pro- Road Map, FY09­11. Creating Opportunity: Doing gram on International Investment). 2007. World Our Part for an Inclusive and Sustainable World." IFC Investment Prospects to 2011: Foreign Direct In- Report, World Bank, Washington, DC. vestment and the Challenge of Political Risk. Lon- ------. n.d. "Guaranteed Offshore Liquidity Facility don and New York: Economist Intelligence Unit. (GOLF)." Product description, IFC, Washington, DC. 1 1 3 T H E W O R L D B A N K G R O U P G U A R A N T E E I N S T R U M E N T S 1 9 9 0 ­ 2 0 0 7 IMF (International Monetary Fund). 2003. "Asian Bond Coverage." Presentation to Berne Union Technical Markets: The Role of Securitization and Credit Panel, July. Guarantees." Presentation to the 2nd ASEAN Cen- Political Risk Insurance Newsletter. 2007. III (1). tral Bank Governors Meeting, June 8­9. Pozzolo, Alberto. 2004. "The Role of Guarantees in ------. 2002. "How Effectively Is the Market for Credit Bank Lending." Temi di discussione del Servizio Risk Transfer Vehicles Functioning?" Global Finan- Studi, No 528, Banka D'Italia. cial Stability Report (March), 36­47. ------. 2003. "Comparative Review of IFI Risk- Keynes, J.M. 1936. The General Theory of Employ- Mitigation Instruments and Direct Sub-Sovereign ment Interest and Money. New York: Harcourt, Lending." Report submitted to World Bank, Wash- Brace and Company. ington, DC. Kiff, Jeff, François-Louis Michaud, and Janet Mitchell. Schwartz, Robert F. 2006. "Risk Distribution in the Cap- 2002. "Instruments of Credit Risk Transfer: Effects on ital Markets: Credit Default Swaps, Insurance and a Financial Contracting and Financial Stability." http:// Theory of Demarcation." Fordham Journal of Cor- www.papers.ssrn.com/sol3/papers.cfm?abstract_id=3 porate & Financial Law XII (1): 167­201. 74640. Shihata, Ibrahim F.I. 1998. "MIGA's Creation and Evo- Kothari, Vinod. 2007. Introduction to Credit Deriva- lution: A Personal Account." In MIGA: The First Ten tives. http://www.credit-deriv.com. Years: 1988­1998. Washington, DC: World Bank. Levitsky, Jacob. 1997. "Best Practice in Credit Guaran- ------. 1988. MIGA and Foreign Investment: Origins, tee Schemes." The Financier 4 (1, 2). Operations, Policies, and Basic Documents of the Meddin, Lee. 2005. "Structured Local Issuance in Emerg- Multilateral Investment Guarantee Agency. Dor- ing Markets Enter Prime-Time." Global Securitisa- drecht and Boston, MA: Martinus Nijhoff. tion Review 2005/06. Essex, U.K.: Euromoney Books. Standard & Poor's. 2007. Supranaturals, Special Edi- Merton, Robert C., and André F. Perold. 1998. "Theory tion--International Finance Corporation. New for Risk Capital in Financial Firms." In The Rev- York: McGraw-Hill Companies. olution in Corporate Finance, ed. Joel M. Stern, Liza Stern, Joel, and Donald Chew, Jr., eds. The Revolution H. Jacobs, and Donald H. Chew, Jr. Oxford, in Corporate Finance. Oxford, UK: Blackwell. UK: Blackwell, 385­402. UNCTAD (United Nations Conference on Trade and De- MIGA (Multilateral Investment Guarantee Agency). velopment). 2005. World Investment Report. New 2007a. "Management Discussion and Analysis." MIGA York: UNCTAD. Annual Report. Washington, DC: World Bank. Winpenny, James. 2005. Guaranteeing Development? ------. 2007b. Operational Regulations. http://www. The Impact of Financial Guarantees. Paris: OECD miga.org/documents/Operations-Regulations.pdf. Development Centre. ------. 1998. MIGA: The First Ten Years: 1988­1998. ------. 2003. Financing Water for All: Report of the Washington, DC: World Bank. World Panel on Financing Water Infrastructure. ------. 1994. MIGA: The First Five Years and Future Global Water Partnership/World Water Council/Third Challenges. Washington, DC: World Bank. World Water Forum. Mintzberg, Harry. 1994. The Rise and Fall of Strategic World Bank. 2007a. Global Development Finance 2007: Planning. New York: The Free Press. The Globalization of Corporate Finance in Devel- Moran, Theodore, and Gerald West. 2007. Interna- oping Countries. Washington, DC: World Bank. tional Risk Management. Washington, DC: World ------. 2007b. "Optimizing Delivery of Insurance and Bank-MIGA. Guarantee Instruments within the World Bank ------. 2005. International Risk Management, Looking Group." World Bank Group Guarantees Note, World to the Future. Washington, DC: World Bank­MIGA. Bank, Washington, DC. ------. 2004. International Risk Management, The ------. 2007c. "World Bank Guarantees." Technical Brave New World. Washington, DC: World Bank­ Briefing, June 24, World Bank, Washington, DC. MIGA. ------. 2005. "Risk Mitigation: Review of World Bank OPIC (Overseas Private Investment Corporation). 2007. Group Guarantees." Technical Briefing to Board, "Regulatory Risks--Developing a New Approach to June 16, World Bank, Washington, DC. 1 1 4 B I B L I O G R A P H Y ------. 2002a. "Lebanon: Power Sector Restructuring Working Paper No. 18925, World Bank, Washing- and Transmission Expansion Project." Implemen- ton, DC. tation Completion and Results Report No. 24274, ------. 1997b. "Overview of the World Bank Group's World Bank, Washington, DC. Guarantee Activities." Report No. SecM97-209, World ------. 2002b. "Private Sector Development Strategy: Bank, Washington, DC. Directions for the World Bank Group." Report No. ------. 1996. "Private Financing of Toll Roads." Work- SecM2002-0047/1, World Bank, Washington, DC. ing Paper No. 16437, World Bank, Washington, DC. ------. 2000. Enhancing the Use of World Bank Guar- ------. (Various years). World Enterprise Survey. http:// antees as an Operational Tool--A Review of the www.enterprisesurveys.org. World Bank Guarantee Program. Washington, DC: World Economic Forum. 2006. "Building on the Mon- World Bank. terrey Consensus: The Untapped Potential of De- ------. 1997a. "Financing Pakistan's Hub Power Proj- velopment Finance Institutions to Catalyze Private ect: A Review of Experience for Future Projects." Investment." 1 1 5 ISBN 978-0-8213-7679-9 9 0 0 0 0 9 780821 376799 SKU 17679