2 ffiA WO R LD B A NK P O L I C Y RE S EA R CH R EP OR T ,t1' ''''n F,,. : ' qfl ,-X '''- ............................ ' -- .". . ' :. -.-, ....... '' ,, ., ', 44 ~ ~ ~ ~ ~ ~ ~ pi 200 A,+-t 0 . r. i t1_ 'N~~~~~~~~~~~~~~~F '/, ~ ~ ~ ~ ~ ~ ~ ~ " or ; - t ! > zj > !¢ ; f - ; f§;!. <| _ '' \"'-N': FINANCE forGROWTH POLICY CHOICES IN AVOLATILE WORLD A World Bank Policy Research Report :I ,I "I FINANCE forG ROWTH POLICY CHOICES IN A VOLATILE WORLD A copublication of the World Bank and OXFORD UNIVERSITY PRESS Oxford University Press OXFORD NEW YORK ATHENS AUCKLAND BANGKOK BOGOTA BUENOS AIRES CALCUTTA CAPE TOWN CHENNAI DAR ES SALAAM DELHI FLORENCE HONG KONG ISTANBUL KARACHI KUALA LUMPUR MADRID MELBOURNE MEXICO CITY MUMBAI NAIROBI PARIS SAo PAULO SINGAPORE TAIPEI TOKYO TORONTO WARSAW and associated companies in BERLIN IBADAN C) 2001 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW., Washington, D.C. 20433, USA Published by Oxford University Press, Inc. 198 Madison Avenue, New York, N.Y. 10016 Cover photo credits The five coins: © Copyright the British Museum. The stock exchange photo: C) Danny Lehman/Corbis. Oxford is a registered trademark of Oxford University Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press. Manufactured in the United States of America First printing April 2001 1 234504030201 The findings, interpretations, and conclusions expressed in this study are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any information storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint, please send a request with complete information to the Copy- right Clearance Center, Inc, 222 Rosewood Drive, Danvers, MA 01923, USA, fax 978-750-4470 All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street N.W, Washington DC, 20433, fax 202-522-2422, e-mail pubrights@worldbank.org Library of Congress Cataloging-in-Publication Data has been appliedfor. ® Textprinted onpaperthatconforms to thenAmerican NationalStandard for Permanence of Paperfor Printed Library Materials Z39.48-1984 Contents Foreword ix The Report Team xiii Acronyms and Abbreviations xv Overview and Summary 1 Summary 4 Policy Implications and Stylized Applications 23 The Next Generation of Research 29 1. Making Finance Effective 31 How Finance Helps 34 Structure 44 Financial Infrastructure 55 Access 65 Conclusions 71 2. Preventing and Minimizing Crises 75 Why Finance Has Been So Fragile ... and Remains That Way 78 Regulating Banks: Harnessing the Market 90 Financial Sector Safety Nets 104 Conclusions 118 3. Government Failure in Finance 123 Bureaucrats as Bankers? 124 Governments as Caretakers 140 Conclusions 153 4. Finance without Frontiers? 157 Capital Account Liberalization: Costs and Benefits 160 Financial Services: Allowing Foreign Provision 162 Opening the Equity Market 169 v CONTENTS Debt Flows and Interest and Exchange Rates 179 Into the Future: Technology and Communications 190 Conclusions 194 References 197 Boxes 1.1 Using regression coefficients to infer policy effects 38 1.2 Finance as an export sector 43 1.3 Time, income, and inflation: stylized facts about financial depth 47 2.1 Poverty and crises 76 2.2 Narrow banking 91 2.3 Subordinated debt proposals 103 2.4 The rise of deposit insurance? 106 2.5 Implicit value of deposit insurance to the bank's shareholders 112 3.1 Political economy and financial policy 129 3.2 Can bank privatization be sustained? 132 3.3 The rise, reprieve, and fall of state banks in Africa 134 3.4 Looting in the Czech Republic 138 3.5 Intervening sets the stage for the next crisis 145 3.6 Lessons from the Reconstruction Finance Corporation 150 3.7 The Swedish experience: a Saab in every garage? 152 4.1 Depositary receipts and country funds 163 4.2 Poor sequencing of Korea's financial liberalization 181 4.3 Theory of twin crises-currency and banking 185 4.4 Derivatives and capital control evasion 186 4.5 Dollarization-asset price and pass-through effects 188 Figures 1 Financial depth and per capita income 5 2 Financial depth and growth 6 3 Bank-to-market ratio and per capita GDP 7 4 East Asia poverty before and after the financial crises 11 5 Government ownership of bank assets and per capita income 15 6 National financial systems ranked by size 19 7 Comparing the share of foreign and state ownership in crisis and non-crisis countries 21 1.1 Financial development and per capita income 35 1.2 Financial development over time 36 1.3 Naive and modeled impact of financial development on growth 37 1.4 Financial depth and macroeconomic volatility 39 1.5 Relative contribution of financial development to productivity and capital intensity 40 1.6 Three measures of the relative development of banks and organized securities markets 45 Vi CONTENTS 1.7 Average leverage of listed firms in industrial and developing economies 50 1.8 Intermediation spreads 51 1.9 Measures of stock market and banking development 52 1.10 Market value of family-owned firms as a percentage of the total equity market value of the top 20 firms 54 1.11 "And the owner is...the Suharto family group" 55 2.1 Total fiscal costs (increases in the stock of public debt) relative to GDP in the year of crisis 83 2.2 Estimates of fiscal cost of and output dip for 39 banking crises 85 2.3 Volatility by region, 1970-99 86 2.4 Volatility in asset markets 87 2.5 Classification of substandard loans, 1997 95 2.6 Subordinated debt in Argentina, 1996-99 104 2.7 Explicit deposit insurance systems: the rise of deposit insurance around the world, 1934-99 105 2.8 Deposit insurance coverage 108 2.9 Deposit insurance: net benefits 114 3.1 State ownership in banking, 1998-99 125 3.2 Government ownership of bank assets 126 3.3 Nonperforming loans, Argentina, 1991 132 3.4 Lending to state-owned enterprises in Argentina 133 3.5 Government ownership of banks during the East Asia crisis 141 3.6 Stylized evolution of three banks through a crisis: sound, salvageable, and doomed 149 4.1 Share of developing economies in aggregate world money stock 159 4.2 Increase in the market share of majority foreign-owned banks, selected countries, 1994 and 1999 165 4.3 Estimated impact of foreign bank entry on domestic bank performance 167 4.4 Stock of foreign holdings of (gross) debt and equity, 1997 170 4.5 Equity market liberalization dates 172 4.6 Impact of liberalization on the amplitude of equity price cycles 174 4.7 Mexico country fund discount, 1993-99 177 4.8 Real treasury bill yields for industrial and developing countries 183 Tables 1.1 Effects on credit availability of adopting a negative-only credit scoring model for various default rates 70 2.1 Selected financial crashes 80 2.2 Typical balance sheet 94 3.1 Estimated individual impact of an accommodating approach to resolution policies 143 4.1 Real interest rates 183 vii Foreword T HE WORLD BANK GROUP HAS LONG RECOGNIZED THAT poverty reduction and growth depend on effective national financial systems. Understanding just how finance contributes to development-and how good policy can help guarantee its contribution-has been the focus of a major research effort at the Bank in recent years. This research has included systematic case-study analyses of the experiences of specific countries, as well as more recent econometric analyses of extensive cross-country data sets. Financefor Growth draws on this research and uses it to develop an integrated view of how financial sector policy can be used in the new century to foster growth and bring about poverty reduction. At its best, finance works quietly in the background; but when things go wrong, financial sector failures are painfully visible. Both success and failure have their origins largely in the policy environment. Policy needs to create and sustain the institutional infrastructure-in such areas as infor- mation, law, and regulation-that is essential to the smooth functioning of financial contracts. Above all, policymakers need to work with the mar- ket to help align private incentives with public interest. As the ever-dimin- ishing cost of communications and information technology leads to greater integration of global financial markets, policymakers face new challenges in ensuring this alignment. Governments must be prepared to recast their policies to take advantage of the opportunities resulting from global inte- gration, and also to guard against the associated risks. This book draws on the latest research to confirm some long-held views and challenge others. Some commentators have long regarded fi- nance as largely irrelevant to the drive for poverty reduction; but the evidence here shows clearly that financial development has a strong and independent role in increasing general prosperity. Countries that build a ix FOREWORD secure institutional environment for financial contracts, making it pos- sible for banking and organized securities markets to prosper, will see these efforts bear fruit in the fight against poverty. Good regulation of financial firms is an essential part of this story. But regulation is also becoming increasingly complex, and this book provides some guidelines for negotiating that complexity. Policymakers must pay special attention to the incentives created by the regulatory system: they should align private incentives with the public interest in such a way that scrutiny of financial institutions by official supervisors is buttressed by supervision by market participants. The book makes it clear that what works best will depend on country circumstances-for example, in some countries introduction of explicit deposit insurance may need to await complementary institutional strengthening. Although there is much for governments to do, there are other areas where the public sector tends not to have a comparative advantage, most notably in ownership of financial firms. Here again the problem is one of incentives and political considerations. Among other problems, deci- sions are too often based not on efficiency considerations, but rather on desires to reward particular interest groups. For this reason, well-crafted privatization can yield considerable social benefits. Even when, in a cri- sis, governments find it expedient to take control of banks, their aim should be to divest again as quickly as practicable-keeping in mind the threats of insolvency and looting by insiders if privatization takes place too rapidly in a weak institutional environment. Many countries are increasingly relying on foreign firms to provide some financial services. It is inevitable that this trend will continue. For one thing, the financial systems of almost all economies are small in relation to world finance. For another, the Internet and related technol- ogy increase the porosity of national financial frontiers. Although gov- ernments may need to adopt capital controls on inflows in some cir- cumstances, they would be wise to make sparing use of policies that protect domestic financial firms from foreign competition. The evidence suggests strongly that growth and stability in national economies are best served by ensuring access to the most efficient and reputable finan- cial services providers. Although financial openness does introduce new channels for importing economic disturbances from abroad, those risks are more than offset by the gains. New developments in communications and information technol- ogy will be an important driver for finance, too. Not only will they x FOREWORD make finance more international, but they will also help extend its reach, thereby crucially increasing the access of small enterprises and others now excluded in practice from the formal financial system. In- formal finance will continue to be important, of course, and that is one of the topics taken up by this year's forthcoming World Develop- ment Report, 2001/2002: Institutions for Markets, which will comple- ment the current volume. If implemented, the financial reforms proposed in this book can have pervasive-if often intangible-effects in expanding economic prosper- ity. At the same time, many of these reforms will be opposed by power- ful interest groups. The stakes in this contest are high. The World Bank Group is committed to continuing to work with member countries to develop and implement reforms by helping them to devise national poli- cies that are firmly based on empirical evidence and that draw on good practices from other countries. Nicholas Stern Senior Vice President and Chief Economist The World Bank March 2001 xi The Report Team T HIS POLICY RESEARCH REPORT WAS WRITTEN BY GERARD Caprio (Development Research Group and Financial Sector Strategy and Policy Department) and Patrick Honohan (Development Research Group), with the editorial assistance of Mark Feige. It takes stock of and synthesizes results to date from a research program on financial sector issues overseen by Paul Collier and Lyn Squire. Original research as background for this report includes work by the authors and by Thorsten Beck, Craig Burnside, Robert Cull, Aslh Demirguc-Kunt, David Dollar, James Hanson, Philip Keefer, Leora Klapper, Aart Kraay, Ross Levine (now at the University of Minnesota), Millard Long, Giovanni Majnoni, Maria Soledad Martinez-Peria, and Sergio Schmukler. The authors benefited from conversations with and comments by the Financial Sector Board and by Amar Bhattacharya, Biagio Bossone, Craig Burnside, Constantijn Claessens, Paul Collier, Simeon Djankov, Bill East- erly, Alan Gelb, Thomas Glaessner, James Hanson, Daniel Kaufmann, Hiro Kawai, Michael Klein, Daniela Klingebiel, Luc Laeven, Carl-Johan Lindgren (IMF), Millard Long, Giovanni Majnoni, Donald Mathieson (IMF), Frederic Mishkin (Columbia University), Ashoka Mody, Jo Ann Paulson, Larry Promisel, Jo Ritzen, Luis Serven, and Mary Shirley, many of whom contributed underlying research as well. A good part of the research originated in two large World Bank research projects, Financial Structure (led by Ash Demirgiiu-Kunt and Ross Levine) and Deposit Insurance (also led by Ash), and data gathered in the more recent Bank Regulation and Supervision project (by James Barth, Ross Levine, and Gerard Caprio) arrived in time to contribute as well. Members of the Financial Stability Forum Working Group on Deposit Insurance made helpful suggestions to chapter 2. The authors would like to acknowledge the excellent research assistance of Anqing Shi, Iffath Sharif, and Ying xiii T HE REPO RT TEANM Lin, and superb administrative support by Agnes Yaptenco. Polly Means made stellar contributions to the graphics and design. Book design, editing, production, and dissemination were coordinated by the World Bank Publications team. The judgments in this policy research report do not necessarily reflect the views of the World Bank Board of Directors, or the governments they represent. xiv } i/ Acronyms and Abbreviations ADR American depositary receipt AMC Asset management company DIS Deposit insurance system DR Depositary receipt EU European Union FDI Foreign direct investment GCB Ghana Commercial Bank GDP Gross domestic product GDR Global depositary receipt GNP Gross national product lAS International Accounting Standard IPO Initial public offering LOLR Lender-of-last-resort LTCM Long Term Capital Management M2 Broad money NAFTA North American Free Trade Agreement NBC National Bank of Commerce NPL Nonperforming loan OECD Organization for Economic Co-operation and Development PPP Purchasing power parity RFC U.S. Reconstruction Finance Corporation SEC Securities and Exchange Commission SME Small and medium-size enterprises SOE State-owned enterprise UCB Uganda Commercial Bank xv Overview and Summary S THE DUST SETTLES FROM THE GREAT FINANCIAL Financial policymaking is A crises of 1997-98, the potentially disastrous conse- one of the key development quences of weak financial markets are apparent. But issues even when there are no crises, having a financial system that does a good job of delivering essential services can make a huge difference to a country's economic development. Ensuring robust financial sector development with the minimum of crises is essential for growth and poverty reduction, as has been repeatedly shown by recent research findings. Globalization further challenges the whole design of the financial sector, potentially replacing domestic with international providers of some of these services, and limiting the role that government can play-while making their remaining tasks that much more difficult. The importance of getting the big financial policy decisions right has thus emerged as one of the central development challenges of the new century. The controversy stirred up by the crises, however, has pointed to the weaknesses of doctrinaire policy views on how this is to be achieved. How then should financial policymakers position themselves? This book seeks to provide a coherent approach to financial policy design-one that will help officials make wise policy choices adapted to local circumstances and seize the opportunities offered by the international environment. With informed policy choices, finance can be a powerful force for growth. This is not a book that relies on the application of some abstract This report presents an principles; rather, our conclusions are based on an analysis of concrete analysis of the evidence evidence. Though much remains to be learned, a huge volume of em- pirical analysis, drawing on a growing body of statistical data, has been conducted on these issues over the past few years. The findings of this research greatly help to clarify the choices that are involved. Many 1 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD long-held beliefs have found detailed empirical confirmation for the first time; some new and perhaps surprising discoveries have been made. In other words, we are asking policymakers to face some facts about finance. It is now possible to define with some confidence the need for a refocusing and deepening of the financial sector policy agenda. In this study, we identify and synthesize what we believe to be the key findings of recent financial sector research, both that conducted at the World Bank and elsewhere, highlighting the policy choices that will maximize growth and restore the financial sector as a key sector for helping to cope with-rather than magnifying-volatility. A few key messages have emerged from this research. Finance contributes to It is obvious that advanced economies have sophisticated financial sys- long-term prosperity teins. What is not obvious, but is borne out by the evidence, is that the services delivered by these financial systems have contributed in an impor- tant way to the prosperity of those economies. They promote growth and reduce volatility, helping the poor. Getting the financial systems of devel- oping countries to function more effectively in providing the full range of financial services-including monitoring of managers and reducing risk- is a task that will be well rewarded with economic growth. Governments are not good Government ownership of banking continues to be remarkably wide- at providing financial spread, despite clear evidence that the goals of such ownership are rarely services- achieved, and that it weakens the financial system rather than the con- trary. The desirability of reducing, even if not necessarily eliminating, state ownership in low- and middle-income countries where it is most widespread, follows from this evidence. However, privatization has to be designed carefully if the benefits are to be gained and the risks of an early collapse minimized. even when a crisis hits Even governments averse to an ownership role in banking may find it foisted on them in a crisis. The authorities' focus then must be on get- ting out as quickly as possible, using the market-rather than govern- ment agencies-to identify winners and losers. Drawing on public funds to recapitalize some banks may be unavoidable in truly systemic crises, but they must be used sparingly to leverage private funds and incentives. Procrastination and half-measures-as reflected in lax policies involving regulatory forbearance, repeated recapitalizations, and their ilk-bear a high price tag that will affect the financial system and the economy for years to come. Achieving an efficient and secure financial market environment re- quires an infrastructure of legal rules and practice and timely and accurate 2 OVERVIEW AND SUMMARY information, supported by regulatory and supervisory arrangements that But well functioning help ensure constructive incentives for financial market participants. Suc- markets need legal and cess here will promote growth in a way that is tilted towards the poor and regulatory underpinning- will stabilize the economy around the higher growth path; direct access to finance by many now excluded will also be expanded. Incentives are key to limiting undue risk-taking and fraudulent behav- ior in the management and supervision of financial intermediaries- especially banks that are prone to costly failure. Instability and crashes are and a strategy based on endemic to financial markets, but need not be as costly as they have been harnessing incentives in recent years. They reflect the results of risk-taking going well beyond society's risk tolerance. These costs are very real: they represent a poten- tially persistent tax on growth. This can raise poverty in the near term, and can have longer-term affects on the poor, both through lower growth and through reduced spending on areas such as health and education. Deposit insurance systems, an important part of the safety net sup- Good safety nets require porting banks, are on the rise in developing countries. It is not hard to good institutions see why: not only will a credible system protect against depositor runs, but they are politically popular-not least with the local owners of small banks. However, recent evidence shows that they also lessen market monitoring of banks. Although this may not have weakened banking systems in developed markets, to the extent that these had already ac- quired reasonably effective regulation and supervision, it is found to heighten the risk of crisis and reduce financial market development where institutions are weak. Thus, authorities considering deposit insurance should make an audit of their institutional framework the first step in the decisionmaking process. Good safety net design needs to go beyond replication of mature systems, and the empirical evidence strongly ar- gues for utilizing known market forces in order to limit the risks that may be associated with introducing deposit insurance. Banks, securities markets, and a range of other types of intermediary Diversity is good for and ancillary financial firms all contribute to balanced financial devel- stability and development opment. A radical preference in favor either of markets or of banks can- not be justified by the extensive evidence now available. Instead, devel- opment of different segments of the financial system challenges the other segments to innovate, to improve quality and efficiency, and to lower prices. They also evolve symbiotically, with expansion of one segment frequently calling for an upgrade in others. The future of some nonbank sectors, notably private pension provision, are heavily dependent on re- lated government policies, whose design needs careful attention. 3 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD Open markets can spur Most developing countries are too small to be able to afford to do with- development- out the benefits of access to global finance, including accessing financial services from foreign or foreign-owned financial firms. Facilitating the entry of reputable foreign financial firms to the local market should be wel- comed too: they bring competition, improve efficiency, and lift the qual- ity of the financial infrastructure. As such, they are an important catalyst for the sort of financial development that promotes growth. Opening up is accompanied by some drawbacks, including a heightening of risk in some dimensions, and will need careful monitoring. It also results in a loss of business for local financial firms, but access to financial services is what matters for development, not who provides them. as can technology- The financial sector has long been an early adopter of innovations in information and communications technology. Internationalization of fi- nance (despite efforts to block it) has been one consequence. This has helped lower the cost of equity and loan capital on average even if it has also height- ened vulnerability to capital flows. The precise future role of e-finance in accelerating the process of internationalization is not easy to predict, but it will surely be substantial. If volatility may have increased, so too have risk management technologies and their associated financial instruments. notably for access Some related credit information techniques, including scoring mecha- nisms, promise to make an important contribution by expanding what is at present very limited access of small-scale borrowers to credit from the formal financial sector. This will be achieved by lowering the barrier of high information costs. At the same time, a degree of subsidization of overhead costs will still likely be appropriate to contribute to the viabil- ity of microcredit institutions targeted at the poor and very poor. In this overview, after summarizing the main arguments of the books four main chapters, we analyze the main policy implications, presenting an illustrative stylized application to contrasting country conditions. The over- view concludes with a prospect offuture research. Summary T HIS SECTION OF THE OVERVIEW SUMMARIZES THE REASONING of the remaining chapters of the report. We focus on the main findings drawn from the empirical research, and the primary implications of these findings. The detailed arguments and caveats are to be found in the succeeding chapters, along with references to the extensive body of research underlying the study. 4 OVERVIEW AND SUMMARY Chapter 1: Making Finance Effective There is now a solid body of research strongly suggesting that improve- Financial depth generates ments in financial arrangements precede and contribute to economic growth performance. In other words, the widespread desire to see an effectively functioning financial system is warranted by its clear causal link to growth, macroeconomic stability, and poverty reduction. Almost regardless of how we measure financial development, we can see a cross-country asso- ciation between it and the level of income per capita (figure 1). Associa- tion does not prove causality, and many other factors are also involved, not least the stability of macroeconomic policy. Nevertheless, over the past few years, the hypothesis that the relation is a causal one (figure 2) has consistently survived a testing series of econometric probes. The reason finance is important for growth lies in what are, despite being less obvious, the key underlying functions that financial institutions perform. At one level, finance obviously involves the trans- fer of funds in exchange for goods, services, or promises of future return, The vertical bar shows the but at a deeper level the bundle of institutions that make up an economy's interquartile range-the financial arrangements should be seen as performing several key eco- financial depth of 50 percent of nomic functions: the countries at each stage of development lie within this range. The median is shown as a horizontal bar. Figure 1 Financial depth and per capita income Private credit as percent of GDP Market capitalization as percent of GDP (median and quartiles by income level) (median and quartiles by income level, percent) Percent Percent 125 - 100 100- 80- 75- 60- 50 - 40 -i 25 - 20 - J. Low Low Upper High Low Low Upper High income middle middle income income middle middle income income income income income Note: This figure represents the average of available dates in the I 990s for each of 87 countries. Source: Beck, Demirgdi-Kunt, and Levine (BDL) database. 5 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD Figure 2 Financial depth and growth Ratio of liquid liabilities to GDP in 1960 Deep (> 0.51 0.25 to 0.5 0.15 to 0.25 Shallow 1< 0.15) i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I 0 1 2 3 Average GDP growth 1960-95 (percent per annuml Source: World Bank data. Mobilizing savings (for which the outlets would otherwise be much more limited). * Allocating capital funds (notably to finance productive investment). * Monitoring managers (so that the funds allocated will be spent as envisaged). * Transforming risk (reducing it through aggregation and enabling it to be carried by those more willing to bear it). Rigorous and diverse econometric evidence shows that the contribu- tion of finance to long-term growth is achieved chiefly by improving the economy's total factor productivity, rather than on the rate of capital accumulation. It is through its support of growth that financial development has its strongest impact on improving the living standards of the poor. Though some argue that the services of the formal financial system only benefit the rich, the data say otherwise. Furthermore, countries with a strong, deep financial system find that, on balance, it insulates them from macrofluctuations. Bank and equity financing The evidence on the importance of each of the two major institutional are complements, not components of finance-banks and organized securities markets-is also substitutes clear. There is no empirical support for policies that artificially constrain one in favor of the other. Indeed, the development of each sector seems to strengthen the performance of the other by maintaining the competitive 6 OVERVIEW AND SUMMARY edge of individual financial firms. While banking is more deeply entrenched in developing economies than securities markets and other nonbank sec- tors (figure 3), distinct challenges face policymakers in trying to ensure that both banks and markets reach their full functional potential. Macro- economic stability is, of course, one key, but other aspects relate more closely to the microeconomic underpinnings of finance. With so much of the borrowings by firms coming from banks, the borrowing cost depends on the operational efficiency and competitive- ness of the banking market. In this respect, too, the performance of developing economies falls behind. Liberalization has been associated not only with higher wholesale interest rates, but also with a widening of intermediation spreads-at least partly reflecting increased exercise of market power by banks. One path to lower financing costs through increased competition in financial markets is through the development of equity financing. Here the challenge is to alleviate the problems of information asymmetry. The complexity of much of modern economic and business activity has greatly increased the variety of ways in which insiders can try to conceal firm performance. Although progress in technology, accounting, and legal practice has also helped improve the tools of detection, on balance the asymmetry of information between users and providers of funds has not been reduced as much in developing countries as it has in advanced economies-and indeed may have deteriorated. Figure 3 Bank-to-market ratio and per capita GDP Ratio of banks' domestic assets to stock market capitalization 8- At lower levels of per capita 6 - income, the value of bank assets * *** * + ++ +tends to be a much larger 4 - * + multiple of stock market *+ *+ ++ capitalization than in higher 2 - * * + * income countries. 0- $200 $3,000 $40,000 GDP per capita Iratio scale) Source: World Bank data. 7 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD Finance needs an The current wave of policy research thus points to the desirability of infrastructure: law and policy measures that could promote the production and communication information of information; limit the exercise of market power, whether in banking or by insiders against shareholders; and ensure an efficient functioning of the organized securities markets. These policies are likely to be more ef- fective if directed to infrastructure rather than directly to the financial structures themselves. It is in the legal area that recent research on effec- tive infrastructure has made most progress-and in areas going beyond the obvious and crucial need to ensure that the creditor's rights can, in the event of default, be expeditiously and inexpensively exercised. Natu- rally, the government has a comparative advantage in the design and imple- mentation of law, and it needs to address itself to updating and refining laws and legal practice as they relate to financial contracts. Yet, to supple- ment-or make up for the absence of-government action, there is a clear and practical scope for market participants to amplify regulatory structures where this is needed. Practice in some of the more successful organized stock markets provides good examples of such private initia- tives. This presents a promising way forward, especially where the devel- opment of public law is difficult. There has been a major scholarly debate on whether the precise de- sign of laws matter, with recent research focusing on the contrasting performance of financial systems with legal structures of differing ori- gins. The evidence indicates that the main families of legal origin do differ in important respects relevant to financial development-notably in the differential protection they tend to provide to different stakehold- ers. These differences have been shown to have had an influence on the relative development of debt and equity markets, on the degree to which firms are widely held, or more generally the degree to which they are financed externally, and thus on overall financial sector development. And the policy message from the econometric results systematically points in one direction: far from impeding growth, better protection of the property rights of outside financiers favors financial market develop- ment and investment. Collective savings media The growth of collective savings-including through investment com- help strengthen and panies and mutual funds, as well as pension funds and life insurance com- upgrade the system panies-can greatly strengthen the demand side of the equity market, as well as widen the range of savings media available to persons of moderate wealth, and provide competition for bank deposits. The impact is not lim- ited to the stock market: in mature and emerging markets, contractual savings institutions have been central in supporting numerous market-based 8 OVERVIEW AND SUMMARY financial innovations such as asset-backed securities, the use of structured finance and derivative products, including index-tracking funds and syn- thetic products that protect investors from market declines. The associated learning and human capital formation, as fund managers tool up to em- ploy such techniques, helps to enhance the quality of risk management throughout the economy. Growth in these funds can also ensure enhanced and stable funding for key niche segments of the financial market, such as factoring, leasing, and venture capital companies. They can also generate a demand for long-term investments, thereby providing a market-based so- lution to a perceived gap that many governments have tried to fill over the years with costly and distorting administered solutions. Regulation of this sector is something that needs attention in many countries. Measures that succeed in deepening financial markets and limiting Policy choices and new the distorting exercise of market power result in more firms and indi- technology may expand viduals securing access to credit at acceptable cost. However, what of the access to finance- poor and of the small or microenterprise borrower? What aspects need special attention to ensure that these do not get passed by despite overall improvement in the performance of financial systems? There is no point in pretending that the problem of access is easily solved. Experience shows that formal financial institutions are slow to incur the set-up costs in- volved in reaching a dispersed, poor clientele (even with minimal deposit- type services). In looking to improvements, however, two aspects appear crucial, namely information and the relatively high fixed costs of small- scale lending. Recent research focusing on technological and policy ad- vances points to how these barriers can be lowered. A range of innovative, specialized microfinance institutions, mostly subsidized, has become established with remarkable success. Loan delin- quency has been low-far lower than in the previous generation of sub- sidized lending programs operated in many developing countries-and the reach of the institutions in terms of sheer numbers, as well as to previously grossly neglected groups, such as women and the very poor, has been remarkable. This success has been attributed to reliance on innovation in, for example, the use of group lending contracts exploit- ing the potentialities of social capital and peer pressure to reduce willful delinquency, dynamic incentives using regular repayment schedules and follow-up loans or "progressive lending," and lighter distributed man- agement structures that reduce costs and enable lenders to keep loan rates down to reasonable levels. Even without subsidy, some of these techniques can be applied to microlending to the nonpoor. Furthermore, efficient use of credit 9 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD -notably in the area of information can reduce the threshold size for cost-effective lending by microfinance the formal, unsubsidized financial sector. Computer technology has greatly reduced the unit costs of collecting information on borrowing history and other relevant characteristics, and has improved the sophis- tication with which these data can be employed to give an assessment of creditworthiness. While the impact of having this information available alters incentives and market power in subtle-and not always favorable- ways, growth in access to credit information improves loan availability and lowers intermediation costs. Chapter 2: Preventing and Minimizing Crises Finance always involves Finance is inherently fragile, largely because of the intertemporal leap in risk the dark that many financial transactions involve. Not only is money handed over now for the promise or expectation of money in the future, but this is done despite the problems of limited and unequal information both as to the characteristics of one's counterparty (adverse selection) and as to their subsequent behavior (moral hazard). Finance cannot be effective without credit, but credit means leverage, and leverage means the risk of failure, sometimes triggering a chain reaction. In these conditions, expectations can change quickly, leading to swings in asset prices, which in turn may be exacerbated by the possibility of crowd behavior. Financial markets are in the business of making efficient use of in- formation, but substantial and even growing deviations from equilib- rium prices are possible, manifesting themselves as bubbles, or specu- lative booms and busts. If the countless historical examples of asset price crashes are not sufficient evidence of this, theory, too, explains why, when acquiring information and contracting are both costly, fi- nancial markets will never be fully efficient and fully arbitraged. Care- fully controlled experiments confirm that individuals are not fully ra- tional in assessing risk: they attach too much weight to recent experience (display myopia), they trade on noise rather than on fundamentals, and they exhibit positive feedback (or momentum) by buying because prices are rising. As well as exacerbating asset-price fluctuations and contributing to euphoric surges of bank lending-followed by revul- sion and damaging credit crunches-such behavioral characteristics also provide fertile ground for fraudulent Ponzi schemes. 10 OVERVIEW AND SUMMARY If finance is fragile, banking is its most fragile part. Bankers have to place a reliable value on the assets they acquire (including the credit- worthiness of borrowers), but banking also adds the complications not only of maturity transformation, but of demandable debt, that is, of- fering debt finance backed by par value liabilities in the form of bank deposits. The particular fragility of finance, and within it of banking, is true for all countries regardless of their income level, as attested to by the occurrence of banking crises in many industrial economies in the 1980s and 1990s. But banking outside the industrial world is more dangerous still, where crises have been enormously costly-in terms of direct fiscal costs, slower growth, and a derailing of stabilization pro- grams and increasing poverty (figure 4). Developing countries face several additional sources of fragility. Not only are information problems in general more pronounced, but develop- ing economies are also smaller and more concentrated in certain economic sectors or reliant on particular export products, and accordingly are less able to absorb shocks or pool isolated risks. In addition, emerging markets have seen a succession of regime shifts altering the risk profile of the oper- ating environment in hard-to-evaluate ways, including most prominently Figure 4 East Asia poverty before and after the financial crises Poverty rate Percent 40 * 1996 * Latest 30- Poverty rises and remains elevated for some Ume following 20 -crises. 10- Indonesia Republic Thailand of Korea Note. The "Latest" column refers to 1999 for Indonesia and Thailand, and 1998 for the Republic of Korea, based on household surveys. Poverty lines are set at $1.50 per day (at 1993 PPP), except for the Republic of Korea, where the national definition of poverty is about $8 per day. Source: World Bank. 11 FINANCE IOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD financial deregulation. Moreover, as banking tends to be the dominant force in emerging financial markets, there is more demandable debt, less access to outside equity for firms, and therefore greater fragility. Collapses in equity prices are not innocuous, but are clearly less disruptive than bank failures, which explains the need to focus on the latter. Financial sector regulation and supervision-the rules of the game in the financial sector, and the way they are enforced-are essential to lim- iting moral hazard, as well as to ensuring that intermediaries have the incentive to allocate resources and perform their other functions pru- dently. Although there has been a remarkable convergence on paper in recent years, stark differences remain in regulatory environments around the world, and weaknesses in this area serve as a potential source of added vulnerability in some emerging markets. Use market-based Necessary though headline regulations may be, a clear lesson from incentives to supplement recent and historical research is that they need to be supplemented by regulations the use of incentives and information to maximize the number of well- informed, well-motivated monitors of financial intermediaries. Diver- sity in the set of monitors for banks is desirable not only because of possible differences in information that they may possess, but also be- cause of the varying and possible opaque incentives that they face. But who can monitor banks? There are three main categories: * Owners, including the board and senior management of a bank, whose net worth should depend on the prudent performance of the institution. * Markets, meaning all nonofficial outside creditors and counterparties, who should not be under the presumption that they will be "bailed out." * Official supervisors, who should operate within a well-constructed incentive structure. The aforementioned factors accounting for enhanced fragility in emerging markets means that they need to ensure that all three moni- tors are performing this function vigorously. Greater information and incentive problems certainly suggest that it is unwise to concentrate on any one of these groups. And the higher volatility of these markets implies that even adopting "best practice" from industrial economies may fall far short of the mark. This report urges that authorities go well beyond the existing Basel guidelines. Ensuring that banks are well diversified, which in many small 12 OVERVIEW AND SUMMARY economies means regional or foreign banking, is important. Motivating There is a need to go creditors, such as mandating that banks issue uninsured subordinated beyond the Basel debt, is a promising part of the solution, but requires that authorities guidelines should focus on improving the information available to these monitors and on the difficult task of ensuring that they are at arm's length from the issuing banks. Also, attention to supervisors' incentives is warranted. Higher present and especially future compensation (through bonuses or loss of generous pensions) need to be coupled with protection from legal prosecution today for effective performance of their job. In the face of financial fragility, governments provide a safety net of sorts, virtually always through lender-of-last-resort facilities and in- creasingly through explicit deposit insurance. Deposit insurance is in- creasingly popular in emerging markets because it appears to be an effective way to stem bank runs, at least in high-income countries, and helps foster indigenous banks. The existence of these schemes, how- ever, may actually worsen the information and incentive environment, increasing the scale and frequency of crises. To some extent, establish- ment of a formal deposit insurance scheme can be expected to result in greater risk-taking-the age-old moral hazard that tends to be associ- ated with most forms of insurance. That would be an argument against establishing a formal scheme, but it has to be recognized that absence of a formal scheme can be equivalent to implicit deposit insurance- perhaps unlimited in its coverage and potentially also entailing moral hazard. Thus, whether to adopt an explicit system, and what kind of system to adopt, are empirical issues. The weight of evidence from recent research suggests that, in prac- tice, rather than lowering the likelihood of a crisis, the adoption of explicit deposit insurance on average is associated with less banking sector stability, and this result does not appear to be driven by reverse causation. Here the qualification "on average" is key: deposit insur- ance has no significant effect in countries with strong institutions, but in weak institutional environments has the potential to destabilize. This result is reinforced by the finding that banks, exploiting the avail- ability of insured deposits, take greater risks. Insurance reduces depositor monitoring, which is not sufficiently compensated by official monitoring where institutions are weak. More- over, in institutionally weak environments, having explicit deposit in- surance is associated with lower financial sector development, in addi- tion to a greater likelihood of crises. Although it may be paradoxical that 13 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD the provision of insurance could lead to less of an activity, it may be that when taxpayers in institutionally weak countries see their authorities providing explicit guarantees, they understand that the environment is not conducive to restraining the cost of these guarantees. The result, then, might be that the real insurers, the taxpayers themselves, choose to hide their assets outside the banking system, and perhaps outside the country to avoid being taxed for coverage. This finding runs sharply counter to the popular doctrine that deposit insurance would promote financial deepening-and hence growth-in poor countries. The role of good institutions-as measured in this research by indi- cators of the rule of law, good governance (a proxy for effective regula- tion and supervision), and low corruption-thus seems crucial in re- ducing the opportunities for risk-taking. Good design of deposit insurance may help lead to better outcomes, but given the delays in improving regulation, supervision, the rule of law, and other basic institutions, au- thorities considering the introduction of deposit insurance should first focus on addressing these related institutions to reduce the likelihood of excessive risk-taking. And for those who already have explicit deposit insurance, it is by no means suggested that they should suddenly end these schemes-doing so would likely induce a crisis-but instead should reconsider the design of their systems in light of the evidence presented herein. In deciding on design features, this report argues that authorities should draw on empirical evidence and in particular utilize market forces to ensure prudence, rather than simply attempting to copy existing practice-itself quite diverse-of high-income countries. It is overwhelm- ingly important that governments do not provide banks with an exces- sively generous safety net, as this will hamper the development of other parts of the sector, as well as potentially underwrite excessive risk-taking. Chapter 3: Government Failure in Finance Government ownership of More than 40 percent of the world's population still live in countries in banks is greater in poor which the majority of bank assets are in majority-owned state banks. Gov- countries emient ownership tends to be greater in poorer countries (figure 5). State ownership in banking continues to be popular in many countries for sev- eral reasons. First, proponents of state control argue that the government can do a better job in allocating capital to highly productive investments. 14 OVERVIEW AND SUMMARY Figure 5 Government ownership of bank assets and per capita income Per capita GDP 1997 in international dollars 30,000 Switzerland * Singapore *Norway 3 * . ~~~~~~~~* Iceland Belgium * Germany 20,000 * *UAE * * * Israel * , * * + State banks are more common * in low income countries. 10,000 * 0 20 40 60 80 100 Share of bank assets in state banks (percent) Source: World Bank Survey of Prudential Regulation and Supervision; La Porta, L6pez-de-Silanes, and Schleifer (2000). Second, there is the concern that, with private ownership, excessive con- centration in banking may lead to limited access to credit by many parts of society. Third, a related popular sentiment-reinforced by abuses at, and governance problems of, private banks in many countries-is that private banking is more crisis prone. Despite the worthy goals often espoused by advocates of state owner- Political incentives make ship-and though there are isolated pockets of success-achievement of governments poor bankers these goals has generally been elusive, to say the least. Government failure as owner is attributed to the incentives imposed on it by the political process, and the few cases of more successful state banks appear to be linked to a stronger institutional environment and dispersed political powers. And important new statistical evidence sum- marized in chapter 3 confirms that state ownership generally is bad for financial sector development and growth. Greater state ownership of banks tends to be associated with higher interest rate spreads, less pri- vate credit, less activity on the stock exchange, and less nonbank credit, even after controlling for many other factors. It is not just financial development that is affected: one study reveals that countries that had greater state ownership of banks in 1970 tended to grow more slowly since then with lower productivity, especially in poor countries and 15 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD where the protection of property rights was weak. Credit allocation is also more concentrated, with the largest 20 firms-often including inefficient state enterprises-getting more credit where the state own- ership is greater. In addition, there is some evidence that greater state ownership is associated with financial instability. To be sure, there are exceptions: Germany, for example, has had little state ownership of the enterprise sector (outside transport and finance), which has reduced the temptation of allocating credit to gov- ernment industries. Moreover, the tough penalties there for default and bankruptcy would make life easy for most banks, even those that are state run. However, although it remains possible for developing countries to find ways to reduce the damage done by state ownership, limiting state ownership likely will be easier to implement than the many institutional and political reforms needed to avoid the abuses and inefficiencies of state banking. Privatization can lead to a The potential scale of gains from bank privatization are borne out more efficient banking from detailed investigation in World Bank research of one country with sector- comprehensive data and a major privatization experience, namely Ar- gentina. This research suggests that in an incentive-compatible environ- ment, the conduct of privatized banks-as reflected in their balance sheets and income statements-over time begins to resemble that of the other private banks. This is especially true in terms of the ratio of their admin- istrative costs to revenues, and most importantly in terms of credit ex- tended to public enterprises, consistent with the evidence above on im- proved allocation of resources. As part of the privatization process, the shedding or more efficient employment of staff, though less significant for the overall economy, works in the same direction. As compelling as the case is for private sector ownership in banking, shifting to private ownership in a weak regulatory environment can lead to crisis-witness the examples of Mexico in the early 1990s, Chile in the late 1970s, and numerous transition economies. While abrupt and pre- mature privatization can be dangerous, so too can be a strategy of hanging on to state ownership. Not only is there the evidence that this lowers growth, but also as the Czech experience points out, continued public sector con- trol of the banking system appears to have facilitated looting-the prac- tice of firms continuing to borrow without the intention of repayment. but the process should be For most countries, abrupt and total privatization is not called for. For handled with care one thing, many countries reached an advanced stage of development with modest state ownership. Also, though, a sudden move to private ownership from a lengthy period of state ownership seems particularly 16 OVERVIEW AND SUMMARY dangerous. The authorities would have to be either quite confident in their level of institutional development, or be selling to foreign banks of impeccable repute-and must be willing to gamble on this bet. Accord- ingly, moving deliberately but carefully with bank privatization-while preparing state banks for sale and addressing weaknesses in the overall incentive environment-would appear to be a preferred strategy. Prepa- ration, in addition to improvements in infrastructure, could include some linkage of compensation for senior managers of state banks to the future postprivatization value of the bank-such as through stock options, an approach that appears to have helped in Poland. To be sure, this approach can only succeed if the process is credible, otherwise the deferred com- pensation will be too heavily discounted to have any value. As also noted below, sale of state banks to strong foreign banks can be a way of bringing good skills, products, and the capacity to train local bankers, and may even facilitate a strengthening of the regulatory environment. As long as the foreign banks are motivated to protect their reputation to behave in line with the highest fiduciary standards, this approach will increase the speed with which allocation decisions are made on market principles while minimizing the odds of a crisis. When a banking crisis occurs, authorities need to decide when and Governments should how to intervene. When the problem is not systemic, bank creditors and intervene only when the supervisors should be left to proceed as usual on a case-by-case basis crisis is systemic- through standing channels. However, widespread bank insolvency may force even a government not disposed to take a significant ownership position in the banking sector to become involved in restructuring banks and even their assets (for example, nonfinancial firms) in the process. In many cases, systemic crisis has led to a substantial increase in govern- ment ownership or "care-taking." Yet the evidence on governments' lim- ited efficacy as owners of banks suggests that they will not excel at restructuring failed or failing banks either. How then can one decide when the crisis has reached systemic propor- tions and when the government should intervene with other assistance? It is not really feasible to speak in terms of mechanical triggers for this kind of judgment. For one thing, the relevant data either come with a lag, or are very imperfect measures of crisis. Besides, as the economy approaches known thresholds, moral hazard increases and bankers and other market participants may take excessive risks. The authorities would then have little option but to bring forward their intervention even though the trigger has not been reached. Because of such problems, most financial authorities have decided on constructive ambiguity as the main solution. 17 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLI) -and prepare a clear exit Once the decision to intervene has been taken, the government has strategy several goals. The first is to maintain or restore a functioning financial system. This goal is difficult to debate, though the best means of doing so are not always clear. Second, the government must contain the fiscal costs of its intervention. Care must be taken in designing restructuring plans, such that a preoccupation with minimizing short-term cash costs does not translate into larger long-term fiscal liabilities. On a related third point, governments must also ensure that their restructuring helps minimize the prospects for subsequent crises-notably in terms of the implicit incentive structures. Unfortunately, as implemented in many countries, government-funded bank recapitalization programs-injecting capital usually in the form of bonds into banks-all too often miss the opportunity to create strong incentives for future prudent behavior. This then suppresses the message that poor performance is costly. Recapitalization without establishing some corresponding financial claim on the bank-and then exercising that claim-is no more or less than a transfer from taxpayers to share- holders, which is the group that keeps the residual value of the bank. So if government funds are to be injected, there has to be some gov- ernment involvement. Governments that inject equity will want to make sure that it is used only where needed to fill an insolvency gap, and certainly that it is not looted. Yet they must recognize that they are not likely to function well as bank owners; accordingly their equity stakes in banks should be for a limited period only. One way of achieving both of these goals is for the authorities to make some amount of funding avail- able for recapitalization of banks, but only to those that Use the market to identify * Secure matching of private sector funds in some ratio. banks to be "rescued" * Agree to restrict dividends and other withdrawals by insiders for some time (likewise, contracts for senior managers should be struc- tured to emphasize deferred performance-linked compensation). * Adhere to stringent transparency requirements. The virtue of such an approach is that it removes from government or government-sponsored agencies the selection of winners, a process that is ripe for abuse. By openly stating the terms on which it will assist banks and their new shareholders, and ensuring that those terms provide good incentives for the restructured bank going forward, the government is making the best use of market forces while minimizing its direct owner- ship involvement. 18 OVERVIEW AND SUMMARY Chapter 4: Finance without Frontiers? Along with the rapid-albeit uneven-expansion of international debt and equity flows, including foreign direct investment (FDI), there has also been a sharp recent increase in the provision of financial services in many developing countries by foreign-owned financial firms. Financial globalization increases the potential for obtaining growth and other ben- efits from finance, but it also increases the risks. In a world where even the largest developing countries have financial systems whose size is dwarfed by the scale and mobility of global fi- nance, policy thinking needs to be refocused on the limited but impor- tant scope for domestic policy actions to maximize each country's capac- ity to secure the best provision of financial services, from whatever source, and to contain the risks of importing volatility. Apart from China, Brazil is the only developing country with as Consequences of being much as 1 percent of the world's financial system. The financial sys- small tems of developing countries are small, and should be managed with that in mind (figure 6). Small financial systems underperform. They suffer from a concentration of risks: the smaller the financial system, the more vulnerable it is to external shocks and the less able its finan- Total assets of the banking cial system is to insulate or hedge those shocks-unless the financial system in about one third of all system is itself securely integrated in the world financial system through countries is smaller than $1 billion; another third have banking systems smaller than $10 billion. Figure 6 National financial systems ranked by size Size of financial system M2, log scale 10,000 * Banking system > $10 billion 1,000 - v $10 billion > Banking system > $1 billion 100 - * Banking system < $1 billion 10 0.01 166 countries Source: International Financial Statistics. 19 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLAT-ILE WORLD ownership and portfolio links. Small financial systems also provide fewer services at higher unit costs, partly because they cannot exploit economies of scale and partly because of a lack of competition. Regu- lation and supervision of small systems is disproportionately costly, and even a well-funded effort would be hard pressed to ensure stability if finance is restricted to domestic institutions operating locally. Many financial systems fall short of minimum efficient scale and thus have much to gain from outsourcing financial services from abroad. It sometimes seems that a boom-and-bust roller coaster has been im- ported when the capital account has been liberalized. Undoubtedly, with the wrong incentives, this has been a threat. There have also been tan- gible gains from external liberalization, and above all there is an inevita- bility about further opening-up to foreign capital markets and financial institutions. However, despite a huge research literature, there is noth- ing near a professional consensus on whether the net impact of full capi- tal account liberalization on growth, poverty, or volatility should be re- garded as favorable or not. Governments can no longer hope to maintain a permanent and wide gap between actual and market-clearing exchange rates and real whole- sale interest rates without a panoply of administrative controls on inter- national trade, as well as on payments, to an extent that is demonstrably damaging to growth and living standards. That premise does not in it- self rule out milder forms of control, including taxes and restrictions on the admission of foreign-owned financial service companies (such as banks), on the purchase by foreigners of local equities, and on interna- tional capital movements. The evidence, however, suggests that such re- strictions should be used very sparingly. The internationalization of the provision of financial services, includ- ing the entry of reputable foreign banks and other financial firms, can be a powerful generator of operational efficiency and competition, and should also prove ultimately to be a stabilizing force (figure 7). Foreign bank entry can Some countries have remained slow to admit foreign-owned financial strengthen the system firms to the local market, fearing that they will destabilize the local finan- cial system and put local financial firms out of business, with the ultimate result that particular sectors and particular national needs will be poorly served. There is no hard evidence, however, that the local presence of for- eign banks has destabilized the flow of credit or restricted access to small firms. Instead, the entry of these banks has been associated with signifi- cant improvements in the quality of regulation and disclosure. The very threat of entry has often been enough to galvanize the domestic banks into 20 OVERVIEW AND SUMMARY Figure 7 Comparing the share of foreign and state ownership in crisis and noncrisis countries Percent of total assets 40 30- 20 - Ownership structures matter for crisis avoidance. 10 Crisis 1 - _ ned _ t o cuntnres Noncrisis Foreign-owned countes banks State-owned banks Source: Barth, Caprio, and Levine (200 1c). overhauling their cost structure and the range and quality of their services, with the result that foreign entry has often proved not to be as profitable for the entrants as they may have anticipated. There may be some downside: pressure on domestic banks from for- eign competition could present prudential risks if it erodes franchise value of high-cost operators to the point where they begin to gamble for resurrection. Also, there is the risk that some less reputable foreign bank entrants might prove to be unsound. Evidently these considerations re- inforce the urgency of strengthening prudential regulation. Actually, the arrival of reputable foreign banks is usually associated with a systemwide upgrading of transparency (especially if the banks bring improved ac- counting practices with them). The most dramatic structural developments in international finance Despite some setbacks- for developing countries over the past decade or so have been the growth in cross-border equity investment, whether in the form of direct for- eign investment (where the investor takes a controlling stake) or in the form of portfolio investment in listed or unlisted equities. The dra- matic stock market collapses in East Asia during 1997 and 1998 took much of the shine off what had seemed an almost trouble-free liberal- ization of several dozen equity markets in the previous two decades, highlighting questions about the consequences, benefits, and costs of equity market liberalization. 21 FINANCE FOR GROWTH: POL ICY CHOICES IN A VOLATILE WORLD For a country that has an active equity market, opening that market to foreign investors is a decisive step that can be expected to influence the level and dynamics of asset pricing. More than thirty sizable stock exchanges in emerging market economies undertook significant liber- alization mostly concentrated in a ten-year period from the mid- 1980s to the mid-1990s. So it is natural to ask: did the expected effects occur in practice? Were stock prices higher on average than they would oth- erwise have been? Was there an increase or a fall in the volatility of stock prices? -equity market In practice, these questions are tougher to answer than might appear liberalization has lowered at first sight. Overall, though, it appears from research findings that eq- the cost of capital without uity prices have increased, thereby lowering the cost of capital, without much increase in volatility an undue increase in volatility. Opening up has also accelerated improve- ments in disclosure and the efficiency of the local stock markets, even though these have lost some of their share of the increased business in the listing and trading of local equities. Before the explosion in international equity investment, the classic form of international finance involved debt flows: international borrowing and lending. Though carefully designed tax-like measures can be somewhat effective in damping short-term debt flows, openness to international flows inevitably impacts domestic interest rates and the exchange rate. Here is where the risks arise, and where macroeconomic, fiscal, and monetary policy has long been directed to containing those risks. Exposure of finan- cial intermediaries and others to exchange rate risks, both direct and indi- rect, can be a particularly severe source of problems. Domestic financial liberalization would be possible even without open- ing up the economy to international capital movements; with the opening-up, it becomes unavoidable. Capital account liberalization weak- ens and distorts a repressed domestic financial sector, eventually forcing domestic liberalization. If the process is long drawn out, partial liberal- ization of external and domestic finance can result in a very risky and unsound situation emerging. Liberalization both of domestic and international finance has resulted in a convergence of interest rate movements, though developing countries are now experiencing some increased interest rate volatility and a struc- tural risk premium, partly reflecting exchange rate and other policy risks. Continuing developments in computing and communications tech- nology seem sure to reshape the way in which financial services are de- livered worldwide. To sorme extent the impact on developing economies 22 OVERVIEW AND SUMMARY will be an acceleration of the trends of recent years, but there will be qualitative changes too. Economies of scale for some financial services are declining, but increasing for others, while the synergies between fi- nancial and other economic services are also changing and often increas- ing. This will alter the organization of the industry, with consolidation in some areas and fragmentation in others. This process may present some opportunities for financial service pro- Increasingly, countries can viders in small developing countries, especially where the unbundling of choose which financial financial products leaves subproducts that can be efficiently produced with services to buy and which low sunk costs, and exploits advantages of location rather than scale. How- to build ever, the greater potential benefit in prospect for developing countries will be for users of financial services, including services that have often not yet been well developed-such as pensions and other forms of collective savings-and international payments. Technology should allow those coun- tries to access these services on terms comparable to consumers in ad- vanced countries, especially insofar as physical distance from the provider begins to lose much of its importance. Undoubtedly, the accelerating pres- ence of the Internet will begin to make direct international financial trans- actions available even to small firms and individuals. The likely speed ofthese developments, and the extent to which they will displace the need for a local presence of financial service companies, remain unclear. The question that will be increasingly asked is whether srnaller de- veloping countries need to have local securities and debt markets in the traditional sense, and even how much of banking needs to be domestic. For policymakers in developing countries the questions will shift to considering the stability of domestic financial institutions in the face of the increased competition. Increased access to foreign financial services will entail more use of foreign currency, and this will accentuate the risks of exchange rate and interest rate volatility for countries that choose to retain their own cur- rency. Once again, heightened prudential alertness will be needed. Policy Implications and Stylized Applications T HE GENERAL APPROACH EMERGING FROM THIS STUDY should be clear. Evidence on the importance of sound financial infrastructure is more important than anyone thought. Unregu- lated financial systems will fail, often catastrophically, but the wrong type of regulation is counterproductive. The right type of regulation is "incentive 23 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD compatible"-that is to say, it is designed with a view to ensuring that the incentives it creates for market participants help achieve its goals rather than hinder them. More specifically, the right type of regulation Financial policies should * Works with the market, but does not leave it to the market. be market-aware * Keeps authorities at arm's length from transactions, lessening the opportunities for conflicts of interest and corruption. * Promotes prudent risk-taking, meaning risks borne by those most capable of bearing it, for example, removing distortions that lead to too little direct investment, too little equity finance, too little long- term finance, and too little lending to small firms and the poor. In short, this is financial policy that is market-aware. The wrong type of regulation includes financial repression-the main- tenance of below-market and often negative real interest rates, and forced credit allocation. Repressive policies, in many cases the wrong response to an earlier round of crises, created some of the problems we see today, including the underinvestment in skills and in the infrastructure that are needed to support a market-based financial system. The design of the financial safety net also requires careful attention if it is not to become another type of misplaced regulation. Another wrong solution is excessive reliance on one type of monitor to oversee intermediation. Prudential supervision is by now a universal fea- ture of financial policy, but supervisors are hard-pressed to keep up with financial technology and the speed with which the risk profile of banks can change. Enlisting the help of private sector participants by arranging for well-funded investors to have something at stake in the continued vi- ability of banks, and hence the incentive to monitor them, will be an increasingly important support to direct official supervision. Establishing appropriate incentives for supervisors themselves-recognized in some cases during the 19th century-will help as well and is an idea whose return is long overdue. Political structures that increase the risk that reforms such as these will be delayed need to be addressed, too; in the opinion of some scholars, it is here that the deepest causes of the wave of crisis of the past two decades should be sought. The recommendations of this report are mutually supportive in some obvious ways. For example, financial systems that are not supported by effective infrastructure and incentives systems will not be entrusted with much of society's savings. A less obvious link is that countries that 24 OVERVIEW AND SUMMARY provide heavily subsidized deposit insurance or a lax regulatory frame- work will miss out on the benefits of a diversified financial system, because nonbank and capital market development will suffer. Simi- larly, excessive state ownership is demonstrably bad for competition and usually features active or passive discouragement of foreign banks. The present condition of the financial system in many countries is far from ideal, and achieving the goals set out here may seem impossibly distant. Yet there are practical implications for all types of countries and all types of initial conditions. Without attempting to provide a detailed tactical design for reform in each case, and without pretending to do justice to the true diversity of country conditions, it is worth briefly sketching the policy implications that can be drawn for policymakers in four contrasting stylized scenarios. Although the initial conditions fac- ing policymakers differ widely, the principles of good policy that emerge from these research findings have an equally wide application. (a) A smalllow-income country dominated by state-owned financial institutions. Here we picture a low-income country, such as many in Africa-but also elsewhere-where the legacy of financial repression and state ownership has hampered the development of a vigorous private financial system. The lessons of chapter 3 are the most immediately relevant for this coun- try. Government ownership has resulted in credit being directed to underperforming state entities; incentives and professional capacity are weak in the banking system, and there may still be a hidden inheritance of doubtful loans. The priority for the state must be to divest itself of its bank holdings and to create a credible policy stance sufficient to attract reputable international bank owners. Legal infrastructure may need upgrading here, too, as discussed in chapter 1, although it is likely that judicial enforcement is the more rel- evant weak spot. In financial regulation, the political independence of the supervisors is an issue (chapter 2). Clear legal protection for them is crucial. The temptation to bolster the emerging private banks with a for- mal deposit insurance system should be resisted in view of the demon- strated moral hazard effects. Although this country needs nonbanking financial services, such as those of securities markets, it is likely too poor and too small to sustain a liquid securities market on its own (chapter 4). The authorities need to be aiming to remove barriers that prevent borrowers and lenders from accessing international capital markets. Evidently this will need to be 25 FINANCE FOR GROWEH: POLICY CHOICES IN A VOLATILE WORLD supported by stable and sustainable macroeconomic policies, as policy- induced macroinstability may be amplified by this opening-up of capital markets. Achieving minimum efficient scale-both in market infrastruc- ture and in such aspects as payments systems-is going to be a challenge. Exploring the possibilities of regional cooperation on these fronts should bear fruit. If democracy is weak and ethnic conflict high, a significant level of uncertainty will likely prevail, which will deter physical entry by good foreign banks, as will low population density. E-finance or joining a regional financial system may be the best hope of getting access to higher- quality financial services. (b) A transition economy with weak rule of law. Where the rule of law is weak, the financial sector cannot be expected to function well. Tackling this situation will be the primary challenge. The message from chapter I is that market participants may have to supple- ment formal law with private contracts that establish bright-line rules that can easily be verified and enforced, possibly using enforcement through external jurisdictions. Because the credibility of domestic institutions is so weak, it is hard to align private incentives with social goals. Certainly it will be undesir- able to institute deposit insurance, as observed in chapter 2, although it may be hard to withdraw insurance from existing state-controlled banks, which retain an important quantity of household savings in several tran- sition economies. Leveraging credibility by allowing foreign institutions to enter and to compete in the retail market is a preferable solution, which is all the more reason to privatize many such banks as expedi- tiously as possible (as proposed in chapter 3), although with care to en- sure that the new owners have significant capital at stake. This economy is likely to have a de facto open capital account, with market participants already obtaining financial services from foreign sys- tems (chapter 4). It would be better to recognize this through a formal liberalization so that such access is not an underground or illegal activity. (c) A lower middle-income bank-dominated country emerging from a crisis. Most bank-dominated, middle-income countries have recendy experienced banking crises associated with an undue burden of debt. As they seek to 26 OVERVIEW AND SUMMARY recover from these crises, the policy messages are clear. Getting the state out of a direct role in restructuring as fast as possible is important, including using the private sector to identify the banks and nonfinancial firms that are fit to survive. In the medium term, authorities need to find ways to lessen reliance on short-term debt finance. Improved protection of minor- ityshareholders, as noted in chapter 1, is needed to help boost the possibil- ity of issuing outside equity. And no doubt, improvements to the availabil- ity and reliability of information will spur nonbank finance. Also important is better monitoring of the banking system. Even to the extent that the crisis was brought on by external factors, virtually every crisis uncovers banks that have ventured far out on the risk frontier, and that may account for a large fraction of the fiscal cost. In addition to ensuring that excessive risk takers are not "bailed out," better monitoring is crucial here to convince financial sector participants that incentives have changed. Often, even if a formal deposit insurance system was not in place before the crisis, blanket coverage may be now, and it is impor- tant that this coverage begins to be limited as soon as possible. If the banks still are fragile or suspect, however, great care is needed, and intro- ducing a subordinated debt requirement-addressing the enforcement problems noted in chapter 2-can both improve monitoring and increase the share of unguaranteed liabilities. Then over time the authorities can announce a schedule of reduction of the ceiling amount of deposits cov- ered by an explicit system. For countries with relatively limited numbers of banks, the German system of private deposit insurance and mutual liability among the private banks in the scheme has much to recommend itself as a way to maximize market monitoring. Official monitoring of banks also will need improvement, and cor- recting the "balance ofterror" noted in chapter 2 will complement greater central bank independence and allow for vigorous oversight. Admitting foreign banks also can help stabilize and improve the sec- tor, and middle-income countries are more likely to have good and eager entrants, while chapter 4 shows how beneficial openness to international equity markets can be. (d) An upper middle-income country with a still-shallow finanial system. The financial development of some upper middle-income countries re- mains below average. They seem to have all the basics, but depth, term finance, and a full range of services are lacking. Here, too, the research findings of each chapter are relevant. Often term credit is absent because 27 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD of uncertranty, both macroeconomic and structural. If high inflation has been a culprit in the past, convincing demonstrations of a longstanding commitment to low inflation is important. Although dollarization (or adoption of some other currency) is one way out of this dilemma, it can create additional problemns to the extent that the country is not an opti- mal currency area with its partner. Another solution, which also can help ensure the quality of regulatory oversight, is fixed and long terms for the central bank governor, ending the ability of finance or prime ministers to remove them without a solid majority of parliament. The development of long-term suppliers of finance-insurance and contractual savings institutions-also will contribute to a deepening of that end of the market, as it has in Chile, without costly distortions. Markets with poor services can benefit from competition. If there is still a significant (20 percent or more) share of the banking sector in state hands, further privatization will help in this regard. Limiting the state banks' role is also shown to increase nonbank financial sector develop- ment, which will improve competition at short and long ends of the fi- nancial market. These more sophisticated financial systems will retain many financial services on-shore, but will also rely on the international market for risk-spreading and for more exotic services. Technology of credit-scoring and credit information can be adopted to help improve the reach of the financial system and the access of small entrepreneurs to it (chapter 1). The incentive conditions and the abil- ity of the authorities to supervise intermediaries effectively can be greatly enhanced in this rather sophisticated environment by relying on care- fully designed requirements that have the effect of bringing additional private sector monitors into the picture. The Next Generation of Research T HIS REPORT REPRESENTS THE CULMINATION OF ONE generation of research on the financial sector, not the first generation, but perhaps the first that has been systematically based on statistical data from across the world. The research findings provide "first-order" solutions to policymakers: overall guiding principles and a sense of strategy. It also highlights key policy issues for the next generation of research. In many cases, the first-order solution needs further amplification and specification beyond overall principles. For example, given the principles of incentive-based regulation from chapter 2, which particular aspects of bank regulation and supervision 28 OVERVIEW AND SUMMARY deserve greater priority at different stages of development? Or, the case for Moving beyond general reducing state ownership in many countries is clear, but how far should principles- authorities go and how quickly? And given the dangers associated with bank privatization, what are the lessons on how to do this process? Al- though research has begun in this area, it comes too soon after the privatizations to provide definitive answers on the long-run effects. Also, although a basic approach to bank restructuring is proposed in this vol- ume, a more systematic exploration of the links between bank and enter- prise restructuring, informed by case studies of systemic crisis countries, would help to guide authorities' decisions in a crisis. Another area of relative ignorance is how corporate governance and ownership in the financial sector affects reform strategies. When insid- ers or "oligarchs" control banks and other important intermediaries, they may be able to so influence, or even seize control of, the regula- tory apparatus that effective oversight is nonexistent. Although many accept that "one size does not fit all" in the reform process, coming up with practical rules and guidelines for authorities to know when it is safe to proceed along different reform paths is an important priority. Case studies of bank restructuring episodes will likely yield useful les- sons in general, but especially in this area, such as by highlighting the fate of different approaches to preventing excessive concentrations in ownership and control. Our discussion of foreign entry also reveals a range of wider ques- tions about the shifting patterns of financial firm ownership and own- ership concentration that need more in-depth research. And, though financial repression is almost a thing of the past in most countries, taxation of finance is still a pressing issue awaiting a synthesis, whether in regard to novel transactions taxes, to international tax competition, or to other aspects. Ongoing developments in e-finance promise to change the financial and monitoring the effects landscape in emerging and mature markets. The likely decline in the of e-finance cost of entering foreign markets may greatly increase the extent to which residents of almost all economies "import" their financial services. This rapidly evolving area needs to be monitored to identify policy problems, options, and solutions. Policymakers will certainly want to know how it will affect credit to small and medium-scale enterprises, though there will be many other effects that also need to be studied. And while e- finance may improve long-run stability, in the near term the increase in competition could have destabilizing consequences. There is already a 29 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD demand to know how countries are handling these pressures and how they are regulating "e-banks" and electronic exchanges. The trends noted or urged here-better infrastructure, improved in- centives, less state ownership, and a more receptive view to importing financial services-will all surely contribute directly or indirectly to a con- siderable expansion in the role of nonbank intermediaries and capital markets. How to regulate efficiently these markets to contain systemic risk could be the key research question of the next few years. The last several years have seen impressive leaps in our understanding of the importance of the financial sector in development and in the knowl- edge base for many key issues, but there is still much to be learned. 30 CHAPTER ONE Making Finance Effective 'Money is the greatest factor in life and the most ill-used. People don 't know how to tend it, how to manure it, how to water it, how to make it grow. " Spoken by Margayya, the "financial expert," in the eponymous novel by R.K. Narayan (1952) M UCH OF THE RENEWED FOCUS IN RECENT Economies are becoming years on the financial systems of developing more finance intensive countries reflects the rapid and often spectacular deepening in the scale and complexity of the the financial systems of advanced economies. This deepening suggests that the nature of contemporary economic progress may be more finance-intensive than previously thought, and that policymakers in developing countries may need to pay more attention to ensuring that their countries' financial systems can and do function effectively. For the good reasons reviewed below, policymakers around the world have now made financial strengthening a priority: everybody seems to want to build deeper, more sophisticated financial systems in the ex- pectation that this will contribute significantly to economic perfor- mance. This perspective is not uncontroversial, but against the con- trary view that finance merely follows and adapts to real economic progress, there is a solid body of empirical research strongly suggesting that improvements in financial arrangements precede and contribute to economic performance. This then raises the question of how a country can develop a more effective financial system. Is bigger always better? Is there a clear-cut preference for the shape of finance in terms of the relative importance 31 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD of different types of intermediary or market? And what of the infrastructural elements needed to support finance? These matters have become the subject of an active research debate, especially over the past decade when financial systems in transition economies had to be built essentially from scratch, requiring policymakers and scholars to go back to first principles. Contrasts between the shape and approach to structure and infrastructure in different advanced economies have become the focus of examination-paradoxically just at a time when these contrasts have begun to erode. There is a clear causal link While there is still much more to be learned from comparative analy- between finance and sis of the causes and consequences of contrasting financial sector per- development formance, recent research allows several important conclusions to be drawn now. The widespread desire to see an effectively functioningfinan- cial system is warranted by its clear causal link to growth, macroeconomic stability, and poverty reduction. Attempts to discriminate between dif- ferent structural types of external finance through a preference for bank- ing over market finance, or vice versa, are unwarranted, though, and could be counterproductive. A well-functioning Efficient functioning of all these markets in intertemporal commit- financial system ments requires a supporting infrastructure for information disclosure, requires a supporting contract enforcement, and competitive behavior. This contractual and infrastructure information infrastructure should, if anything, be biased in the direc- tion of directly protecting the interests of the external funds provider: the long-term interests of the would-be user will be poorly served by an infrastructure that gives potential providers so little protection that they withhold their funds. In addition, the infrastructure should be fashioned in such a way as to limit the exercise of market power not only in bank- ing, but also by insiders-whether in a firm or in the securities markets- against outside shareholders. To what extent all of these financial services will-or need to be- provided at home by domestic financial firms and markets, instead of being imported or supplied by foreign-owned firms, is a key question to which we return in chapter 4. Here it is worth pausing to clarify what we mean by financial devel- opment, which subsumes both institutions and functions. Starting with money itself, specialized institutions, including intermediaries, mar- kets, and agents, tend to become increasingly pervasive in an economy's financial activities, displacing bilateral arrangements. However, it is worth bearing in mind that, especially in developing economies, much 32 MAKING FINANCE EFFECTIVE of finance is provided within the family, through partnerships or unin- corporated business: Still, while finance can and does exist without specialized financial firms, our discussion is confined to organized fi- nance, that is, with funds processed, intermediated, or managed by specialized financial firms or traded in organized markets.' More important than the institutional form taken by these firms and markets are the underlying functions of finance that they perform. While the most evident financial activity relates to the transfer of funds in ex- change for goods, services, or promises of future return, it is essential to dig deeper. In fact, the bundle of institutions that make up an economys financial arrangements can be seen as providing the bulk of the economy's need for several functions deeper than that of simply trading and trans- fer (Levine 1997; cf. Merton and Bodie 2000): * Mobilizing savings (for which the outlets would otherwise be much more limited). * Allocating capital funds (notably to finance productive investment). The main functions of * Monitoring managers (so that the funds allocated will be spent as finance envisaged), * Transforming risk (reducing it through aggregation and enabling it to be carried by those more willing to bear it). Most textbooks, in addition to the focus on payments systems, dwell on the mobilization and allocation functions, but the monitoring and risk transformation functions are crucial as well. Though the financial sector has no monopoly on the economy's stock of intellectual capital, it is these deeper functions that justify characterizing the sector as func- tioning like the brain of the economy. Monitoring means that intermediaries do not merely collect infor- mation on firms and allocate loans or investments to them, but also continue to keep track of the recipients' activity and to exert corporate control, whether by enforcing covenants on existing contracts or ulti- mately by withdrawing or not renewing their financing. These activities are valuable precisely because information is difficult to acquire and costly to verify. In this way, intermediaries serve as "delegated monitors" (Dia- mond 1984), without which it would be difficult to separate firm own- ership and management. Risk transfer and mitigation likewise can be underrated; the variety of the associated financial instruments and the fact that they occasion- ally backfire often results in this function receiving less weight than it 33 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD deserves. Some risks can be alleviated simply by access to liquidity. Thus, entrepreneurs with access to liquid savings-their own and oth- ers-may be more willing to undertake riskier but high-yielding projects that raise growth, or investors may be willing to finance a project (or country) if they know that they can get out of it by selling, without their action so driving down the price that the option ceases to have value. Specific risks that all face-from typhoons or El Nin1o events for farmers, to technological shifts for e-entrepreneurs-also can be eased by sharing the risks with investors. Specialized instruments are invented regularly to unbundle and repackage various risks. How Finance Helps W T>E SHOULD THUS SEEK THE MAJOR CONTRIBUTIONS OF finance to economic performance in three dimensions. The first, and likely the most important, is whether there is an overall contribution to long-term average economic growth; second, whether it contributes additionally to poverty reduction; and third, whether finance succeeds in stabilizing economic activity and incomes. In all three dimensions, recent research findings suggest an unambiguously Financial development positive role for the formal institutions of finance. causes growth Almost regardless of how we measure financial development, we can see a cross-country association between it and the level of per capita income (figure 1.1). Association, however, does not prove causality and, as the charts show, there is even a very wide variation between the level of financial development between countries at comparable income lev- els, and this variation persists over time (figure 1.2). Nevertheless, over the past several years, the hypothesis that the relation is a causal one has consistently survived a testing series of econometric probes. Formal empirical exploration of this issue dates back over 30 years, and there has been a steady accumulation of evidence.2 Possibly the most striking basic indication that the relationship is one of causality is the fact-evident in figure 2 of the overview-that the level of financial development back in 1960 can help to predict subsequent economic growth even after account is taken of other known determinants of growth (including the catch-up effect of a low initial level of per capita income and the 1960 level of school enrollment). First displayed by King and Levine (1993a), for economic growth up to 1989, this predictive power has continued to be present as growth data for subsequent years is added. 34 MAKING FINANCE EFFECTIVE Figure 1.1 Financial development and per capita income Stock market turnover Liquid liabilities as percent of GDP Percent Percent 80 100 - 60 80s I 60- 0 ~~~~~~~~~~~~0- Low Low Upper High Low Low Upper High income middle middle income income middle middle income income income income income Government bond capitalization Private bond capitalization Percent Percent 60 - 60 40 - 40 - 20 - - 4 20 f 0 inoeI0 j Low Low Upper High Low Low Upper High income middle middle income income middle middle income income income income income Note: This figure represents the average of available dates in the 1 990s for each of 87 countries. Increasing income is generally The vertical bar shows the interquartile range; the financial depth of 50 percent of the countries at associated with greater financial each stage of development lie within this range. The median is shown as a horizontal bar. Data for depth. Stock market turnover, the bond market in low-income countries is available only for China and India. Source: BDL database. liquid liabilities of banks and near-banks, and bond capitalization are also generally Perhaps the most persuasive of the more recent studies (Levine, Loayza, associated with greater financial and Beck 2000) uses a richer data set for the period 1960-95 to make a depth. more comprehensive assessment in particular of the key issue: could it be that the process of economic growth itself feeds back on financial develop- ment, rather than the other way around? Some aspects of the financial 35 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD Figure 1.2 Financial development over time Bank credit to private sector as percent of GDP 80 I Vertical bar= interquartile range Mean Median 40- 20 - 1960 1965 1970 1975 1980 1985 1990 1995 The ratio of bank credit to the Note: The vertical bar shows the interquartile range; the financial depth of 50 percent of the private sector is up relative to countries at each date lie within this range. The median and mean are shown as upper and lower GDP, but the variance has horizontal bars. widened. sector clearly are determined prior to recent growth, and what Levine, Loayza, and Beck show, in essence, is that each country's level of financial development can be partly predicted by one such prior factor, namely the origin of its legal system (that is, which of the British, French, German, or Scandinavian traditions it is based on).' As we will see, this is not the only role for legal factors in our discussion, but in the present context they provide the essential econometric instrument. Levine, Loayza, and Beck go on to show that the predicted level of financial development is also correlated with long-term growth (even after also controlling for other standard determinants of growth), thereby seeming to rule out the idea that the finance-development link is all or mostly reverse causality. Levine, Loayza, and Beck's favored measure of financial development is the ratio of bank (and near-bank) credit to the private sector, expressed as a share of GDP. The emphasis on the private sector reflects the fact that credit to government does not much involve the functions of allo- cation, monitoring, and risk management. Thus it is not just a measure of financial depth, such as the ratio of money to GDP. Nor is it exactly a measure of the private sector's role in the allocation of credit; a better measure of that is the ratio of the central bank's to other banks' assets. It is worth noting that both of these alternative measures also correlate with GDP growth, and both survive the test for reverse causality. And 36 MAKING FINANCE EFFECTIVE similar results are available for measures of stock market capitalization and liquidity, as discussed below. Even higher than would be predicted by a naive simple regression of growth rates on financial development, the size of the estimated effect is substantial (figure 1.3): a doubling of the ratio of private credit (say, from 19 percent of GDP to the sample average of 38) is associated with an average long-term growth rate almost 2 percentage points higher (box 1.1). To be sure, neither this nor other financial sector ratios are policy vari- ables, and the healthy development of the sector depends on the quality of the infrastructure and incentive environment in which it functions.4 It is through its support of growth that financial development has its Finance-led growth is strongest effect on improving the living standards of the poor. Some argue, pro-poor- however, that the services of the formal financial system only benefit the rich, and even suggest that there may be a price to pay for finance-supported growth in the form of a worsening of the income-distribution in financially developed economies. Nevertheless, available empirical evidence is against any such tradeoff: on the contrary, measures of financial development are, if anything, positively (albeit weakly) correlated with the share of the bottom quintile of the income distribution. Note that poverty in a country is deter- mined by the numbers of the poor and their income. Faster growth affects both. The recent literature on the interrelation between inequality and growth points to the importance of wealth inequality in dampening growth and to Figure 1.3 Naive and modeled impact of financial development on growth Average GDP growth 1960-95 (percent per annum) * Actual Model 4- Naive 0 * -4 - 0 2 4 6 Private credit as percentage of GDP (log) Source: Levine. Loayza, and Beck (2000). 37 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD Box 1.1 Using regression coefficients to infer policy effects IT IS TEMPTING TO USE ESTIMATED REGRESSION this can considerably alter the measured impact, of- coefficients to project what might happen in a country ten-though not always-reducing it. if one of the explanatory variables were to be changed The second issue is also very important. Indeed, by policy. To begin with, one needs to recognize the it is clear that the private credit variable is only a limitations of regression analysis; the hazards of noisy proxy for a multidimensional but unmeasured im- and incomplete data and the probability that pact of financial intermediation on productivity. It alternative modeling specifications can alter the works in the regressions, because it tends to be cor- quantitative results. Even when we have satisfied related with the other dimensions, but if it becomes ourselves that the selected regression is as good as we a focus of govemment policy, the traditional correla- can hope for in the current state of knowledge, three tion with the unobserved, omitted variables will cer- further tests must be satisfied before we use the tainly break down, and the hoped-for impact on estimated relationship to predict the impact of policy. growth will not occur. (An analogous problem is well First, the relationship must be free of endogeneity known in the theory of monetary stabilization, where bias. Second, there must not be relevant omitted it is known as "Goodhart's Law.") In a nutshell, sim- variables that are correlated with the variable being ply boosting credit growth is no guarantee of healthy manipulated. Third, the variable really must be con- long-term output growth. trollable.' Controllability may be more of an issue when it Take the relationship discussed in the text: pri- comes to the institutional factors discussed later in vate credit -e growth. At least the first two problems the chapter and elsewhere. Do we really have reliable are clearly relevant and have to be navigated. ways of measurably improving legality, say, or the The first issue of endogeneity (loosely referred to quality of administration? Here the direction of ap- in the text as "reverse causality) is what has been propriate policy change may be easier to identify than handled in the cited literature by means ofinstruments; its likely quantitative impact. L. Controllability does not imply an endogeneity bias unless the variable has been manipulated by policy in response to distur- bances in the estimnated relationship. a feedback from growth in the direction of reducing inequality. Analysis by Li, Squire, and Zou (1998) of data on inequality in 49 countries suggests that financial development is a positive catalyst in both of these relation- ships. It is statistically associated across countries not only with higher growth whatever the level of wealth inequality, but also with lower income inequal- ity (as measured either by the Gini coefficient, or by the share of the top quintile) whatever the rate of growth. Further research is needed to see if these results are valid over time, as well as across countries. The results are not implausible, however. For example, even having access to secure forms of savings can protect poor farmers from the 38 MAKING FINANCE EFFECTIVE various idiosyncratic shocks that they face, reducing the likelihood that a bad year will put them into the poverty statistics, and access to other financial services can allow them to adopt more advanced technologies. What of the risk-reduction function? At the microeconomic level, a -and generally a wider range of financial instruments, including insurance contracts, can stabilizing fore- pool risk, as well as shifting it to those more willing to bear it.' And it seems from recent research that financial development also tends to re- duce aggregate economic volatility. For example, Easterly, Islam, and Stiglitz (2001) find the level of financial development (here measured again by the private credit indicator discussed above) to be a strong and significant explanatory factor in a regression explaining the output growth volatility of some 60 countries. A doubling of private credit from 20 percent of GDP to 40 percent is predicted in this regression to reduce the standard deviation of growth from 4 to 3 percent per annum (figure 1.4). Interestingly this improvement is not sustained with further finan- cial deepening: indeed, the authors' estimates suggest that very high values of the private credit measure of financial development could be associated with higher volatility of output growth, though the data in this range are sparse. Another warning sign fromn this study is that vola- tility in monetary aggregates is also associated with output volatility. Figure 1.4 Financial depth and macroeconomic volatility Standard deviation of growth (predicted) 0.3 - 0.2 - . * The deeper the financial system, .~~4 the smaller is economic volatility 0.1 . . ( Iexcept perhaps where the share 0.1 -. ,, ,, .+ .... * of credit in GDP is very high). o 7~~- _4 *..... - 4 _ * 4 4* :4 O- . 0 50 100 150 200 Credit to the private sector/GDP (percent) Note: The figure shows the data and fitted value in a regression of the standard deviation of annual GDP growth on financial depth (conditional on other control variables). Source: Based on Easterly, Islam, and Stiglitz (2001). 39 FLNANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD -but can amplify Finance is better at protecting against some sorts of shock than oth- inflationary shocks in ers. As shown by Beck, Lundberg, and Majnoni (2001), financial low- and middle-income development (measured with the private credit indicator) insulates out- countries put growth from terms of trade shocks, but it actually seems to mag- nify the impact of inflationary shocks on output volatility in low- and middle-income countries. Perhaps it is not surprising that inflationary shocks can matter more in a more deeply monetized economy, but this, too, is a warning sign that expansion of finance is not risk-free. Indeed, deeper finance without the institutional and incentive features recommended in this report can lead to a poor handling or even mag- nification of risk, rather than its mitigation.6 The aggregate empirical evidence thus points to financial develop- ment having an unambiguous long-term growth effect, and also to be stabilizing and pro-poor. The finance-growth link is especially well sup- ported by a range of different methodologies, but what mechanisms are involved, and can they be strengthened by judicious policy design? Furthermore, are there risks in seeing financial deepening as a quick fix or short-term engine of growth? Further empirical exploration by Beck, Levine, and Loayza (2000) has helped to pinpoint the most likely channels through which finance contrib- utes to long-term growth. They show that financial development is not reli- ably correlated with either national savings ratios or with capital deepening (figure 1.5: where the arrow is almost vertical rather than almost horizontal). Figure 1.5 Relative contribution of financial development to productivity and capital intensity Productivity Y |Path of financial \ \ \ \ development It's through productivity-not volume-of investment that y/l finance helps growth. YA Y/A Output levels I -v/l. (isoquants) Capital intensity kll Source: Based on the econometric findings of Levine, Loayza, and Beck (2000). 40 MAKING FINANCE EFFECTIVE Therefore, the contribution offinance to long-term growth is to improve the economys totalfactor productivity rather than the quantity of capitaL This meshes well with the finding of Bandiera and others (2000- based on a detailed examination of the multidimensional process of fi- nancial liberalization in eight countries over a quarter century) that liberalizing reforms aimed at financial development do not reliably in- crease aggregate saving. On the contrary, the indications are that liberal- ization overall, and in particular those elements that relax liquidity con- straints, may be associated with a fall in saving. Even a lower overall rate of saving is quite compatible with more rapid growth if it comes with an improved efficiency in the allocation of investable funds. Although the role of finance in contributing to growth comes through More finance means its contribution to productivity rather than the quantity of capital, more external funding for more developed financial systems do make external finance available to more finns and in particular firms, and specifically tend to favor economic sectors where, for one sectors reason or another, firms need to call on outside finance. Thus, when Demirguc,-Kunt and Maksimovic (1998) compared the actual growth rate of several thousand firms from 30 countries with each firm's esti- mated capacity to finance long-term growth from internal resources, they found that a greater proportion of the firms in financially devel- oped economies were growing faster than this benchmark. This suggests that financial development is in this sense associated with wider access to external finance. Likewise, looking at the aggregate financing of firms in 36 manufacturing sectors in more than 40 countries, Rajan and Zingales (1998) found that it was the economic sectors that, based on U.S. experience, need to rely most on external finance that grow more rapidly in more financially developed economies. If specialized financial firms are good at monitoring the users of funds, more reliance on external finance could represent an improvement. In themselves, however, these firm-level and sectoral findings do not neces- sarily point to a greater efficiency in the allocation of investable resources. After all, efficiency improvements may require diverting finance away from certain firms or sectors. A recent study by Wurgler (2000), also using sectoral data, goes some way to closing the circle by showing that sectoral investment is more responsive to sectoral output growth in fi- nancially developed economies. Put another way, in financially less- developed economies, an output slump in a certain sector is less likely to result in investment being cut back in that sector, and vice versa. Evi- dendy this, too, is an imperfect measure of allocative efficiency, but it does cast additional light on the processes at work, which apparently 41 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD include the mobilization of-though not necessarily an increase in- savings, reallocation of investable funds, and increased reliance on exter- nal providers of finance to firms, and as such to more external monitor- ing of firms' managers. Finance works in Finance also impacts growth positively over shorter periods. Levine, the short run too- Loayza, and Beck slice their 35 years of data into 7 equal subperiods and find that the correlation of growth rates with the level of financial devel- opment is still as high as it is over the longer term. Indeed, if we shorten the period even further, we can find even stronger correlations between private credit expansion and economic growth-but these can be mis- leading. It is, after all, necessary to distinguish between (a) sustainable growth based on steady productivity gains helped by shrewd allocation of capital resources and monitoring of managers, and (b) transitory growth based on unsustainable rates of borrowing. but beware: bigger is not This alerts us to a need for caution in pushing for credit expansion as necessarily better a way of achieving finance-driven growth. Bigger is not necessarily bet- ter. To be sure, it is a distinguishing mark of the high-income, advanced economies that their financial systems are large in terms of * The amount of funds intermediated and processed. * The number and range of firms they embody and the services that they provide. * The economic resources they employ. The econometric results we have described suggest that the associa- tion is to some extent causal, but it is almost obvious that a headlong rush by developing economies to emulate the scale of advanced financial systems is unwise and potentially costly. For example, attempts to make a dash for financial depth (to improve economic growth through invert- ing the finance-growth equation) can and have misfired badly: * Engineering too rapid a rapid growth in domestic credit leads to inflation and depreciation or to institutional insolvency (getting this right is partly a matter of macroeconomic stabilization policy- some of the biggest failures here have been associated with inflow surges of foreign capital, as discussed in chapter 4). * Creation of publicly owned banks to force the pace of interme- diation may instead stifle the creation of financial capacity (see chapter 3). 42 MAKING FINANCE EFFECTIVE Box 1.2 Finance as an export sector THERE IS A VALID ALTERNATIVE PERSPECTIVE HERE finance to concentrate in a small number of centers, that has been seen as relevant, especially for some each reflecting a pool of liquidity and expertise, and smaller economies. This is to see finance as a several regional financial centers have benefited from potentially important export sector. To some extent the existence of repressed financial systems in their a successful export finance business can be expected geographical neighborhood. (Furthermore, routine as a spin-off from effective domestic finance. The back-office financial services can be exported without United Kingdom's huge net export earnings from requiring such a high systemic investment in human financial services reflect the legacy of a quarter- resources.) Unfortunately, few have been successful. millennium of technological leadership in finance, Many countries attempting to develop finance as an built on the Dutch experience in the late 17th century export business have not developed the requisite legal, and placed initially at the service of government war regulatory, and supervisory structures and have, finance. Technological and human capital instead, sought to employ aggressive tax or regulatory sophistication has also helped Hong Kong and competition. What were envisaged as centers of Singapore achieve comparable roles in their region. financial expertise have sometimes become little more There is some natural tendency for international than centers for money laundering. Protection of the financial services sector as an "infant industry" can lead to excess costs and poor services (box 1.2). Instead, the policy lessons must be derived from closer observation of Policymakers should focus the processes by which the financial systems of advanced economies have on the effectiveness of evolved to provide solutions to the financial requirements of firms- financial systems and households-in increasingly complex economic environments, thereby providing the platform for further productivity advances. To a large extent, these processes have been market-driven, and many of them occurred in periods when there was little direct government activism in financial markets. Governments in developing countries that want to build on this success can best do so by responding to market needs-not indeed to the particular needs of individual market participants, but to the needs of market functioning overall. In other words, the aim should be not to try to engineer directly an expansion of the financial sector, but to adopt policies that enable financial system participants to deliver the services in which they specialize with the maximum effectiveness, and to ensure in particular that the deep functions most needed by each economy that can be provided by finance are adequately catered to. 43 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD Since developing these services involves considerable externalities and network effects, a passive stance is not enough (cf. Stiglitz, 1994). Policymakers must work with the market to ensure optimal financial devel- opment, both helping to coordinate the development of interlinked mar- ket structures, and also creating the necessary infrastructure for finance. We look in turn at structures and infrastructure. Structure Debt and equity-the basic HIS QUESTION OF HOW TO CHOOSE THE OVERALL DESIGN structural elements of a financial system emerged suddenly and acutely with the collapse of the planned economy, and the urgent requirement to create a new structure for finance in more than a score of transition economies. Almost immediately, a latent debate between the merits of intermediated and nonintermediated "market" finance came to the fore. This debate has taught us much about the way in which financial systems work, in particular how apparently different institutional structures can perform in apparently quite different ways, but with similar efficiency, the same deep economic functions. Though the premises of the debate-that bank-based financial sys- tems and market-based systems can be unequivocally ranked in their ability to deliver financial services needed for growth and prosperity- seems, as we will see, to be a false one, it is certainly true that the institu- tional structures surrounding banks and securities markets have often and for long periods evolved quite differently in different countries. How are the funding needs of a venture to be externally financed when they are too large to be provided by the promoter themselves? External providers of finance need to satisfy themselves in advance that the returns are commensurate with the risks involved and to continue keeping an eye on things thereafter. One financing strategy is for providers and users of funds to rely on a bank: a more or less large specialist intermediary that takes the risk of financing the venture on its own books, pools the risk from many ven- tures, and achieves economies of scale by avoiding duplication in ap- praisal and monitoring. The bank in turn is kept on a tight leash through its reliance on short-term financing from a large number of depositors, which also has the advantage of giving the depositors liquidity. Bank financing clearly makes sense when information about creditworthiness 44 MAKING FINANCE EFFECTIVE is easy to interpret but costly to acquire-as for example in a mature industry; the depositors have no reason to disagree with the bank management's information-based judgments. An alternative financing strategy is for financial claims on the venture to be sold directly to fund providers. If there is a lot of disagreement on the prospects of the venture, this might be a better way of enabling fin- anciers to be matched to the ventures they believe in. Selling financial claims on the open market, where there are both optimists and pessi- mists among the suppliers of funds, might be a better bet for those seek- ing finance for innovation than trying to rely on the judgment of a mono- lithic intermediary (Allen and Gale 2000). The liquidity of these claims is enhanced by having them listed on an organized securities exchange; without such liquidity, the pool of open market investors will be limited. Formal financial systems in most countries are dominated by banks, but in some of the most advanced countries the ratio of stock market capitaliza- tion to banking assets is very high, and there is a general tendency for the market-to-bank ratio to increase with the level of development both over time and cross-sectionally (figure 1.6, box 1.3). Does this imply that an , ~~As countries get wealthier, the increased role for market-based finance should be a goal of policy? relative scale, activity, and This question has been extensively analyzed in recent econometric efficiency of the stock market to research, with a striking conclusion: the trend for a general increase in the banking sector all increase. Figure 1.6 Three measures of the relative development of banks and organized securities markets Bank domestic assets/market capitalization 100 _ : { * + * (a) Relative scale of assets: bank domestic assets and market 10 | + * ** * capitalization (log scale) 0.1- $200 $3,000 $40,000 GDP per capita (figure continues on following page) 45 FINANCE FOR GROWTH: P'OLICY CHOICES EN A VOLATILE WORLD Figure 1.6 (continued) Bank private assets/value traded (log) 10,000 - (b) Relative activity: bank claims 1,000 - * . * on private sector and volume 1,000 - *+ *traded (log scale) 100 10 * 1 ~ ~ ~ * ** $200 $3,000 $40,000 GDP per capita Value traded x net interest margin 0.000001 - | * (c) Relative efficiency: measured 0.00001 by the ratio of volume traded on the stock market to (the reciprocal 0.0001 ** *of) bank net interest margin 0.0001 0 * $4t < E (log scale) 0.01 0.1 - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ $200 $3,000 $40,000 GDP per capita Source: BDL database. Banking and market the share of market finance with economic development does not appear finance both support to be causal. If one takes the regression models of growth, discussed economic growth above, and adapts them by adding various measures of the market-to- bank ratio, the results consistently fail to show any statistically signifi- cant impact of these measures of structure on growth. There appears to be no effect, whether on the sectoral composition of growth or on the proportion of firms growing more rapidly than could be financed from internal resources; even bank profitability does not appear to be affected. This is the case regardless of whether the ratio one employs relates to the 46 MAKING FINANCE EFFECTIVE Box 1.3 Time, income, and inflation: stylized facts about financial depth EVEN BEFORE EXAMINING INSTITUTIONAL ' but falls by about half a percentage point for determinants of financial development and the causal every percentage point rise in the medium-term relationship between finance and growth, certain inflation rate. empirical regularities can be detected in the macroeconomic dat, likn , iaca depth X , The importance of inflation, especially high infla- macroeconomic data, linking financial depth) tion, in hindering financial development is stressed inflation, and pet capita GDP with a significant by Boyd, Levine, and Smith (forthcoming). Their Fo m alone dth time ,, rato), r regressions, which include a wider list of controls and For monetary depth (M2/GDPl ratio), regres- examine nonlinear effects of inflation, confirm the sion analysis of a pooled cross-section and time- sion analyis ofapoledcoss-sctioandime average size of the inflationary impact on financial series with some 2,700 observations covering more depth noted above. The also show that high- than 120 countries for up to 35 years from the inlti ntrie have. mhy low fian deeh BD dtbs(Bc, Demrg. un , an .evn inflation countries have much lower financial devel- BDL database (Beck, Demirgihg-Kunr, and Levine 2()00a) allows us to quantify these macroeconomic opment, but that, beyond a certain point, addltional 2000aealltowsh usand quggestia number of stylized increases in inflation have little further impact. relationships and suggests The trend increase in financial depth has not pre- vented a secular slowdown in the world rate of growth. Monetary depth This could appear paradoxical in light of the proposi- * increases by about three percentage points for tion that deeper financial systems help generate growth. every $1,00 increase in GDP per capita, . One interpretation is that technological changes over every $1,ou0 icrease in GDP per capita, time have increased the finance-intensity of growth, . and by about a quarter of a percentage point implying that deeper financial systems are now re- every year, quired to achieve the same rate of growth as before. Note: These results are the estimated long-term (cointegrating) relationship from an error-correction mechanism imposing common coefficients across countries except for a country fixed effect. Estimation method: GLS with cross-section weights. volume of assets (bank deposits, stock market capitalization) or efficiency (net interest margin, stock turnover).7 We need not conclude that the type of financing does not matter. One aspect of what seems to have happened is that firms in successful economies have found a mixture of equity market and bank develop- ment that suits their own particular financing needs and institutional structures; the higher the level of income, the more likely that mixture will be weighted toward equity. The production technology and prod- uct market conditions facing different firms certainly play a role in prompting different approaches to financing decisions. So this evidence by no means runs counter to the need for an appropriate degree of diver- sity in channels for financing in each country. Instead, the message must 47 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD be that both development of banking and of market finance help eco- nomic growth: each can complement the other. To be sure, policy has also influenced the relative importance of banks and stock markets in some countries. Often policies favoring one segment have had the effect of stifling another, with the result that the financial system has not developed the optimal range of structures for its needs. For example, in many countries, policies of restricting information or subject- ing dividends to multiple taxation stifles the development of equities and fosters relationship debt finance, particularly when interest paid is deduct- ible. Similarly, the relative repression of the Indian and Korean banking systems compared with less interventionism in the nonbank sector clearly contributed to the development of the latter in the 1 980s and 1 990s. The restriction, from the mid-1930s, preventing U.S. commercial banks from taking significant ownership stakes in nonfinancial firms helped make at least larger firms more reliant on the stock market, as had restrictions from the country's inception on geographical diversification of banks. The striking thing is that these restrictions did not prevent the U.S. financial system from adequately supporting subsequent U.S. growth. A contrasting case is that of the U.K. banks: even though public policy did not impose any comparable restrictions on their activities, they too left room for a sub- stantial contribution from the stock market to the development of the U.K. financial system and economy. The reasons for contrasting behavior of different financial institu- tions in different countries will continue to be debated. What recent research findings have established is that they matter much less than was previously thought, and that it is the financial services themselves that matter more than the form of their delivery. Indeed, the variety of the needed services goes well beyond what can be measured in the aggregate data for the scale and activity of banks and markets. Information asymmetry One reason that the dichotomy between banks and markets may not limits access to equity in help much to predict growth is that it does not closely correspond to the developing countries- dichotomy between debt and equity.8 A range of different financial in- struments is necessary to enable firms in different circumstances to ob- tain an adequate structure of their financing. Debt is the classic instru- ment that can be used to deter insiders in a successful firm from pretending that they are unable to remunerate external financiers.9 With simple debt contracts, the payment is not supposed to be conditional on the firm's performance, and default will trigger a transfer of control (whether of a collateral, or of the firm itself) to the external financiers. Provided this transfer of control can indeed be relied upon to take place (and 48 MAKING FINANCE EFFECTIVE as discussed below, this is by no means guaranteed for developing coun- tries), this gives lenders the confidence that, even if they cannot monitor the firm's performance very reliably, they will be able to move in, take control, and realize the firm's value in the event of a default. An interesting historical reflection helps confirm that verifiability of outcomes is the central issue. Some of the earliest debt-type contracts, specifically the ship-voyage (bottomry) loans of antiquity, did actually make payment partly conditional on one of the readily verifiable aspects of success inasmuch as the debtor did not have to service the loan unless and until the financed ship returned, in which case, as with the junk bonds of more recent times, they paid off handsomely. And it is important that there should be something left over to take control of: debt is much more available to firms with tangible assets (esti- mates by Demirgiiu-Kunt and Maksimovic (1999) based on the financial accounts of a large sample of listed companies in developed and developing countries indicate that replacing $100 of intangible assets with tangible will increase reliance on debt finance on average by between $36 and $51). The complexity of much of modern economic and business activity has greatly increased the variety of ways in which insiders can try to conceal firm performance. Although progress in technology, account- ing, and legal practice has also helped improve the tools of detection, on balance the asymmetry of information between user and provider of funds has not been reduced as much in developing countries as it has in advanced economies, and may have deteriorated. The problem of moni- toring limits the potential for firms to have access to outside equity, and this problem is more acute in developing countries. Indeed, access to long-term debt financing is limited in developing -and along with legal countries, even for the leading firms. Likely contributory causes are problems, limits the not only information asymmetries and general opaqueness, but also availability of long-tenn poor collateral law and weak judicial efficiency, making it hard either debt to write strong contracts or to enforce them in a court of law. Exami- nation of the financial statements even of listed companies clearly shows that the proportion of total assets financed by debt is smaller-and much smaller if we confine attention to long-term debt-in those de- veloping countries for which data are available than in the major in- dustrial countries (figure 1.7). The low average maturity of the debt issued by firms in developing countries is not wholly explained by higher inflation-though there's nothing like inflation for stifling a long-term debt market. That is not to say that these firms have substituted out- side equity. While the data do not allow us to identify outside equity 49 FINANCE FOR GROWTH: POLICY CHOICES IN A VOLATILE WORLD Figure 1.7 Average leverage of listed firms in industrial and developing economies Long-term debt to total assets Debt to total assets Norway ovm Y fYY sY Norway soa orissuvsoso sh fi Finland c-- Y'a. f mYec Finland Sweden Y'YY a mamemY,Y, o- Sweden CerTnany Y Republic of Korea Canada T 4- a - m India United States YYY YY nwYY Yao.ama Italy i Austria r fse_ France 02?SOt, Switzerland