WORLD BANK DISCUSSION PAPER NO. 428 (ha WORLD BANK WDP428 Work in progress for public discussion January 2002 Managing the Real and Fiscal Effects of Banking Crises .- -A . X~ '' Edital bI Dalliela K/in ue/bid Liite Laevum FILE COPY Recent World Bank Discussion Papers No. 351 From Universal Food Subsidies to a Self-Targeted Program: A Case Study in Tunisian Reform. Laura Tuck and Kathy Lindert No. 352 China's Urban Transport Development Strategy: Proceedings of a Symposium in Beijing, November 8-10, 1995. Edited by Stephen Stares and Liu Zhi No. 353 Telecommunications Policiesfor Sub-Saharan Africa. Mohammad A. Mustafa, Bruce Laidlaw, and Mark Brand No. 354 Saving across the World: Puzzles and Policies. Klaus Schmidt-Hebbel and Luis Serven No. 355 Agriculture and German Reunification. Ulrich E. Koester and Karen M. Brooks No. 356 Evaluating Health Projects: Lessonsfrom the Literature. Susan Stout, Alison Evans, Janet Nassim, and Laura Raney, with substantial contributions from Rudolpho Bulatao, Varun Gauri, and Timothy Johnston No. 357 Innovations and Risk Taking: The Engine of Reform in Local Government in Latin America and the Caribbean. Tim Campbell No. 358 China's Non-Bank Financial Institutions:Trust and Investment Companies. Anjali Kumar, Nicholas Lardy, William Albrecht, Terry Chuppe, Susan Selwyn, Paula Perttunen, and Tao Zhang No. 359 The Demandfor Oil Products in Developing Countries. Dermot Gately and Shane S. Streifel No. 360 Preventing Banking Sector Distress and Crises in Latin America: Proceedings of a Conference held in Washington, D.C., April 15-16, 1996. Edited by Suman K. Bery and Valeriano F. Garcia No. 361 China: Power Sector Regulation in a Socialist Market Economy. Edited by Shao Shiwei, Lu Zhengyong, Norreddine Berrah, Bernard Tenenbaum, and Zhao Jianping No. 362 The Regulation of Non-Bank Financial Institutions: The United States, the European Union, and Other Countries. Edited by Anjali Kumar with contributions by Terry Chuppe and Paula Perttunen No. 363 Fostering Sustainable Development: The Sector Investment Program. Nwanze Okidegbe No. 364 Intensified Systems of Farming in the Tropics and Subtropics. J.A. Nicholas Wallis No. 365 Innovations in Health Care Financing: Proceedings of a World Bank Conference, March 10-11, 1997. Edited by George J. Schieber No. 366 Poverty Reduction and Human Development in the Caribbean: A Cross-Country Study. Judy L. Baker No. 367 Easing Barriers to Movement of Plant Varietiesfor Agricultural Development. Edited by David Gisselquist and Jitendra Srivastava No. 368 Sri Lanka's Tea Industry: Succeeding in the Global Market. Ridwan Ali, Yusuf A. Choudhry, and Douglas W. Lister No. 369 A Commercial Bank's Microfinance Program: The Case of Hatton National Bank in Sri Lanka. Joselito S. Gallardo, Bikki K. Randhawa, and Orlando J. Sacay No. 370 Sri Lanka's Rubber Industry: Succeeding in the Global Market. Ridwan Ali, Yusuf A. Choudhry, and Douglas W. Lister No. 371 Land Reform in Ukraine: The First Five Years. Csaba Csaki and Zvi Lerman No. 373 A Poverty Profile of Cambodia. Nicholas Prescott and Menno Pradhan No. 374 Macroeconomic Reform in China: Laying the Foundation for a Socialist Economy. Jiwei Lou No. 375 Design of Social Funds: Participation, Demand Orientation, and Local Organizational Capacity. Deepa Narayan and Katrinka Ebbe No. 376 Poverty, Social Services, and Safety Nets in Vietnam. Nicholas Prescott No. 377 Mobilizing Domestic Capital Marketsfor Infrastructure Financing: International Experience and Lessonsfor China. Anjali Kumar, R. David Gray, Mangesh Hoskote, Stephan von Klaudy, and Jeff Ruster No. 378 Trends in Financing Regional Expenditures in Transition Economies: The Case of Ukraine. Nina Bubnova and Lucan Way No. 379 Empowering Small Enterprises in Zimbabwe. Kapil Kapoor, Doris Mugwara, and Isaac Chidavaenzi No. 380 India's Public Distribution System: A National and International Perspective. R. Radhakrishna and K. Subbarao, with S. Indrakant and C. Ravi No. 381 Market-Based Instrumentsfor Environmental Policymaking in Latin America and the Caribbean: Lessonsfrom Eleven Countries. Richard M. Huber, Jack Ruitenbeek, and Ronaldo Ser6a da Motta. No. 382 Public Expenditure Refor"m under Adjustment Lending: Lessonsfrom World Bank Experiences. Jeff Huther, Sandra Roberts, and Anwar Shah No. 383 Competitiveness and Employment: A Frameworkfor Rural Development in Poland. Garry Christensen and Richard Lacroix (Continued on the inside back cover) WORLD BANK DISCUSSION PAPER NO. 428 Managing the Real and Fiscal Effects of Banking Crises Edited by Daniela Klingebiel Luc Laeven T he World Bank Washington, D.C. Copyright i) 2002 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing January 2002 1-234040302 Discussion Papers present results of country analysis or research that are circulated to encourage discussion and comment within the development community. The typescript of this paper therefore has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) 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 World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. The World Bank encourages dissemination of its work and will normally grant permission promptly. Permission to photocopy items for internal or personal use, for the internal or personal use of specific clients, or for educational classroom use, is granted by the World Bank, provided that the appropriate fee is paid directly to Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, U.S.A., telephone 978-750-8400, fax 978-750-4470. Please contact the Copyright Clearance Center before photocopying items. For permission to reprint individual articles or chapters, please fax your request with complete information to the Republication Department, Copyright Clearance Center, fax 978-750-4470. All other queries on rights and licenses should be addressed to the World Bank at the address above or faxed to 202-522-2422. ISBN: 0-8213-5056-0 ISSN: 0259-210X Daniela Klingebiel is a Senior Financial Economist at the World Bank. Luc Laeven is a Financial Economist at the World Bank Cover photo: Bank Polanska Kasa Opieki SA Library of Congress Cataloging-in-Publication Data has been applied for. iii Contents Foreword v Abstract vii Financial Restructuring in Banking and Corporate Sector Crises: Which Policies to Pursue? Stqn Claessens, Daniela Klingebiel, and Luc Laeven 1 Controlling the Fiscal Costs of Banking Crises Patrick Honohan and Daniela Klingebiel 15 Episodes of Systemic and Borderline Banking Crises Gerard Caprio and Daniela Klingeblel 31 v Foreword In recent decades many countries have These papers are not intended to reflect the Bank experienced systemic banking crises requiring Group's policies, but rather to stimulate debate in major restructurings of their financial systems. and solicit views from the development community These restructurings have often had high fiscal at large. costs, with budget outlays sometimes exceeding While the papers in this volume were motivated 50 percent of GDP. The recent East Asian crisis by events that took place during the East Asian spurred a debate on policies needed to restore crisis, they also draw on experiences from other financial stability and avert and mitigate future regions. Although many questions remain to be financial crises. Managing and resolving a financial answered, this volume contributes to the literature crisis is a complex undertaking-and one that raises by providing an overview of the lessons learned important questions about government's role. To from past government policies aimed at managing advance the dialogue on these issues, World Bank and resolving financial crises. The volume will be Group staff have prepared a number of papers, of particular interest to policymakers involved with three of which are presented in this volume. financial and corporate sector reform. Cesare Calari Vice President, Financial Sector The World Bank vii Abstract T his volume provides two recent analyses, In the second chapter Patrick Honohan spurred by the recent East Asian crisis, of and Daniela Klingebiel use cross-country T government responses to financial distress. evidence to determine whether specific crisis It also presents a comprehensive database on containment and resolution policies systematically systemic and borderline banking crises. influence the fiscal costs of resolving a crisis. In the first chapter Stijn Claessens, Daniela The authors find that accommodating policies- Klingebiel, and Luc Laeven review the tradeoffs such as blanket deposit guarantees, open- involved in public policies for systemic financial and ended liquidity support, repeated (and so corporate sector restructuring. The authors find that partial) recapitalizations, debtor bailouts, and consistent policies are crucial for success, though regulatory forbearance-significantly increase such consistency is often missing. This consistency fiscal costs. covers many dimensions and entails, among other The third chapter, by Gerard Caprio and things, ensuring that there are sufficient resources for Daniela Klingebiel, is a comprehensive database absorbing losses and that private agents face on 113 systemic banking crises that have occurred appropriate incentives for restructuring. The authors in 93 countries since the late 1970s. The database also find that sustainable restructuring requires deep also includes information on 50 borderline structural reforms, which typically require that (nonsystemic) banking crises in 44 countries political economy factors be addressed upfront. during the same period. Financial Restructuring in Banking and Corporate Sector Crises: Which Policies to Pursue? Stijn Claessens, Daniela Klingebiel, and Luc Laeven Stijn Claessens is professor of international finance of this chapter was presented at the National Bureau at the University ofAmsterdam and research fellow at for Economic Research conference on Management the Centre for Economic Policy Research. Daniela of Currency Crises, held in Monterey, California, Klingebiel is senior financial economist in the on 28-31 March 2001. The authors are grateful Financial Sector Strategy and Policy Department to Gerard Caprio, Jeffrey Frankel, Peter Kenen, at the World Bank. Luc Laeven is financial economist Tom Rose, and other conference participants for in the Financial Sector Strategy and Policy helpful comments, and to Ying Lin for help with the Department at the World Bank. An earlier version data. R esolving a systemic banking and corporate Characteristics of banking and crisis involves many policy choices ranging corporate crises R from macroeconomic (including monetary and fiscal policy) to microeconomic (including A systemic banking and corporate crisis is a capital adequacy rules and corporate governance situation where an economy faces large-scale requirements), with reforms varying in depth.' financial and corporate distress within a short These choices involve tradeoffs that influence the period.2 Recent examples include the crisis in amount of government resources needed to resolve Nordic countries in the early 1990s, in Mexico the crisis, the speed of recovery, and the recovery's in 1994-95, in East Asian countries after 1997, sustainability. Despite considerable analysis, these and in transition economies in the 1990s (though tradeoffs are not well known-an oversight that for transition economies, financial distress occasionally leads to conflicting policy advice and and structural problems had been longer-term larger than necessary economic costs. Even less is phenomena). Banking and corporate crises appear known about the political economy factors that to have become more common since the early make governments choose certain policies. 1980s: Caprio and Klingebiel (in this volume) This chapter reviews knowledge about the identify 93 countries that experienced a systemic tradeoffs involved in policies related to systemic financial crisis during the 1980s or 1990s (figure 1). financial and corporate restructuring. It finds that a It also appears that crises became deeper in the consistent framework is the key factor for successful 1 990s relative to earlier periods (Bordo and others restructuring-and one that is often missing. 2001). Consistency is needed in many areas and involves, . In a systemic crisis, partly as a result of a among other elements, ensuring that there are general economic slowdown and large shocks to sufficient resources for absorbing losses and that foreign exchange and interest rates, corporate and private agents face appropriate incentives for financial sectors experience a large number of restructuring. Moreover, sustainable restructuring defaults and difficulties repaying contracts on time. requires deep structural reforms, which often As a result nonperforming loans increase sharply. require addressing political economy factors This situation is often accompanied by depressed upfront. asset prices (such as equity and real estate prices) The next section provides an overview of on the heels of run-ups before the crisis, sharp banking and corporate crises. After that the chapter increases in real interest rates, and a slowdown or reviews the literature on such crises. The final section concludes. 2 We do not try to identify the exact causes of systemic I In this chapter systemic is used to refer to a crisis that is distress or determine whether currency crises are caused large relative to a national economy, not necessarily large by systemic financial distress in banks and corporations relative to the global economy or one that has other or vice versa. For such analysis, see Edwards and Frankel global spillovers. (forthcoming). 2 Managing the Real and Fiscal Effects of Banking Crises Figure 1 Frequency of systemic banking crises, 1980-98 Number of crisis episodes, by year crisis started 14 12 10 -__ - - I 8 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Note: The sample contains 93 cnsis countries; some countries expenenced more than one cnsis. Source: Caprio and Klingebiel (in this volume); authors calculations. reversal in capital flows (table 1). In countries frameworks are often deficient. Disclosure and with longer-term financial distress and other accounting rules may be weak for financial large-scale structural problems-such as several institutions and corporations. Equity and creditor transition economies-a systemic crisis may not rights may be poorly defined. And the judiciary is be accompanied by such changes in asset prices and often inefficient. There is usually also a shortage of capital flows, partly because run-ups in prices and qualified managers in the corporate and financial capital flows may not have occurred. sectors, as well as a lack of qualified domestic Developments in crisis countries highlight restructuring and insolvency specialists-partly the complicated coordination problems that arise because there may be no history of corporate and between corporations, between the corporate and financial sector restructuring. The government itself financial sectors, between the government and the may face credibility problems because it may have rest of the economy, and with respect to domestic been partly to blame for the crisis, and in general and foreign investors. In a systemic crisis the fate of faces many time consistency problems-such as an individual corporation and the best course of how to avoid large bailouts while also restarting the action for its owners and managers will depend economy. on the actions of many other corporations and These complicated coordination problems financial institutions as well as the general suggest that systemic crises are difficult to resolve. economic outlook. The financial and corporate Many observers have tried to develop best practices sectors, always closely intertwined, both need for resolving such crises. We next review that restructuring in a systemic crisis, and the actions literature. taken affect their liquidity and solvency. The government must set the rules of the game and be a Literature on bankig and prominent actor in restructuring. And investors, corporate crises domestic and foreign, will await the actions of Governments have used many approaches to try to owners, the government, labor, and others-often resolve systemic bank and corporate distress. implying a shortage of foreign and domestic capital Resolving systemic financial distress is not easy, when it is most needed. and opinions differ widely on what constitutes best A crisis and its coordination problems are practice. Many different and seemingly typically aggravated by institutional weaknesses, contradictory policy recommendations have been many of which likely contributed to the crisis in the made to limit the fiscal costs of crises and speed first place. Bankruptcy and restructuring recovery. Empirical research supporting particular Financial Restructuring in Banking and Corporate Sector Crises: Which Policies to Pursue? 3 Table 1 Patterns of systemic banking crises, various countries Percent Peak nonperforming Crisis Fiscal cost loans (share of Real change Change in Peak real Decline in real Country year (share of GDP) total loans) in GDP exchange rate interest rate asset prices Finland 1992 11.0 13 -4.6 -5.5 14.3 -34.6 Indonesia 1998 50.0 65-75 -15.4 -57.5 3.3 -78.5 Korea, Rep. of 1998 37.0 30-40 -10.6 -28.8 21.6 -45.9 Malaysia 1998 16.4 25-35 -12.7 -13.9 5.3 -79.9 Mexico 1995 19.3 30 -6.2 -39.8 24.7 -53.3 Philippines 1998 0.5 20 -0.8 -13.0 6.3 -67.2 Sweden 1991 4.0 18 -3.3 1.0 79.2 -6.8 Thailand 1998 32.8 33 -5.4 -13.7 17.2 -77.4 Source: Crisis year is the peak crisis year , from Caprino and Klingebiel (in this volume). Fiscal cost is from Honohan and Klingebiel (2000). Peak nonperforming loans is from Caprio and Klingebiel (in this volume) for Indonesia, Republic of Korea, Malaysia, the Philippines, and Thailand; Claessens, Djankov, and Klingebiel (2001) for Finland and Sweden; and Krueger and Tomell (1999) for Mexico. Real change in GDP is the percentage change in real fourth-quarter GDP in the crisis year relative to real fourth-quarter GDP the year before. Consumer price index inflation is used to calculate real GDP growth, and growth is in terms of local currency. GDP data are from the Intemational Monetary Fund s International Financial Statistics. The inflation rate is the percentage change in the consumer price index during the crisis year and is from litternational Financial Statistics. Change in exchange rate is the percentage change in the exchange rate relative to the U.S. dollar during the first quarter of the crisis year. An increase in the exchange rate indicates appreciation. Exchange rate data are from International Financial Statistics. Peak real interest rate is the peak real money market rate during the crisis year. For the Philippines the real discount rate is reported instead of the money market rate due to data unavailability. Interest rate data are from International Financial Statistics. Decline in real asset prices is the lar gest monthly drop in the stock market index during the crisis year relative to the stock market index in January of the year before. The return is in local currency and corrected for inflation. Datastream global market indexes were used for Finland, Mexico, and Sweden; Standard and Poor s and Intemational Finance Corporation global market indexes were used for the other countries. views remains limited, and most research is limited financial system. The second phase involves to individual cases. the actual financial, and to a lesser extent Sheng (1996) was the first attempt to distill operational, restructuring of financial institutions lessons from several banking crises. Caprio and and corporations. The third phase involves structural Klingebiel (1996) expanded on those lessons using reforms, including changes in laws and regulations, additional crises. The main lesson from both privatization of any nationalized financial efforts is that managing a financial crisis is much institutions and corporations, and so on. Here we different in industrial countries than in emerging discuss the containment phase, the restructuring of markets because emerging markets have weaker financial institutions, and the restructuring of institutions, crises are often larger, and other initial corporations. circumstances differ. As a result best practices from industrial countries do not easily transfer to Ccntahimen/plase developing countries. Another key lesson is that Policymakers often fail to respond effectively to there are many tradeoffs between various policies. evidence of an impending banking crisis, hoping In reviewing the literature on financial that banks and corporations will grow out of restructuring, especially in emerging markets, it their problems.3 But intervening early with a is useful to differentiate between three phases comprehensive and credible plan can avoid a of systemic restructuring. During the first phase, systemic crisis, minimize adverse effects, and limit which can be called the containment phase, the financial crisis is still unfolding. During this phase governments tend to implement policies aimed at 3 There are many political economy reasons why restoring public confidence to minimize the policymakers may not wish to act-thereby giving rise repercussions on the real sector of the loss of to a crisis-but we do not discuss them here (for such confidence by depositors and other investors in the analysis, see Haggard 2001). 4 Managing the Real and Fiscal Effects of Banking Crises overall losses (Sheng 1996). Early intervention meet capital adequacy requirements in the future- appears to be especially important in stopping requires careful government oversight and good the flow of financing to loss-making financial financial statements. But such features are often institutions and corporations and in limiting moral missing in developing countries. Instead of relying hazard in financial institutions and corporations on rehabilitation that requires good oversight and gambling for survival. data, regulators could apply a 100 percent Experience also suggests that intervention and (marginal) reserve requirement on deposit inflows closing of weak financial institutions need to be and other new liabilities, limiting weak banks' properly managed. Uncertainty among depositors ability to reallocate resources in a detrimental way. needs to be limited; otherwise the government may There are two schools of thoughts on whether have to try to resolve a loss of confidence by to use liquidity support and unlimited guarantees providing an unlimited guarantee on the liabilities during the containment phase.5 Some argue that of banks and other financial institutions. But in crisis conditions make it almost impossible to practice, ad hoc closures are more the norm and distinguish between solvent and insolvent often add to uncertainty, triggering a systemic institutions, leaving the authorities with little choice crisis. For example, in late 1997 the closing of but to extend liquidity support. Moreover, it is 16 banks in Indonesia triggered a depositor run argued that an unlimited deposit guarantee because depositors were aware that some politically preserves the payments system and helps stabilize connected banks known to be insolvent were kept institutions' financial claims while restructuring is open (Lindgren and others 2000). Similarly, the being organized and carried out (Lindgren and suspension of finance companies in Thailand in others 2000). 1997 increased uncertainty among depositors as Others argue that open-ended liquidity support well as borrowers. provides more time for insolvent institutions to Reviewing several cases, Baer and Klingebiel gamble (unsuccessfully) on resurrection, facilitates (1995) suggest that, to avoid uncertainty among continued financing of loss-making borrowers, depositors and limit their incentives to run, and allows owners and managers to engage in policymakers need to deal simultaneously with all looting. Supporters of this view also argue that a insolvent and marginally solvent institutions. government guarantee on financial institutions' Intermittent regulatory intervention makes liabilities reduces large creditors' incentives to depositors more nervous and undermines regulatory monitor financial institutions, allowing bank credibility-especially if regulators had previously managers and shareholders to continue gambling on argued that the institutions involved were solvent.4 their insolvent banks and increasing fiscal costs. Moreover, in emerging markets regulations are often They further point out that extensive guarantees weak, supervision is limited, and data on financial limit government maneuverability in allocating solvency are poor, so intervention tools need to be losses, often with the end result that government fairly simple. incurs most of the costs of the systemic crisis For example, a rehabilitation program for (Sheng 1996). undercapitalized financial institutions-which In practice, there is a tradeoff between restoring involves institutions indicating how they plan to confidence and containing fiscal costs. Evidence on these tradeoffs comes from Honohan and Klingebiel (in this volume), who show that much of the variation in the fiscal costs of 40 crises in industrial 4 Baer and Klingebiel also point out that a comprehensive and developing economies in 1980-97 can be approach places less demand on supervisory resources, explained by goverment approaches to resolving Under a piecemeal approach, insolvent and marginally liquidity crises. The authors find that governments solvent institutions continue to exist while other insolvent institutions are being closed or restructured. Marginally solvent institutions are subject to moral hazard and fraud while being unable and unwilling to 5 A third school argues that the granting of government raise additional capital. Especially in an environment guarantees is the outcome of political economy with weak supervision, comprehensive approaches are circumstances, and so is often a foregone conclusion thus more necessary. (see Dooley and Verma forthcoming). Financial Restructuring in Banking and Corporate Sector Crises: Which Policies to Pursue? 5 that provided open-ended liquidity support and value of their claims), to employees (through reduced blanket deposit guarantees incurred much higher wages) and suppliers, and to the government or the costs in resolving financial crises. They also find public (through higher taxes, lower spending, or that these costs are higher in countries with weak inflation). Here we discuss the restructuring of institutions. financial institutions; the next section discusses the Most important, Honohan and Klingebiel find restructuring of corporations. no obvious tradeoff between fiscal costs and To minimize moral hazard and strengthen subsequent economic growth (or overall output financial discipline, governments can allocate losses losses). Countries that used policies such as not only to shareholders but also to creditors and liquidity support, blanket guarantees, and large depositors who should have been monitoring regulatory forbearance did not recover faster. the banks. Governments often assume all losses Rather, liquidity support appears to make recovery through their guarantees. But there are exceptions to from a crisis longer and output losses larger-a the model of governments guaranteeing all finding confirmed by Bordo and others (2001). liabilities in an effort to restore confidence. Baer Thus it appears that the two most important and Klingebiel (1995) show that in some crises- policies during the containment phase are to limit notably in the United States (1933), Japan (1946), liquidity support and not extend guarantees. And Argentina (1980-82), and Estonia (1992)- where institutions are weak, governments may need governments have imposed losses on depositors to use simple methods in dealing with weak banks with little or no adverse macroeconomic and a loss of confidence to avoid higher fiscal consequences or flight to currency. In these cases contingencies and costs. economic recovery was rapid and financial intermediation, including household deposits, Restructuring flnanclal 11stitut,'ons was soon restored. Thus allocating losses to Once financial markets have been stabilized, the creditors or depositors will not necessarily lead to second phase of systemic restructuring involves runs on banks or end in contraction of aggregate restructuring weak financial institutions and money, credit, and output. In a related vein, Caprio corporations. Restructuring is complex because and Klingebiel's (1996) review of country cases policymakers need to take into account many issues. indicates that financial discipline is further Financial and corporate restructuring will depend strengthened when bank management-often part on the speed at which macroeconomic stability can of the problem-is changed and banks are be achieved because that determines the viability of operationally restructured. corporations, banks, and other financial institutions, Besides loss allocation, financial and corporate and more generally the reduction in overall restructuring crucially depend on the incentives uncertainty. But macroeconomic stability often under which banks and corporations operate. requires progress on financial and corporate Successful corporate debt workouts require proper restructuring, and so cannot be viewed incentives for banks and borrowers to come to the independently of the restructuring process (see negotiating table (Dado and Klingebiel 2000). The Burnside, Eichenbaum, and Rebelo forthcoming incentive framework for banks includes accounting, and Park and Lee forthcoming). classification, and provisioning rules-that is, Restructuring refers to several related processes: financial institutions need to be asked to recognizing and allocating financial losses, realistically mark their assets to market. The restructuring the financial claims of financial framework also includes laws and prudential institutions and corporations, and restructuring the regulations. Regulators should ensure that operations of financial institutions and corporations. undercapitalized financial institutions are properly Recognition involves the allocation of losses and disciplined and closed. The insolvency system associated redistribution of wealth and control. should enable financial institutions to enforce their Losses-that is, differences between the market value claims on corporations, allow for speedy financial of assets and the nominal value of liabilities held by restructuring of viable corporations, and provide for financial institutions and corporations-can be the efficient liquidation of enterprises that cannot be allocated to shareholders (through dilution), to rehabilitated. Proper incentives also mean limited depositors and creditors (by reducing the present ownership links between banks and corporations 6 Managing the Real and Fiscal Effects of Banking Crises (since otherwise the same party could end up being because they may resist government interference. both debtor and creditor). But without some support, good banks may not be Adequately capitalized financial institutions are able to provide financial intermediation to a key component of a proper incentive framework, corporations, aggravating the crisis. because financial institutions need to have sufficient loss absorption capacity to engage in sustainable Restructuring corporations corporate restructuring. In a systemic crisis, capital Providing the right incentives. The nature of a will often have to come from the government systemic crisis, and the already close links between through recapitalization. But general experience- the solvency and performance of the corporate supported by recent events in East Asia-suggests and financial sectors in normal times, make it clear that recapitalization of financial institutions needs to that bank restructuring needs to be complemented be structured and managed to limit moral hazard. In by corporate restructuring. To start corporate their analysis of 40 bank crises, Honohan and restructuring, corporations should quickly be Klingebiel (in this volume) find that repeated, triaged into operationally viable and not financially incomplete recapitalizations tend to increase the distressed corporations, operationally viable but fiscal costs of resolving a crisis. One possible financially distressed corporations, and financially explanation is that marginally capitalized banks and operationally unviable corporations. In a normal tend to engage in cosmetic corporate restructuring- restructuring of an individual case of financial such as maturity extensions or interest rate distress, private agents will make these decisions reductions on loans to nonviable corporations- and start the operational and financial rather than writing off debts. restructuring.6 But in a systemic crisis case-by-case Besides adequate capitalization, preferably by restructuring is difficult because the incentives private shareholders, banks' incentives to undertake under which agents operate are likely not corporate restructuring can be strengthened by conducive, private capital is typically limited, and linking government financing to the restructuring. coordination problems are large.7 For example, a capital support scheme in which Nevertheless, the starting point is providing additional fiscal resources are linked to corporate proper incentives for private agents to allow and restructuring through loss sharing arrangements can encourage market-based, sustainable corporate induce banks to conduct deeper restructuring. restructuring. Given that the crisis was likely partly Regardless, especially in weak institutional settings, induced by weaknesses in the environment in which limits on the actions of marginally capitalized banks the corporate sector operated, the first step for will typically be necessary. government has to be creating an enabling In principle, governments should only environment. Depending on country circumstances, capitalize or strengthen the capital base of financial this can imply undertaking corporate governance institutions with charter and franchise value. But reforms, improving bankruptcy and other apart from political economy problems, it is often restructuring frameworks, making the judicial difficult for governments to distinguish good banks system more efficient, liberalizing entry by foreign from bad. Risk sharing mechanisms with the private sector, such as cofinancing arrangements with government equity infusion (in the form of 6 Financial restructuring for corporations can take many preferred shares) when the private sector provides forms: loan reschedulings (extensions of maturities), capital, can help identify better banks. This setup lower interest rates, debt-for-equity swaps, debt still requires decent institutions to avoid misuse. forgiveness, indexing interest payments to earnings, and so on. Operational restructuring, an ongoing process, Especially in a weak institutional environiment with includes improvements in efficiency and management, limited private capital, governments may want to reductions in staff and wages, asset sales (such as a rely more on hard budget constraints on weak banks reduction in subsidiaries), enhanced marketing efforts, (such as a 100 percent marginal reserve requirement and the like, with the expectation of increased on new deposits) to prevent a large leakage of fiscal profitability and cash flow. resources, including through excessive guarantees 7 For other papers on systemic corporate restructuring, on financial institutions' liabilities. And good banks including specific case studies, see Claessens, Djankov, may need to be actively coerced to receive support, and Mody (2001). Financial Restructuring in Banking and Corporate Sector Crises: Which Policies to Pursue? 7 investors, changing the competitive framework for coordination problems and weaknesses in other the real sector, or introducing other supportive aspects of the institutional framework. Thus structural measures. In general, the political governments have created special frameworks for economy of reform suggests that a crisis can often corporate restructuring, such as the "London rules" be a time to get difficult structural reforms accepted first used in Mexico and then in several East Asian or at least initiated (Haggard 2001). countries (Indonesia, Republic of Korea, Malaysia, Most crisis countries reform the incentives for Thailand).8 The London rules involve an out-of- restructuring, though the strengths and depth of the court accord, under regular contract or commercial reforms differ (see Claessens, Djankov, and law, that all or most creditor institutions are coerced Klingebiel 2001; Dado and Klingebiel 2000; Stone to sign. With such an accord, agreements reached 2000a, b; and World Bank 2000 for different groups among most creditors can often be enforced on of crisis countries). For example, Indonesia adopted other creditors without formal judicial procedures. a new bankruptcy system to replace its pre-World Arbitration with specific deadlines-and War II Dutch code in August 1998, 12 months after penalties for failing to meet the deadlines-can also its crisis started. Similarly, Thailand's Senate be part of the accord, avoiding a formal judicial approved the Act for the Establishment of and process to resolve disputes.9 The degree of such Procedure for Bankruptcy Court, intended to enhancements to the London rules has varied increase the efficiency of judicial procedures in among countries. In East Asia the frameworks in bankruptcy cases, in February 1999, 19 months Republic of Korea, Malaysia, and Thailand were the after its crisis began. But despite the act's adoption, most conducive to out-of-court restructuring, while bankruptcies in Thailand remain infrequent and the framework in Indonesia was the least fraught with difficulties (Foley 2000). (Claessens, Djankov, and Klingebiel 2001). These Beyond fixing the environment, it can be differences appear to partly explain the variations in necessary to provide extra incentives for private the speed of restructuring in these four countries. agents to engage in (quick) corporate restructuring. The most far-reaching proposal for enhancing These incentives can involve tax, accounting, and the restructuring framework is "super-bankruptcy" other measures. Banks, for example, may be given (or "super Chapter 11"), a temporary tool that more tax relief for provisioning or restructuring allows corporate management to stay in place and loans. Corporations may be given more favorable forces debt-to-equity conversions (Stiglitz 2001). accounting relief for recognizing foreign exchange This tool can preserve firms' value as going losses. In the wake of its crisis, the Republic concems by preventing too many liquidations and of Korea adopted more favorable tax rules for keeping in place existing managers, who arguably corporate restructuring, though they ended up most often know best how to run the firms. An being misused through cosmetic rather than real important issue is when to call for a super Chapter restructuring. Some countries have offered 11-that is, when is a crisis systemic, and who has guarantees on exchange rate behavior, such as the authority to call for such a suspension of Indonesia's INDRA scheme and Mexico's FICORCA scheme; see Stone (2000a). The efficiency of such measures should be evaluated 8 The London rules are principles for corporate from various perspectives, taking into account their reorganization first proposed in the United Kingdom benefits for restructuring and public finance as well in the early 1990s. Because the rules were not designed as their possible redistributive effects. But while for systemic corporate distress, countries have tightened such measures may speed recovery, they often do them in various ways. not contribute to fundamental reforms. In any case, 9 Out-of-court negotiations and bankruptcy or other legal the general opinion is that such measures should be resolution techniques are not the only ways of dealing temporary (that is with sunset clauses). with financial distress. Economists have been proposing alternative procedures for some time, centering on versions of an asset sale or cash auction. Cash auctions Improving the frameworkfor restructuring. Even are easy to administer and do not rely on the judicial when adequate for normal times, a revamped system (Hart and others 1997). While attractive from a bankruptcy and restructuring framework might not theoretical perspective, these proposals have not had be sufficient during a systemic crisis given the recent followers except Mexico in 1998. 8 Managing the Real and Fiscal Effects of Banking Crises payments? Political economy factors should be Indonesia and Thailand's large family-controlled taken into account, because some debtors could gain conglomerates. These firms ended up receiving disproportionately from a suspension of payments. disproportionately large financing during the first To date no country has taken this approach.'° phase of the crisis while smaller firms lacked even Even with a better enabling environment, working capital (Domac and Ferri 1999). Thus it is agents will likely be unable to triage corporations crucial to choose a lead agent that ensures proper quickly and proceed with restructuring. The analysis of corporations' prospects as well as resulting debt overhang or deadlock in claims can durable operational and financial restructurings. be especially risky when institutions are weak, The main choice for the lead agent in and can greatly increase the final costs to the restructuring is between the government and the public sector of resolving the crisis. Weak banks private sector. Many approaches are possible. A may continue to lend to corporations that are "too centralized asset management corporation puts the big to fail," partly as a way of gambling for government in charge. Recapitalization of private resurrection, and so delay sustainable corporate banks puts the banks in charge. Under other models restructuring. Owners of defunct enterprises may investors and corporations can become the lead strip assets, leaving only liabilities for creditors, agent, with the government sharing the risks. Banks Even financially viable corporations may stop can work out nonperforming loans, for example, but paying promptly if faced with an insolvent banking with some stop-loss arrangements with the system. government. Or nonperforming loans can be In such cases it may be necessary in the short transferred to a number of corporate restructuring run to use hard budget constraints to limit the flow vehicles that, though state-owned, can be privately of resources to weak corporations from weak run by asset managers with incentive stakes. financial institutions or other sources. To increase Most important is that the lead agent have the credit to corporations that can actually repay and capital needed to absorb losses as well as the limit lending to weak corporations, it may also be institutional capacity, incentives, and external necessary to have temporary across-the-board enforcement mechanisms needed to effect mechanisms for certain types of borrowers (such as restructuring. Undercapitalized banks, for example, small and medium-size enterprises) or certain will not be very effective restructuring agents. And activities (such as trade financing). The need for without a working bankruptcy regime, private such blunter tools will increase with a country's agents will not be able to force recalcitrant debtors institutional weakness. Indonesia's market-based to the negotiating table-as in Indonesia and in approach to corporate restructuring, for example, Thailand, where the restructuring of Thai seems to have had little impact and probably only Petrochemical Industry took three years. led to further asset stripping. Countries often choose a mix of these approaches when dealing with a systemic crisis. Choosing a lead agent. As a next step it is often In 1995 Mexico tried both an asset management necessary for government to more directly support corporation and a more decentralized approach. corporate restructuring. As with support for the The four East Asian crisis countries (Indonesia, financial system, it is essential to restructure strong Republic of Korea, Malaysia, Thailand) all and viable corporations, and not weak ones. But all eventually used asset management corporations, too often, unviable corporations (such as those all used out-of-court systems for corporate considered too big too fail) receive support instead restructuring, and most used, after some initial of deserving, operationally viable corporations. This period, fiscal stimulus and monetary policy to foster was the case with Korea's large chaebols and with economic growth. In addition, all enhanced, to varying degrees, their basic frameworks for private sector operations, including bankruptcy and 10 While bankruptcy laws differ considerably even among corporate governance frameworks, liberalization industrial countries, there has been a general move from of foreign entry in the financial and corporate more creditor-friendly regimes that are liquidation- sectors, and so on. But success has varied with the oriented toward more debtor-friendly regimes that are intensity of these measures (Claessens, Djankov, restructuring-oriented (Westbrook 2001). and Klingebiel 2001). Financial Restructuring in Banking and Corporate Sector Crises: Which Policies to Pursue? 9 Empirical evidence on these mechanisms is institutional framework, including accounting and limited but tends to favor the decentralized model. legal rules, and on initial conditions, including A study of seven centralized approaches using the capital positions of banks and ownership links. asset management corporations found that most In Norway the government built on favorable initial did not achieve their stated objectives with conditions to attain a solid overall framework corporate restructuring (Klingebiel 2001). The for the decentralized approach. The biggest study distinguishes corporate restructuring asset improvement to the overall framework was made management corporations from bank rehabilitation in Chile, with favorable results. Poland and Hungary asset management corporations. Two of the three ranked behind Chile, though Poland improved its corporate restructuring corporations did not achieve framework much faster than Hungary. Thailand their narrow goal of expediting restructuring. Only made little progress on strengthening its framework. Sweden's successfully managed its portfolio, acting In Japan, despite many reforms to the overall in some instances as the lead agent in restructuring. framework, efforts remained blocked by large Rapid asset disposition vehicles fared ownership links. And Argentina relied solely on somewhat better, with two of four-in Spain and the public debt relief programs and did not change its United States-achieving their objectives. overall framework for restructuring. These successes suggest that asset management corporations can be effective, but only for narrowly Changing ownership structures. Just as a crisis can defined purposes of resolving insolvent and offer a window for structural reform, it can provide unviable financial institutions and selling their an opportunity to reform a country's ownership assets. But even achieving these objectives requires structures. As a direct party to the restructuring many ingredients: a type of asset that is easily process, the state often becomes the owner of liquefied (such as real estate), mostly professional defunct financial institutions and corporations. This management, political independence, a skilled development severely complicates the resolution of human resource base, appropriate funding, adequate the crisis, because government may not have the bankruptcy and foreclosure laws, good information right incentives or capacity to effect the needed and management systems, and transparent operational and financial restructuring. At the same operations and processes. time, large indirect ownership by the state of the The findings by Klingebiel (2001) on asset financial and corporate sectors provides an management corporations are corroborated by a opportunity to change ownership structures as part review of three East Asian countries (Dado 2000). of restructuring. This move can have several The centralized asset management companies in benefits. Indonesia and Republic of Korea did not appear First, the changes can correct ownership likely to achieve their narrow goal of expediting structures that contributed to the crisis and so help bank or corporate restructuring, while Malaysia's prevent future crises. To the extent, for example, was relatively successful, aided by that country's that ownership concentrated in the hands of a few strong bankruptcy system. Success has also varied families contributed to the crisis-as argued by when a mix of approaches is tried. In Mexico some for East Asia-government can try to widen neither the asset management company nor the ownership structures. enhanced restructuring framework was effective, Second, government can try to obtain possibly because fundamental reforms were lacking political support for restructuring by reallocating (Mexico's bankruptcy regime, for example, was not ownership." One option is to reprivatize financial revamped until four years after its crisis). Export-led institutions or corporations in a way that growth appears to have led Mexico's recovery after 1995 (though growth did not resolve banking problems; see Krueger and Tomeell 1999). 11 Regardless of the changes in ownership and the Dado and Klingebiel (2000) analyze relationships between debtors and creditors, the decentralized restructuring in seven countries- government may want to create a special social safety Argentina, Chile, Hungary, Japan, Norway, net for laid-off workers to help sustain political support Poland, and Thailand. They find that the success for restructuring over time. See Levinsohn, Berry, and of this approach depended on the quality of the Friedman (forthcoming) for the case of Indonesia. 10 Managing the Real and Fiscal Effects of Banking Crises redistributes ownership among the general public or from weak banks are affected by tighter regulation employees of the restructured institution. Another and supervision (BIS 1999). Given the unbalanced option is to use some of the state ownership to financial systems in East Asia-where banks endow unfunded pension obligations from a pay-as- dominate and little alternative financing was you-go system. In this way government can create available, and many banks were fragile even before ownership structures that over time will reinforce its the crisis (Claessens and Glaessner 1 997)-it is reforms. likely that, at least initially, banking weaknesses and Third, changing ownership structures can tighter regulation and supervision led to a credit introduce third parties who have better incentives crunch for East Asian corporations (Domac and and skills in restructuring individual corporations Ferri 1999). Following this initial crunch, and determining financial relief. One option is to corporations may have ended up with a debt transfer nonperforming loans to a fund jointly overhang, with a consequent need for financial owned by private and public shareholders, but with restructuring. the private stake having lower seniority. Private shareholders in the fund would then have the right Conclusion incentives when deciding on the financial viability The literature on systemic restructuring emphasizes of a corporation, but without having full formal the need for governments to actively intervene to ownership of the assets. Public resources would be overcome the many coordination problems in a provided only when all parties-creditor banks, systemic crisis and to relieve the shortage of other creditors, new private investors, the financial capital, both of which impede progress government, and the private shareholders in the with case-by-case restructuring. The core issue in fund-had reached agreement with the corporation. dealing with a systemic crisis then becomes how to resolve coordination issues while preserving or Pursuing supportive macroeconomic policies. enhancing incentives for normal, market-based Another common theme in the literature is that restructuring and transactions. Achieving both goals corporate restructuring should occur in the context requires consistent government policies, both of supportive macroeconomic policies. The right among issues and sectors, and over time. macroeconomic policies (fiscal and monetary) can The literature also stresses that fiscal and speed the recovery of overall activity and corporate monetary policies have to support the recovery output. The appropriate fiscal stance has been process in a systemic crisis. Policies must strike extensively reviewed, especially in the context of the the right balance between supporting the exchange East Asian crisis. A review by the International rate and avoiding a serious credit crunch created Monetary Fund suggests that East Asian countries' by high interest rates. Supportive policies also fiscal stance was too tight initially (Lane and others cover other dimensions, such as the strictness of 1999). The appropriate monetary stance has been capital adequacy requirements and whether an more controversial and is still being debated (see allowance should be made for automatic rollover of Drazen forthcoming and Cho and West payments by small and medium-size enterprises forthcoming), but mainly in terms of defending the during the early phases of a crisis. As extensively exchange rate. debated in the context of the East Asian crisis and An important related aspect is the effect on the earlier (for example, following Chile's 1982 crisis), corporate sector through a possible credit crunch. these supportive policies have not always been in Microeconomic-based empirical literature suggests place during systemic crises. evidence of a credit crunch early in the East Asian Especially during the containment phase of a crisis (Claessens, Djankov, and Xu 2000; Colaco, systemic crisis, but also afterward, governments Hallward-Driemeier, and Dwor-Frecaut 2000; have to balance achieving stability with aggravating Dollar and Hallward-Driemeier 2000). The crunch moral hazard. One dimension is avoiding the was likely the result of tighter capital adequacy extension of government guarantees of financial requirements and the monetary policies being institutions' liabilities, which can create moral pursued. More generally, it has been found that hazard and reduce freedom in future loss while tighter capital adequacy rules have minimal allocations. Another dimension is the closing or effects on aggregate credit provision, borrowers suspension of some financial institutions. Though it Financial Restructuring in Banking and Corporate Sector Crises: Which Policies to Pursue? 11 Consistent reform is also needed for public WVhile many of these lessons are often recapitalizations. Any public recapitalization of mentioned in the literature we reviewed, best banks must take into account the availability of practice policies are often not applied. Mistakes can fiscal resources. In several crisis countries the be made in the middle of a crisis. Afterward, it recapitalization of financial institutions with is easy to point out these inconsistencies. But government bonds did not restore public confidence even before there have been many clear cases of because limited fiscal resources were available to inconsistent financial restructuring programs. back the bonds. A related intertemporal consistency These inconsistencies usually develop because issue in any crisis is government credibility. We policymakers are trying to overcome political did not address this issue directly in this chapter, constraints, and it is hard to judge whether but ex ante consistency is a precondition for they do so in the most efficient manner. But credibility. inconsistencies can also reflect genuine differences Finally, approaches to restructuring must be of opinion among policymakers and advisers on consistent with a country's institutional capacity. what constitutes best practice-as with the need to Institutional deficiencies can rule out approaches guarantee all liabilities during the early stages of a in some countries that may be best practices in crisis. The end result is similar, in that consistency other countries. These best practices can include is often lacking. heavy reliance on a market-based approach Government efforts to restructure need to take to corporate restructuring-where banks are into account the political economy factors behind recapitalized and asked to work out debtors. But the causes of a crisis and its resolution. In this where corporate governance and financial context there might be ways to change ownership regulation and supervision are weak, such an structures in a systemic crisis so that recovery is approach may be a recipe for asset stripping or expedited and a more sustainable outcome results. looting rather than sustainable restructuring. Thus But while we lack complete understanding of emerging markets and industrial countries will systemic crises, we know even less about the need different approaches to systemic restructuring. political economy of systemic crises. References Baer, Herbert, and Daniela Klingebiel. 1995. "Systematic Risk When Depositors Bear Losses: Five Case Studies." In G. G. Kaufmian, ed., Research in Financial Services: Private and Public Policy. vol. 7. Greenwich, Conn.: JAI Press. BIS (Bank for International Settlements). 1999. "Capital Requirements and Bank Behaviour: The Impact of the Basle Accord." Working Paper. Basle Committee on Banking Supervision, Basel, Switzerland. Bordo, Michael, Barry Eichengreen, Daniela Klingebiel, and Maria Soledad Martinez-Peria. 2001. "Is the Crisis Problem Growing More Severe?" Journal of Economic Policy (April). Burnside, Craig, Martin Eichenbaum, and Sergio Rebelo. Forthcoming. "On the Fiscal Implications of Twin Crises." In Michael Dooley and Jeffrey Frankel, eds., Managing Currency Crises. Proceedings of a National Bureau for Economic Research conference. Chicago, Ill.: University of Chicago Press. Caprio, Gerard, and Daniela Klingebiel. 1996. "Bank Insolvencies: Cross-Country Experience." Policy Research Working Paper 1620. World Bank, Washington, D.C. Cho, Dongchul, and Kenneth D. West. Forthcoming. "Interest Rates and Exchange Rates in the Korean, Philippine and Thai Exchange Rate Crises." In Michael Dooley and Jeffrey Frankel, eds., Managing Currency Crises. Proceedings of a National Bureau for Economic Research conference. Chicago, Ill.: University of Chicago Press. Claessens, Stijn, and Thomas Glaessner. 1997. Are Financial Sector Weaknesses Undermining the EastAsian Miracle? A Directions in Development book. Washington, D.C.: World Bank. 12 Managing the Real and Fiscal Effects of Banking Crises Claessens, Stijn, Simeon Djankov, and Daniela Klingebiel. 2001. "Financial Restructuring in East Asia: Halfway There?" In Stijn Claessens, Simeon Djankov, and Ashoka Mody, eds., Resolution of Financial Distress. Washington, D.C.: World Bank Institute. Claessens, Stijn, Simeon Djankov, and Ashoka Mody, eds. 2001. Resolution of Financial Distress. Washington, D.C.: World Bank Institute. Claessens, Stijn, Simeon Djankov, and Lixin Colin Xu. 2000. "Corporate Performance in the East Asian Financial Crisis." The World BankResearch Observer 15 (1): 23-46. Colaco, Francis, Mary Hallward-Driemeier, and Dominique Dwor-Frecaut. 2000. "Asian Corporate Recovery: A Firm-Level Analysis." In Dominique Dwor-Frecaut, Francis Colaco, and Mary Hallward-Driemeier, eds., Asian Corporate Recovery: Findings from Firm-Level Surveys in Five Countries. Washington, D.C.: World Bank. Dado, Marinela. 2000. "Note on Centralized Asset Management Companies in Indonesia, Korea and Thailand." World Bank, Trust Funds and Cofinancing Department, Washington, D.C. Dado, Marinela, and Daniela Klingebiel. 2000. "Decentralized, Creditor-Led Corporate Restructuring: Cross-Country Experience." World Bank, Financial Sector Strategy and Policy Department, Washington, D.C. Dollar, David, and Mary Hallward-Driemeier. 2000. "Crisis, Adjustment, and Reform in Thailand's Industrial Firms." The World Bank Research Observer 15 (1): 1-22. Domac, Ilker, and Giovanni Ferri. 1999. "The Credit Crunch in East Asia: Evidence from Field Findings on Bank Behavior." World Bank, East Asia and Pacific Region, Washington, D.C. Dooley, Michael, and Jeffrey Frankel, eds. Forthcoming. Managing Currency Crises. Proceedings of a National Bureau for Economic Research conference. Chicago, Ill.: University of Chicago Press. Dooley, Michael, and Sujata Verma. Forthcoming. "Rescue Packages and Output Losses Following Crises." In Michael Dooley and Jeffrey Frankel, eds., Managing Currency Crises. Proceedings of a National Bureau for Economic Research conference. Chicago, Ill.: University of Chicago Press. Drazen, Allan. Forthcoming. "Interest Rate Defense against Speculative Attack as a Signal: A Primer." In Michael Dooley and Jeffrey Frankel, eds., Managing Currency Crises. Proceedings of a National Bureau for Economic Research conference. Chicago, Ill.: University of Chicago Press. Edwards, Sebastian, and Jeffrey Frankel. Forthcoming. Preventing Currency Crises in Emerging Markets. Proceedings of a National Bureau for Economic Research conference. Chicago, Ill.: University of Chicago Press. Eichengreen, Barry, and Andrew Rose. Forthcoming. "Does it Pay to Defend Against a Speculative Attack?" In Michael Dooley and Jeffrey Frankel, eds., Managing Currency Crises. Proceedings of a National Bureau for Economic Research conference. Chicago, Ill.: University of Chicago Press. Foley, Fritz. 2000. "Going Bust in Bangkok: Lessons from Bankruptcy Law Reform in Thailand." Harvard Business School, Cambridge, Mass. Haggard, Stephen. 2001. "The Political Economy of Financial Restructuring in East Asia." In Stijn Claessens, Simeon Djankov, and Ashoka Mody, eds., Resolution of Financial Distress. Washington, D.C.: World Bank Institute. Hart, Oliver, Rafael La Porta Drago, Florencio Lopez-de Silanes, and John Moore. 1997. "A New Bankruptcy Procedure That Uses Multiple Auctions." European Economic Review 41: 461-73. Honohan, Patrick, and Daniela Klingebiel. 2000. "Controlling the Fiscal Costs of Banking Crises." Policy Research Working Paper 2441. World Bank, Washington, D.C. Financial Restructuring in Banking and Corporate Sector Crises: Which Policies to Pursue? 13 International Monetary Fund. Various editions. International Financial Statistics. Washington, D.C. Klingebiel, Daniela. 2001. "The Role of Asset Management Companies in the Resolution of Banking Crises." In Stijn Claessens, Simeon Djankov, and Ashoka Mody, eds., Resolution of Financial Distress. Washington, D.C.: World Bank Institute. Krueger, Anne, and Aaron Tomell. 1999. "The Role of Bank Restructuring in Recovering from Crises: Mexico 1995-98." NBER Working Paper 7042. National Bureau for Economic Research, Cambridge, Mass. Lane, Timothy, Atish Ghosh, Javier Hamann, Steven Phillips, Marianne Schultze-Ghattas, and Tsidi Tsikata. 1999. "IMF-Supported Programs in Indonesia, Thailand and Korea." IMF Occasional Paper 178. International Monetary Fund, Washington, D.C. Levinsohn, James, Steven Berry, and Jed Friedman. Forthcoming. "Impacts of the Indonesian Economic Crisis: Price Changes and the Poor." In Michael Dooley and Jeffrey Frankel, eds., Managing Currency Crises. Proceedings of a National Bureau for Economic Research conference. Chicago, Ill.: University of Chicago Press. Lindgren, Carl-Johan, Gillian Garcia, and Matthew I. Saal. 1996. Bank Soundness and Macroeconomic Policy. Washington, D.C.: International Monetary Fund. Lindgren, Carl-Johan, Tomas J. T. Balinlo, Charles Enoch, Anne-Marie Gulde, Marc Quintyn, and Leslie Teo. 2000. "Financial Sector Crisis and Restructuring: Lessons From Asia." IMF Occasional Paper 188. International Monetary Fund, Washington, D.C. Park, Yung Chul, and Jong Hwa Lee. Forthcoming. "Recovery and Sustainability in East Asia." In Michael Dooley and Jeffrey Frankel, eds., Managing Currency Crises. Proceedings of a National Bureau for Economic Research conference. Chicago, Ill.: University of Chicago Press. Sheng, Andrew, ed. 1996. Bank Restructuring: Lessons from the 1980s. Washington, D.C.: World Bank. Stiglitz, Joseph. 2001. "Bankruptcy Laws: Some Basic Economic Principles." In Stijn Claessens, Simeon Djankov, and Ashoka Mody, eds., Resolution of Financial Distress. Washington, D.C.: World Bank Institute. Stone, Mark. 2000a. "The Corporate Sector Dynamics of Systemic Financial Crises." Policy Discussion Paper 00/114. International Monetary Fund, Washington, D.C. 2000b. "Large-Scale Post-Crisis Corporate Sector Restructuring." Policy Discussion Paper 00/7. International Monetary Fund, Washington, D.C. Westbrook, Jay. 2001. "Systemic Corporate Distress: A Legal Perspective." In Stijn Claessens, Simeon Djankov, and Ashoka Mody, eds., Resolution of Financial Distress. Washington, D.C.: World Bank Institute. World Bank. 2000. East Asia: Recovery and Beyond. Washington, D.C. 15 Controlling the Fiscal Costs of Banking Crises Patrick Honohan and Daniela Klingebiel Patrick Honohan is lead economist in the Beck, Gerard Caprio, Stijn Claessens,AsliDemirgu,c- Development Economics Research Group at the Kunt, Danny Leipziger, Giovanni Majnoni, Sole World Bank. Daniela Klingebiel is seniorfinancial Martinez, Eric Rosengren, David Scott, and economist in the FinancialSectorStrategy andPolicy participants at conferences at the Chicago (US.) Department at the World Bank. This chapter Federal Reserve (Bank Structure Conference) and summarizesapaper with the same namepublished as the World Bank for helpful comments, and to World Bank Policy Research Working Paper 2441 in Marinela Dado, Andrea Molinari, and Pieter van September 2000. The authors are grateful to Thorsten Oijen for excellent research assistance. In recent decades most countries-rich and poor from around the world, a sample representing all alike-have experienced systemic banking those for which data are available on both the fiscal crises requiring major-and expensive- costs of the crises and the nature of the crisis overhauls of their banking systems. Banking crises management policies pursued. not only hit government budgets with outlays that We find no evidence that accommodating have to be absorbed through higher taxes (or lower policies reduce fiscal costs. Indeed, each of the spending), they are also costly in terms of forgone accommodating measures examined-open-ended economic output. liquidity support, blanket deposit guarantees, When crises break, as typically revealed regulatory forbearance, repeated (and thus initially through audits uncovering widespread bank inadequate or partial) recapitalizations, and debtor insolvencies or through liquidity squeezes and bailout schemes-appears to significantly increase depositor withdrawals, governments are faced with the costs of banking crises. Using regression results the tasks of containment and resolution. Either of to simulate the effects of these accommodating two broad approaches can be pursued. One is policies, we find that if the countries in our sample accommodating, involving measures such as liberal had not pursued any such policies, fiscal costs liquidity support to banks with cash-flow would have averaged about 1 percent of GDP-little difficulties, guarantees to depositors and creditors more than one-tenth of what was actually spent. of financial institutions, regulatory forbearance by Moreover, there is no indication that incurring these tolerating violations of bank solvency and minimum higher costs reduced the scale of the output dips that capitalization rules, and debtor support schemes followed the crises. But things could have been that prop up bank borrowers who might otherwise worse: had every country pursued all the above default. The other approach sticks to the rules, policies, the regression results imply that fiscal requiring banks to meet standard capitalization costs would have reached more than 60 percent of requirements or face official intervention that GDP. constrains their operations. The accommodating Our interpretation of these findings is in terms approach can restore or sustain depositors' of the moral hazard created by accommodating confidence and buy time for the situation to correct policies. Our model also takes into account the itself. It is often thought that this approach saves independent role of macroeconomic shocks in taxpayers' money in the long run and limits the contributing to and revealing bank insolvencies, as wider economic costs of a crisis. But the heightened well as the fact that a bad resolution strategy can be moral hazard entailed by the accommodating more damaging when the origins of a crisis are approach can be just as costly-if not more so. primarily microeconomic. This chapter examines the empirical evidence The next section reviews the nature and extent on the two approaches. Specifically, it quantifies the of the costs of banking crises. The chapter then extent to which fiscal outlays incurred in resolving discusses the different tools for resolving crises- bank distress can be attributed to crisis management that is, the choice between strict and measures adopted by governments in the early years accommodating policies. After that we present of a crisis. It does so by analyzing some 40 crises empirical evidence on the extent to which costs are 16 Managing the Real and Fiscal Effects of Banking Crises influenced by these policy choices. The final section banking crises have occurred in 93 countries, and 50 concludes. borderline crises have arisen in 44 countries (Caprio and Klingebiel in this volume). The costs of banking crises Governments-and thus ultimately taxpayers- In the past quarter-century no type of country has have shouldered most of the direct costs of these managed to avoid costly banking crises: banking crises. These costs have been large: in our sample of system failures have been at least as prevalent in 40 countries, governments spent an average of developing and transition economies as in industrial nearly 13 percent of GDP cleaning up their financial countries. Since the late 1 970s, 113 systemic systems (figure I shows some of the higher costs in Figure 1 Fiscal costs of banking crises Percentage of GDP 0 10 20 30 40 50 Indonesia 1997 Chile 1981 ; - Thailand 1997 Uruguay 1981 ,_, Korea, Rep. of 1997 -`7 CWte d'lvore 1988 Venezuela, RB 1994 Japan 1992 _ _ Mexico 1994 *.-_ U - Malaysia 1997 - Slovenia 1992 Brazil 1994 Philippines 1983 L -- Bulgaria 1996 _IlL Ecuador 1996 . - . Czech Republic 1989 ____ Finland 1991 Hungary 1991 _ _ Senegal 1988 ___ - - __i Norway 1987 ,±-41--3': Spain 1977 _- - Paraguay 1995 Colombia 1982 Sri Lanka 1989 Malaysia 1985 Sweden 1991 Note: Shows costs of 26 crises used in our sample. Dates indicate the year each crisis began. Source: Caprio and Klingebiel (t996), Claessens, Djankov, and Klingebiel (2001). Controlling the Fiscal Costs of Banking Crises 17 the sample). Costs were even higher (just over 14 * Bond or equity injections into an insolvent percent of GDP) in developing countries. And some bank to restore its capital or make it salable to a banking crises have been far more costly: in the sound bank; this is often done by buying part of early 1980s Chile's crisis cost 41 percent of GDP, the bank's loan portfolio at face value even and Argentina's cost 55 percent. Many of the costs though the recoverable value of the loans is of East Asia's recent financial crisis-estimated at much lower. 25-50 percent of GDP for the three hardest-hit * The capitalized value of subsidized loans to countries (Indonesia, Republic of Korea, and insolvent banks or their borrowers. Thailand)-will ultimately fall on government * Payouts to depositors and other claimants, budgets. Despite their small economies, developing including foreign creditors.' countries have suffered cumulative fiscal costs of Most of our data on these costs come from more than $1 trillion. Among industrial countries, evaluations made when a crisis has been detected Japan's drawn-out banking crisis has been the and is being contained. They represent estimates costliest. of the net present value of prospective government Fiscal outlays are not the only economic costs outlays to restore banks' capital positions and of bank collapses. Indeed, to the extent that bailing to make depositors and creditors whole in cases out depositors amounts to a transfer from taxpayers where the government has extended guarantees to depositors, this is not even a net economic cost. to them. But when a government makes a bank's claimants While it is hard to obtain reliable data on the whole, its net costs tend to be correlated with the fiscal costs of banking crises, it is even harder true economic costs. For one thing, the losses to pinpoint the other dimensions of crisis costs. covered-which are caused by bad loan decisions- Attempts have been made to roughly estimate the reflect wasted investible resources. Furthermore, a additional flow economic costs, typically by government's assumption of large, unforeseen comparing actual output with some hypothetical bailout costs can destabilize fiscal accounts, "no crisis" output path. But it is extremely difficult triggering high inflation and a currency collapse- to guess what part of an output slump is caused costly in themselves-as well as adding to the by a banking crisis-a latent banking crash often deadweight cost of taxation. becomes evident only when it is triggered by an Moreover, fiscal costs do not include the exogenous economic shock that also directly costs borne by depositors and other creditors of contributed to recession. IMF (1998) offers one failed banks (in some cases) and do not take widely used approach to estimating the costs of the into account the burden imposed on depositors output dips that follow banking crises.2 Using this and borrowers by the higher interest rate spreads measure, output dips are correlated with measured that result from bad loans left on banks' balance fiscal costs and on average are of the same order of sheets. In addition, fiscal costs do not reflect the magnitude (figure 2).3 costs of granting borrowers some monopoly Examining the influence of policies on fiscal privilege or other means to improve their profits costs is of interest regardless of whether fiscal costs and so repay their loans. Finally, estimates of are a good measure of total crisis costs. But if fiscal costs do not capture the slowdown in economic activity that occurs when resources are driven out of the formal financial sector (into I These avenues do not reflect the net fiscal flows to less efficient uses) and stabilization programs are banking systems. Outside crisis times, repressed derailed. banking systems often involve a sizable flow of Estimating the fiscal costs of banking crises is resources to the government. not easy. There is no universal methodology, and 2 Hoggarth, Reis, and Saporta (forthcoming) discuss obtaining the underlying components of the alternative ways of measuring the subsequent output information required is usually problematic. Such dip. costs typically arise through: 3 If three outliers are discarded, the correlation is 0.7, * Defaults on liquidity loans made by the and a regression line implies an approximate one-to- monetary authority to a bank that proves to be one relationship between flow output costs and fiscal insolvent. costs. 18 Managing the Real and Fiscal Effects of Banking Crises Figure 2 crisis is unfolding, and those for the rehabilitation Estimated fiscal costs and output dips and restructuring phase. For both phases policy after 40 banking crises choices can be strict or more accommodating and Percentage of GDP gradual. In general, strict policies emphasize Fiscal cost decisive preventive action. A gradual approach can 60 be defended when the authorities have other ways of limiting further risk taking. 50 Con/at8iment phase In the early stages of a financial crisis, governments 40 typically implement policies aimed at restoring public confidence in the banking system to 30 * minimize repercussions on the real sector. As they struggle to contain the crisis, governments face 20 - (among others) two key strategic questions: . ; . ' . * Should open-ended liquidity support be 10 * extended to all financial institutions, including 4t * . * insolvent ones? 0 ..1,. ., , , Should blanket guarantees be provided to 0 10 20 30 40 50 60 70 depositors and creditors of financial institutions Output dip (in times of severe disruption) to stem a loss of Source: Authors calculations. confidence in the system as a whole? Open-ended liquidity support. The classic doctrine incurring fiscal costs helps reduce other crisis costs, is that central banks should abstain from providing then the policy implications would be quite banks with open-ended emergency liquidity support different. In what follows we use the IMF (1998) unless they are satisfied that the banks are viable approach to examine the influence of and oversight is adequate. Proponents of this view accommodating policies on crisis-induced output point out that governments have often used liquidity dips. There is no indication that such policies reduce support to delay crisis recognition and to avoid such dips; some accommodating measures actually intervening in de facto failed institutions. They increase them. argue that open-ended liquidity support is doomed Accommodating and strict to fail because manager and shareholder incentives shift when a financial institution becomes insolvent. policies for resolving crises Thus, unless public loans to insolvent banks are Although all banking systems are subject to conditioned on restructuring and recapitalization, ongoing supervision, awareness of emerging they prolong such institutions' ability to gamble for solvency problems typically triggers intensified resurrection, facilitate continued financing to loss- management. Starting with a diagnosis of the scope making borrowers, and allow owners and managers of the crisis-especially whether it should be to engage in looting. The alternative view considered systemic-the authorities make a series recognizes that crisis conditions make it nearly of decisions ending with actions such as closing impossible to distinguish between solvent and financial institutions, nationalizing them, insolvent institutions, and argues that a generalized liquidating them, disposing of their assets, and so crisis leaves the authorities with little choice but to on. There are many possible policy responses to extend liquidity support. banking distress, and the right decisions depend on factors such as the causes of the crisis, other Blanket deposit guarantees. A second contentious prevailing conditions, and political constraints point is whether governments should extend explicit facing the regulatory authorities. blanket guarantees to depositors and creditors It is convenient to distinguish between policies immediately after the onset of a crisis, to stem the for the short-term containment phase, while the loss of confidence in the financial system. Some Controlling the Fiscal Costs of Banking Crises 19 analysts take a strict line here as well, arguing that on what state banks are in, and that relaxing guarantees-if they are credible-reduce large regulation in response to macroeconomic downturns creditors' incentives to monitor financial can provide better ex ante risk sharing as well as institutions, providing ready funds for managers and shelter bank customers from the disruptions to shareholders to use in gambling to resurrect financial services (including credit crunches) that insolvent banks. They also point out that extensive may result from widespread bank suspensions and guarantees limit governments' maneuverability in closures. allocating future losses, with the result that they may end up absorbing most of the costs. Others Repeated recapitalizations. Instead of relying on a reason that, by extending timely and temporary flow of future profits, stock solutions immediately guarantees, the authorities can avoid the much inject capital, and are supported by the government greater fiscal and economic costs of a widespread and aimed at restoring the solvency of viable but panic, such as could be triggered or exacerbated by insolvent or marginally solvent institutions. If the closing of a few banks. recapitalization needs to be repeated, this suggests that not enough was done the first time, and thus Re/iab,iitat,on and r-es/ruc/U1179 pha8se that the bank was allowed to operate without During rehabilitation and restructuring the enough capital. Opponents of repeated authorities are focused on restoring the capital recapitalizations point to the moral hazard involved. position of banks and resolving bad assets. Key Banks' incentives to collect their loans and strategic questions include: borrowers' incentives to repay are undermined as * Is it safe for governments to engage in implicit both await the next bailout, increasing budget costs or explicit regulatory forbearance so that banks and delaying corporate restructuring. Proponents of can strengthen their capital base over time partial (and hence repeated) recapitalizations point through higher profits? to the fiscal pressures that can result from * Should the authorities insist on accomplishing immediate recognition of the full need for additional complete recapitalization immediately, or can capital. recapitalization be done in stages? * Should governments intervene to help Debtor bailouts. If bank bailouts are politically borrowers cover their debts? unpopular, an indirect way of relieving a crisis-and possibly restarting real economic activity-is Regulatory forbearance. If banks still have a introducing a public debt relief program for bank franchise value, they could in principle restore their borrowers. Critics argue that, in addition to obvious capital over time by retaining profits. But such a moral hazard, such programs risk being open-ended, flow solution allows banks to function while attracting borrowers who never would have been undercapitalized and so typically requires able to repay even in good times and diverting forbearance on strict application of prudential investible resources to uncreditworthy firms.4 regulatory requirements. Forbearance is a matter of Proponents of public debt relief schemes contend degree. In its most accommodating form, banks that they are a good tool for mitigating external known to be insolvent are allowed to remain open. shocks beyond the control of corporations. Less accommodating forbearance policies include E e allowing severely undercapitalized banks to remain Empirical evidence open under existing management or temporarily Having considered the various intervention and relaxing any of a range of other regulations-such resolution policies that governments can adopt and as loan classification and loan loss provisioning that may influence the fiscal costs of a crisis, we requirements. Opponents of regulatory forbearance point to the apparent contradiction of relaxing prudential 4 In the empirical analysis we also experimented with a requirements just when they bite, and again note the dummy for the use of asset management companies danger of allowing insolvent or undercapitalized (Klingebiel 2000) and a freeze on deposits. Both of these banks to gamble for resurrection. Proponents of dummies were never significant, so they are not further forbearance counter that regulation should depend discussed here. 20 Managing the Real and Fiscal Effects of Banking Crises now turn to the empirical evidence. Perhaps percentage of GDPR The explanatory variables fall there are no universal solutions to these issues: into three groups: crisis resolution policy variables, country circumstances may determine the macroeconomic indicators, and indicators of the right policy choices. Still, we can examine the nature of the bank failures. (Fuller definitions are statistical relationship between policy choices and provided in the appendix.) crisis costs. Modeling the cross-country variation in fiscal Crisis resolution policy variables. In line with the costs requires us to take into account both policy discussion in the previous section, we used six variables and exogenous variables. The severity of a variables measuring resolution policy tools (table 1). triggering macroeconomic recession and other These are all dummy variables with a value of 0 factors unrelated to management and resolution when policy was strict and 1 when policy was more policies can obviously increase financial distress relaxed: independently of the policies adopted, so we need to * LIQSUP indicates whether central banks or other take that into account to avoid assigning too much government agencies (such as deposit insurance importance to policy. But resolution policies can agencies) provided liquidity support to also deepen losses. Their influence will depend financial institutions. It has a value of I if the on the extent to which a crisis is caused by government provided open-ended, extensive microeconomic factors, including management support (often at below-market rates) to deficiencies in banks. Finally, governments financial institutions regardless of their can choose to cover more or less of the overall financial position. Support is open-ended and losses. Thus the estimating equation may need extensive if it was extended longer than a year to include macroeconomic factors as controls, and was greater than total banking capital (as as well as policy variables. Furthermore, the happened in 23 of our 40 cases)-at which size of the coefficients may depend on whether point it is no longer temporary liquidity support microeconomic weaknesses are pervasive. but rather solvency support. This section describes the data we assembled * GUAR has a value of I if the government issued to estimate these effects and reports on the an explicit blanket guarantee to depositors and regression results. creditors after the onset of the crisis or if market participants were implicitly protected Samp/e and vaniab/es from losses because public banks accounted for It was a major challenge to develop a dataset that at least 75 percent of the market (also 23 cases). identified not just regulatory policies and other * There are two measures of regulatory causal factors but also actual fiscal costs, for which forbearance. FORB-A has a value of I if some most data sources are not very reliable. The sources banks were permitted to continue functioning and methods for the data are described in the despite being technically insolvent (9 cases). appendix. FORB-B has a value of I if either FORB-A was The sample consists of 34 countries (25 of applied or some prudential bank regulations- them developing or transition economies, with such as for loan classification and loan loss 9 in Latin America, 6 in Asia, 5 in Eastern provisioning-were suspended or not fully Europe, and 5 in Africa or the Middle East) that applied (26 cases). experienced significant fiscal costs from bank * An additional indicator of forbearance, REPCAP, failures between 1970 and 2000. Why these has a value of 1 if banks were repeatedly countries? Simply because they are the entire set for which we were able to gather sufficient information on both regulatory practices and fiscal costs. Six of the countries experienced 5 The results reported use the functional form logy. With two distinct crisis episodes. These episodes are this transformation the skewness of the dependent treated separately, giving 40 distinct country variable is greatly reduced, but it has the drawback that experiences. it is undefined as costs approach 0. Alternative The variable to be explained is the estimated functional forms such as log (1 + cost) and cost/(l + total direct fiscal cost of the banking crisis as a cost) gave qualitatively similar results. Controlling the Fiscal Costs of Banking Crises 21 Table 1 Table 2 Government responses to Correlation matrix for individual 40 banking crises policy tools Policy tool (name of variable) Number of times used LIQSUP GUAR FORB-A FORB-B REPCAP PDRP Liquidity support (LfQSUP) 23 LIQSUP 1 0.28 -0.02 0.22 0.10 0.10 Blanket guarantee (GUAR) 23 GUAR 1 -0.14 0.32 0.46 -0.02 Forbearance type A (FORB-A) 9 FORB-A 1 0.27 -0.14 0.28 Forbearance type B (FORB-B 26 FORB-8 1 0.27 0.27 Repeated recapitalizations (RECAP) 9 REPCAP 1 0.00 Public debt relief program (PDRP) 9 PDRP Source: Honohan and Klingebiel (2000). Source: Honohan and Klingebiel (2000). recapitalized. Such events suggest that the Regresslon results initial recapitalization was inadequate and Starting with the parameter estimates for the effectively amounted to capital forbearance (9 macroeconomic indicators, we find that cases). macroeconomic difficulties-as indicated by high * Finally, PDRP has a value of I if the government real interest rates and falling equity prices-tend to implemented an across-the-board public debt increase the fiscal costs of a banking crisis. But relief program. Such programs can be seen as a these variables are mainly included to ensure that further form of accommodation likely to the omission of macroeconomic factors does not generate moral hazard (9 cases). bias the estimate of policy variables. When interpreting the main results, bear in Thus among these tools the most common in mind that the sign of the policy parameters is our sample were regulatory forbearance (type B), set so that a positive coefficient indicates that liquidity support, and blanket government an accommodating policy choice increased guarantees on bank deposits. But authorities were fiscal costs. The main finding is that every selective when being accommodating, so policy significant coefficient is positive. In other words, choices are not strongly correlated (table 2). That we found no specification where an accommodating means, for example, that governments that used policy choice significantly reduced fiscal costs. liquidity support did not necessarily use any Varying the specification by including or excluding particular other policy tool. explanatory variables does not significantly affect the size of the coefficients. This outcome Macroeconomic indicators. Many crises also holds whether or not the macroeconomic were triggered or exacerbated by exogenous variables are included (Honohan and Klingebiel macroeconomic conditions. We explored a 2000). variety of indicators to control for the impact LIQSUP and the two FORBS are the most of macroeconomic shocks on fiscal costs (table 3). consistently significant explanatory variables; GUAR From this set two were consistently significant: is also consistently significant. Replacing FORB-B by the real deposit interest rate (REALINT) and the its product with the dummy MICRO achieves a small change in equity prices (sTocKPRIcE, taken to improvement, modestly supporting the hypothesis the third power to increase the contribution of that using regulatory forbearance as a crisis large values). resolution tool will result in even higher fiscal costs in countries with weak microeconomic Indicators of the nature of the bankfailures. We environments (Honohan and Klingebiel 2000). use a composite variable, MICRO, that aggregates The policy message from these results seems the indicators of microeconomic management clear: open-ended liquidity support, regulatory described in the appendix. This variable is forbearance, and a blanket guarantee for depositors used as a slope dummy with some of the policy and creditors all significantly contribute to the fiscal variables. cost of banking crises. 22 Managing the Real and Fiscal Effects of Banking Crises Table 3 Macroeconomic, microeconomic, and government intrusion indicators before the onset of 40 banking crises Percent Indicator Quartile I Median Quartile IlIl Max/Min Macroeconomic Real deposit interest rate*, 4.2 2.5 0.8 Real GDP growth* -1.6 -0.2 0.9 9.3 Change in equity prices* -27.0 -10.8 20.0 211.0 Current accountGDP -5.8 -3.9 -0.6 2.3 Fiscal balance/GDP -4.7 -1.2 0.3 5.1 Change in terms of trade* -5.7 -0.6 3.4 21.2 Short-term external debUGDP* 56.3 14.4 9.2 7.9 Microeconomic Growth in crediUGDP 407.0 214.0 147.0 116.0 Loans/deposits* 190.5 138.9 111,4 87.6 Government intrusion Bank reserves/deposits 47.3 16.7 8.4 4.4 Share of government in total bank claims 91.3 17.6 11.0 4.0 Bank borrowing from central bank/total bank deposits 80.0 15.9 6.0 2.7 * Average for one year before crisis. Average for two years before crisis. a. Also a microeconomic indicator. Source: World Bank. There is one obvious potential problem of years; each dummy takes a value of I for countries simultaneity here: really big crises may have whose crisis began that year, 0 otherwise). The time triggered the adoption of policies such as blanket dummies could be valid instruments to the extent guarantees or liquidity support (especially if these that the choice of accommodating policies in a policies can be seen to some extent as being particular year is influenced by global trends or analogous to the government burying its head in overall world conditions. the sand) in countries that otherwise relied on According to Honohan and Klingebiel (2000), strict policies. To verify that our results are not two-stage least squares estimates of the main contaminated by such reverse causality, we used equations using these instruments come out close to an instrumental variables approach. the ordinary least squares results. This finding Our two types of predetermined instruments implies that the predicted degree of accommodation used data on the political and institutional from the first-stage regressions is just as strong a environment and on timing. The political and predictor of fiscal costs as the actual degree of institutional data are those published by the accommodation. A regression of the residuals on the International Country Risk Guide measuring instruments is not significant, providing some corruption in government (coRRuPT) and the reassurance that the instruments are indeed tradition of law and order (LAwoRDER). These predetermined. In all, then, this evidence suggests were used because it is assumed that countries that reverse causality is not a problem for the with weak institutions, as measured by either interpretation of our results. variable, are more likely to adopt accommodating We also experimented with alternative policies. functional forms. Several different forms give a The timing instruments are dummy variables similar fit without dominating the one shown for the years when crises began (there are 14 such (though as noted below, the exact functional form Controlling the Fiscal Costs of Banking Crises 23 does have implications for the size of out-of-sample equation predicts that if deposit guarantees, one predictions). form of forbearance, and repeated recapitalizations are used, not extending liquidity support could halve Sca/e of tle cost Implications the expected fiscal cost. Our empirical findings show that accommodating Another caveat worth repeating is that the final policies significantly increase the fiscal costs of regressions do not include variables measuring banking crises. If we were to take the regression policies in place before the interventions considered results literally and simulate the effects of uniformly here. To the extent that such policies are important strict and uniformly accommodating policy (and to the extent that they would be correlated with packages, we would obtain rather extreme results. the included policy variables), their omission may These results imply that a crisis country bias the estimated coefficients of the included that did not provide open-ended liquidity support, policy variables. Accommodating pre-crisis policies blanket deposit guarantees, regulatory forbearance, that allowed financial institutions to take big risks repeated recapitalizations, or debtor bailouts would might well be associated with accommodating have incurred a fiscal cost of about I percent of intervention and resolution policies that raised post- GDP. On the other hand, a crisis country that crisis costs. provided all these forms of support would have a predicted fiscal cost of more than 60 percent of Is there a tradeoff between fiscal costs and GDP (Honohan and Klingebiel 2000). But because economic recovery? they are calculated beyond the range of the sample, We also explored whether there was an obvious and taking into account their sensitivity to the tradeoff between fiscal costs and subsequent functional form of the equation, these limiting economic growth. In other words, did countries that projections should probably not be taken too used accommodating policies experience faster literally. macroeconomic recovery? Using a standard Still, more realistic calculations showing the approach to measure the size and duration of estimated impact of each policy tool-assuming the the output dips following the crises (IMF 1998), other tools are used with their actual frequency- regressions using the same structure as for fiscal also show sizable effects (table 4). These costs fail to uncover evidence that this was the case calculations indicate that the most expensive (Honohan and Klingebiel 2000). Except for accommodating measures are liquidity support and liquidity support, all the policy variables proved the various forms of forbearance, each of which insignificant. And for liquidity support, the positive costs several percentage points of GDP. The coefficient indicates that such support actually appears to have prolonged crises, because recovery Table 4 took longer. Furthermore, the estimated output dip Estimated fiscal costs of was bigger. accommodating policy tools Policy tool Cost of adopting each tool Conclusion (percentage of cases where used) (percentage of GDP) In this chapter we have made a first attempt Forbearance type A (24) 6.7 to understand whether and what kind of crisis Repeated recapitalizations (24) 6.3 resolution policies are effective in lowering the Liquidity support (58) 6.3 fiscal costs of banking crises. While much discussion suggests that most of the costs of Forbearance type B (84) 4.1 banking crises come from exogenous shocks, we Public debt relief program (21) 3.1 find evidence that resolution policies matter-and Blanket guarantee (55) 2.9 that strict resolution policies contain fiscal costs. Note: The table shows how much each accommodating measure can add Of course, it may also be that the underlying to fiscal costs. For example, permitting insolvent banks to stay open policy philosophy that tends to generate strict (forbearance type A; see text) raises predicted fiscal costs by 6.7 percent policies is associated with an environment that of GDP twice the sample mean (each calculation uses the sample mean helps contain costs before the recognition phase- value of the other variables). Source: Honohan and Klingebiel (2000). that is, before a crisis is recognized as such. By the time containment and resolution policies come into 24 Managing the Real and Fiscal Effects of Banking Crises play, some of the damage of an accommodating significantly contribute to the fiscal costs of policy stance will have already been done. banking crises. Countries that avoid these policies But while we have emphasized intervention and can expect to considerably cut the costs of future resolution policies, it is not really possible to draw crises. an unambiguous line between these and prevention Containing and resolving banking crises is not policies. To the extent that prevention policies have easy, and the exact policy approach cannot be been explicitly included, our estimates may dictated by the results of a model simplified for somewhat exaggerate the role of intervention and econometric testing. We can hardly claim to have resolution. proven the best policy choices in all circumstances. The data on which we depend are tentative, and Nevertheless, our findings clearly favor a strict one should not rely too heavily on the precise approach to resolving crises, rather than an coefficient estimates. Still, the effects we model are accommodating one. At the very least, they statistically significant, have a consistent sign, and emphasize that regulatory authorities who choose an are economically large. In particular, open-ended accommodating or gradual approach to an emerging liquidity support, regulatory forbearance, and a crisis need to be sure that they have some other way blanket guarantee for depositors and other creditors of controlling risk taking. Appendix: Descriptions of variables and data Dependent variab/es remains below the trend value. As in IMF (1998), Fiscal cost recovery time is I plus the duration of the dip (in years). As noted in the text, Hoggarth, Reis, and The dependent variable fiscal cost is the estimated Saporta (forthcoming) propose alternative dip net present value of the budget cost of the crisis measures. based on official or expert assessments, expressed as a percentage of GDP (table Al). The first date shown Data on criissresolution too/s for the crisis is the date when the existence of the To characterize the main components of a crisis crisis became publicly known. The fiscal cost figure resolution strategy, we use dummy variables (shown includes both fiscal and quasi-fiscal outlays for in table Al) characterizing each government's financial system restructuring, including the cost of approach along five dimensions: recapitalizing banks, bailout costs related to covering * Issuance of a blanket government guarantee depositors and creditors, and debt relief schemes for (GUAR). Did the government issue an explicit and bank borrowers. unlimited guarantee for depositors and creditors, Data on fiscal costs and crisis dates are from or were market participants implicitly protected World Bank estimates assembled from published because deposits in state-owned institutions sources and from recent discussions with national accounted for more than 75 percent of banking experts. The estimates here draw on those reported deposits? by Caprio and Klingebiel (1996 and in this volume) * Open-ended, extensive liquidity support to and Lindgren, Garcia, and Saal (1996). Conflicts insolvent institutions (LIQsuP). Did the between sources have been reconciled with the help government or one of its agencies (typically a of country experts. central bank or deposit insurance agency) provide open-ended, extensive liquidity support Output dip (at preferential rates) to financial institutions Dips in output growth following banking crises are regardless of their financial standing? (Support calculated using the approach of and data from IMF is open-ended and extensive if it was provided (1998), with updates for more recent crises (table for longer than a year and was greater than total A2). This approach calculates the output dip as the banking capital.) cumulative deviation of output from its previous * Forbearance (FoRB). Did the government trend growth during the duration of the dip, defined forbear in either of the following progressively as the period over which the output growth rate less liberal ways? Under forbearance type A, Controlling the Fiscal Costs of Banking Crises 25 Table Al Crisis intervention and resolution policy tools Blanket guarantee for Extensive Repeated Public debt Fiscal cost depositors and liquidity support Forbearance? recapitalizations relief for Country Period (% of GDP) creditors (GuAR)? (LIQsuP)? FoRD-A FoRB-B (RECAP)? borrowers (PDRP)? Argentina (I) 1980-82 55.1 Yes No No Yes No Yes Argentina (II) 1995 0.5 No No No No No No Australia 1989-92 1.9 No No No Yes No No Brazil 1994-96 13.2 No No Yes Yes No Yes Bulgaria 1996-97 13.0 Yes Yes Yes Yes No No Chile 1981-83 41.2 No Yes No Yes No Yes Colombia 1982-87 5.0 Yes Yes No No No No CWte d'lvoire 1988-91 25.0 No Yes Yes Yes No No Czech Republic 1989-91 12.0 Yes No No Yes Yes No Ecuador 1996-present 13.0 No No Yes Yes No Yes Egypt 1991-95 0.5 Yes Yes No Yes No No Finland 1991-94 11.0 Yes Yes No Yes No No France 1994-95 0.7 No No No Yes No No Ghana 1982-89 3.0 Yes Yes Yes Yes No Yes Hungary 1991-95 10.0 Yes Yes No Yes Yes No Indonesia (I) 1992-94 3.8 No No No Yes No No Indonesia (II) 1997-present 50.0 Yes Yes No Yes Yes No Japan 1992-present 20.0 Yes Yes No Yes Yes No Korea, Rep. of 1997-present 26.5 Yes Yes Yes Yes Yes No Malaysia (I) 1985-88 4.7 No Yes No Yes No No Malaysia (II) 1997-present 16.4 Yes No No Yes Yes No Mexico 1994-present 19.3 Yes Yes No Yes Yes Yes New Zealand 1987-90 1.0 No Yes No No No No Norway 1987-93 8.0 Yes Yes No Yes No No Paraguay 1995-present 5.1 Yes Yes No Yes No No Philippines (I) 1983-87 13.2 No Yes Yes Yes No Yes Philippines (II) 1998-present 0.5 No No No No No No Poland 1992-95 3.5 Yes Yes No Yes No No Senegal 1988-91 9.6 Yes Yes No Yes No Yes Slovenia 1992-94 14.6 Yes No Yes Yes No No Spain 1977-85 5.6 No Yes No Yes No No Sri Lanka 1989-93 5.0 Yes No No Yes Yes No Sweden 1991-94 4.0 Yes No No No No No Thailand (I) 1983-87 2.0 No No No Yes No No Thailand (II) 1997-present 32.8 Yes Yes No Yes No No Turkey (I) 1982-85 2.5 No No No No No No Turkey (II) 1994 1.1 Yes No No Yes No No United States 1981-91 3.2 No No Yes Yes No No Uruguay 1981-84 31.2 Yes Yes No Yes Yes Yes Venezuela, RB 1994-97 22.0 No Yes No Yes No No Source: Compiled by authors. 26 Managing the Real and Fiscal Effects of Banking Crises banks observed to be in open distress-such of borrowers, including through an exchange as those unable to pay depositors, with no access rate guarantee program or rescue of to the interbank market, or widely believed corporations? to be insolvent (except public banks)-are Data on these measures drew on the dataset allowed to continue operating without any from Caprio and Klingebiel (1996) and added restrictions for at least a year. Under forbearance countries and policy variables. Information type B, either forbearance type A was applied on the policy variables came from official or regulations (particularly those for loan country sources, from the World Bank Regulatory classification and loan loss provisioning) are Database (Barth, Caprio, and Levine 2001), relaxed or the current regulatory framework from Garcia (1999) and other IMF reports, and is not enforced for at least a year. from interviews with country experts. Complete * Repeated recapitalizations (REPcAP). Did banks data are available for 40 episodes involving undergo more than one round of government- 34 countries. sponsored recapitalizations? * Public debt reliefprogram (PDRP). Did the Co0nro/ variables government implement a broad debt relief We used data summarizing macroeconomic program for corporations or other types conditions, the regulatory and management Table A2 Crisis country scores for microeconomic indicators and output dips Growth in Real deposit Loan Enforcement Loan to MICRO Output Duration credit/GDP interest rate classificatlon, of creditor deposit ratio (0 if mean of growth dipc of dip, Country (I) (1I) (III) rightsb (IV) (V) I-V 2 2.4) (% of GDP) (years) Argentina (I) 3 1 2 2 3 1 16.6 4 Argentina (II) 1 2 3 4 2 0 11.9 3 Australia 3 3 3 4 2 0 0 1 Brazil 2 . 3 3 2 1 1 0 Bulgaria 4 1 3 . 4 0 20.4 3 Chile 1 3 3 4 1 0 45.5 9 Colombia 3 2 2 1 3 1 65.1 5 Cote d'lvoire 4 1 1 2 1 1 Czech Republic 2 3 1 3 . 1 0 1 Ecuador 1 4 3 2 3 0 0.9 1 Egypt 4 1 . 1 4 1 6.5 5 Finland 3 2 4 4 1 0 23.1 7 France 4 2 4 4 1 0 0 1 Ghana 4 1 1 1 4 1 6.6 2 Hungary 4 2 1 3 1 1 13.8 3 Indonesia (I) 1 4 1 1 2 1 42.3 9 Indonesia (II) 3 4 2 3 2 0 33.0 4 Japan 2 2 4 3 2 0 27.7 9 Korea, Rep. of 2 3 2 2 1 1 16.5 3 Malaysia (I) 1 4 2 3 2 0 13.7 4 Malaysia (II) 2 3 2 3 2 0 22.8 4 Mexico 1 4 2 2 1 1 9.6 2 New Zealand 2 2 4 2 4 0 18.5 7 Norway 1 4 1 4 2 0 19.6 8 Controlling the Fiscal Costs of Banking Crises 27 environment affecting bank management (referred stay afloat), loan classification rules (proxy to as microeconomic indicators), and the degree of for quality of regulation; see note to table government intrusion: A2), enforcement of creditor rights (proxy * Macroeconomic indicators, as averages for one for the effectiveness of the legal system; (*) or two (t) years before the crisis year: real see note to table A2), and average ratio of deposit interest rate* (also a microeconomic bank loans to deposits* (proxy for liquidity indicator), real GDP growth*, percentage risk). change in equity (stock market) prices*, current * Government intrusion indicators: bank account as a percentage of GDPt, fiscal reserves (cash plus with central bank) as balance as a percentage of GDPt, percentage percentage of deposits, share of government change in the terms of trade*, and short-term in total bank claims, and bank borrowing external debt as a percentage of GDP*. from central bank as a percentage of total * Microeconomic indicators: growth in bank deposits. bank credit relative to GDPt (as proxy for Each continuous control variable was normalized relaxed credit risk standards), real deposit to zero mean and unit standard deviation. The interest rate (possible proxy for financial variable MICRO iS a composite of the microeconomic system distress as banks bid up rates to indicators: it takes a value of I when the country Table A2 Crisis country scores for microeconomic indicators and output dips-continued Growth in Real deposit Loan Enforcement Loan to MICRO Output Duration credit/GWP interest rate classification' of creditor deposit ratio (o if mean of growth dip, of dip, Country (I) (II) (III) rightsb (IV) (V) I-V 2 2.4) (% of GDP) (years) Paraguay 2 3 3 4 3 0 0 1 Philippines (I) 3 3 2 1 1 1 25.7 5 Philippines (II) 1 3 3 2 3 0 7.5 3 Poland 2 1 1 2 4 1 0 1 Senegal 4 4 1 1 1 1 0 1 Slovenia . 4 1 4 3 0 2.1 2 Spain 3 1 1 2 4 1 0 1 SriLanka 1 2 . 1 3 1 0.5 3 Sweden 1 2 3 4 1 1 6.5 3 Thailand (I) 2 3 1 1 3 1 8.7 2 Thailand (II) 3 4 1 2 1 1 31.5 4 Turkey (I) 3 1 1 4 4 0 0 1 Turkey (II) 4 1 3 4 4 0 9.1 2 United States 2 3 4 4 2 0 5.4 3 Uruguay 3 1 . 2 3 1 41.7 6 Venezuela, RB 4 4 2 1 4 0 14.1 4 Note: Except where otherwise indicated, each country was scored 1, 2, 3, or 4 for each of the microeconomic variables, with lower values indicating weaker conditions. a. Scored as follows: 4 indicates forward-looking provisioning criteria, 3 indicates that provisioning is required when loans are 90 days overdue, 2 indicates that provisioning is required when loans are 120 days overdue, and I indicates that provisioning is required when loans are 360 days overdue. b. Based on La Porta, Lopez-de-Silanes, and Shliefer(1998), where thresholds are set at scores of 6 (changed to I in this table), 12 ( changed to 2), and 18 (changed to 3). c. IMF (1998) methodology; see text. Source: Compiled and calculated by authors. 28 Managing the Real and Fiscal Effects of Banking Crises has a low average value of the microeconomic Note that, of the macroeconomic and indicators relative to other countries; otherwise the government control variables, only the real interest value is 0.6 rate and change in equity prices were significant in the regressions. The others were then excluded from all reported regressions. Data on control variables came from the 6 Specifically, each country was scored 1, 2, 3, or 4 for Data onterntiol Fanial Sati (ban dat each of the microeconomic variables (with lower values indicating weaker conditions; see table A2). Each refers to deposit money banks), the International country's mean of these scores was then computed and Finance Corporation's Emerging Markets Database, micro set to I for countries lower than the median across and La Porta and other (1998) (for enforcement countries. Thus micro is I when microeconomic of creditor rights), supplemented by national conditions are weak. sources. References Baer, Herbert, and Daniela Klingebiel. 1995. "Systemic Risk When Depositors Bear Losses: Five Case Studies." In G. G. Kaufman, ed., Research in Financial Services: Private and Public Policy. vol. 7. Greenwich, Conn.: JAI Press. Barth, James R., and Gerard Caprio, and Ross Levine. 2001. "The Regulation and Supervision of Banks around the World: A New Database." Policy Research Working Paper 2588. World Bank, Washington, D.C. Benston, George J., and G. G. Kaufman. 1995. "Is the Banking and Payments System Fragile?" Journal of Financial Services Research 9: 209-40. Caprio, Gerard, and Patrick Honohan. 1999. "Restoring Banking Stability: Beyond Supervised Capital Requirements." Journal of Economic Perspectives 13 (4): 43-64. 2000. "Reducing the Cost of Banking Crises: Is Basel Enough?" Paper presented at the American Economic Association annual meetings, Boston, Mass. Caprio, Gerard, and Daniela Klingebiel. 1996. "Bank Insolvencies: Cross-Country Experience." Policy Research Working Paper 1620. World Bank, Washington, D.C. 1997. "Bank Insolvency: Bad Luck, Bad Policy, or Bad Banking?" In Michael Bruno and Boris Pleskovic, eds., Annual World Bank Conference on Development Economics 1996. Washington, D.C.: World Bank. Claessens, Stijn. 1999. "Experiences of Resolution of Banking Crises." In Strengthening the Banking System in China. BIS Policy Paper 7. Bank for International Settlements, Basel, Switzerland. Claessens, Stijn, Simeon Djankov, and Daniela Klingebiel. 2001. "Financial Restructuring in East Asia- Halfway There?" In Stijn Claessens, Simeon Djankov, and Ashoka Mody, eds., Resolution of Financial Distress. Washington, D.C.: World Bank. Demirgfi-Kunt, Asli, and Enrica Detragiache. 1998. "The Determinants of Banking Crises in Developing and Developed Countries." IMF Staff Paper. International Monetary Fund, Washington, D.C. 1999. "Monitoring Banking Sector Fragility: A Multivariate Logit Approach with an Application to the 1996-97 Banking Crises." The World Bank Economic Review 14 (2): 287-307. Demirguc-Kunt, Asli, and Harry Huizinga. 1999. "Market Discipline and Financial Safety Net Design." Policy Research Working Paper 2183. World Bank, Washington, D.C. Controlling the Fiscal Costs of Banking Crises 29 Garcia, Gillian. 1999. "Deposit Insurance-A Survey of Actual and Best Practices." IMF Working Paper 99/ 54. International Monetary Fund, Washington, D.C. Hoggarth, Glen, Ricardo Reis, and Victoria Saporta. Forthcoming. "Cost of Bank Instability: Some Empirical Evidence." Journal of Banking and Finance. Honohan, Patrick. 1999. "A Model of Bank Contagion through Lending." International Review of Economics and Finance 8 (2): 147-63. 2000. "Banking System Failures in Developing and Transition Countries: Diagnosis and Prediction." Economic Notes 29 (1): 83-109. Honohan, Patrick, and Daniela Klingebiel. 2000. "Controlling the Fiscal Costs of Banking Crises." Policy Research Working Paper 2441. World Bank, Washington, D.C. IMF (International Monetary Fund). 1998. World Economic Outlook (May). Washington, D.C. Kaufman, G. G. 1994. "Bank Contagion: A Review of the Theory and Evidence." Journal of Financial Services Research 8: 123-50. Klingebiel, Daniela. 2000. "The Use of Asset Management Companies in the Resolution of Banking Crises-Cross-Country Experience." Policy Research Working Paper 2284. World Bank, Washington, D.C. La Porta, Rafael, Florencio Lopez-de-Silanes, and Andrei Shleifer. 1998. "Law and Finance." Journal of Political Economy 106 (6): 13-55. Lindgren, Carl-Johan, Gillian Garcia, and Matthew I. Saal. 1996. Bank Soundness and Macroeconomic Policy. Washington, D.C.: International Monetary Fund. McKinnon, Ronald. 1996. The Rules of the Game: International Money and Exchange Rates. Cambridge, Mass.: MIT Press. Merton, Robert C. 1977. "An Analytic Derivation of the Cost of Deposit Insurance Using Option-Pricing Estimates." Journal of Banking and Finance 1: 3-11. Sheng, Andrew, ed. 1996. Bank Restructuring: Lessonsfrom the 1980s. Washington, D.C.: World Bank. Shleifer, Andrei, and Robert Vishny. 1993. "Corruption." Quarterly Journal of Economics 108: 599-617. 31 Episodes of Systemic and Borderline Banking Crises Gerard Caprio and Daniela Klingebiel Gerard Caprio is director and Daniela Klingebiel World Bank. The authors are grateful to World is senior financial economist in the Financial Bank staff who provided information for this Sector Strategy and Policy Department at the database. he following table presents information on being detected. The dates attached to the crises 113 systemic banking crises (defined as reviewed here are those generally accepted by T much or all of bank capital being exhausted) finance experts familiar with the countries, but their that have occurred in 93 countries since the late accuracy is difficult to determine in the absence of 1970s. The table also provides information on 50 the means to mark portfolios to market values. borderline and smaller (nonsystemic) banking crises Similarly, it is not always clear when a crisis is over. in 44 countries during that period. The data are as of In countries that have experienced multiple crises, 1999. Some judgment has gone into the compilation later events may just be a continuation of earlier of this list, not only for countries lacking data on the events. size of the losses but also for countries where As the table shows, the costs of banking crises official estimates understate the problem. For vary widely. But the data on losses and costs should instance, at some point in the 1 990s nearly every be treated with caution. Some of the data include transition economy experienced a banking crisis, corporate restructuring, while others relate only to the but not all of these were included to limit the restructuring and recapitalization of the financial number of countries with missing information. system. Moreover, we are not able to include the Moreover, it is difficult to identify the burden borne by depositors and borrowers in the timeframes of banking insolvencies. Overt crises- form of wider interest rate spreads resulting from such as those involving a run on banks, on a bad loans left on banks' balance sheets. Finally, country's currency, or both-are fairly easy to date, most of the data on costs do not include costs but these are only a subset of the cases listed here. resulting from indirect methods of bailing out Financial distress, in which the banking system has banks. For example, a government may subsidize a negative net worth, can occur over a period of time. borrower by granting it monopoly privilege or other Indeed, a crisis may persist for some time before means to improve profits and so repay loans. 32 Managing the Real and Fiscal Effects of Banking Crises Systemic Banking Crises Sub-Saharan Africa Economy Scope of crisis Estimated losses or costs Benin 1988-90 All three commercial banks collapsed; 80 percent of banks' CFA 95 billion, equivalent to loan portfolios were nonperforming. 17 percent of GDP. Burkina Faso 1988-94 Banking system nonperforming loans estimated at 34 percent. Burundi 1994-? Banking system nonperforming loans estimated at 25 percent in 1995; one bank was liquidated. Cameroon 1987-93 In 1989 banking system nonperforming loans reached 60-70 percent. Five commercial banks were closed and three banks were restructured. 1995-98 At the end of 1996 nonperforming loans accounted for 30 percent of total loans. Three banks were restructured and two were closed. Cape Verde 1993-? At the end of 1995 commercial banks' nonperforming loans reached 30 percent. Central African Rep. 1976-92 Four banks were liquidated. 1988-99 The two largest banks, accounting for 90 percent of assets, were restructured. Banking system nonperforming loans reached 40 percent. Chad 1980s Banking sector experienced solvency problems. 1992 Nonperforming loans to the private sector reached 35 percent. Congo, Dem. Rep. of (former Zaire) 1980s Banking sector experienced solvency problems. 1991-92 Four state-owned banks were insolvent; a fifth bank was to be recapitalized with private participation. 1994-present Nonperforming loans to the private sector reached 75 percent. Two state-owned banks have been liquidated and two other state banks privatized. In 1997, 12 banks were having serious financial difficulties. Congo, Rep. of 1992-present Two large banks were liquidated. The three remaining banks are insolvent. Situation aggravated by the civil war. CWte d'lvoire 1988-91 Four large banks affected, accounting for 90 percent of Government costs estimated at banking system loans; three definitely and one possibly CFA 677 billion, equivalent to insolvent. Six government banks closed. 25 percent of GDP. Episodes of Systemic and Borderline Banking Crises 33 Systemic Banking Crises Sub-Saharan Africa-continued Economy Scope of crisis Estimated losses or costs Equatorial Guinea 1983-85 Two of the country's largest banks were liquidated. Eritrea 1993 Most of the banking system was insolvent. Ghana 1982-89 Seven of eleven audited banks insolvent; rural banking Restructuring costs estimated sector affected. at 6 percent of GNP. Guinea 1985 Six banks-accounting for 99 percent of system Repayment of deposits deposits-deemed insolvent. amounted to 3 percent of 1986 GDP. 1993-94 Two banks deemed insolvent; one other bank had serious financial difficulties. Together these three banks accounted for 45 percent of the market. Guinea-Bissau 1995-? At the end of 1995 nonperforming loans accounted for 45 percent of commercial banks' loan portfolio. Kenya 1985-89 Four banks and twenty-four nonbank financial institutions-accounting for 15 percent of financial system liabilities-faced liquidity and solvency problems. 1992 Intervention in two local banks. 1993-95 Serious solvency problems with banks accounting for more than 30 percent of financial system assets. Liberia 1991-95 Seven of eleven banks not operational; in mid-1995 their assets accounted for 60 percent of bank assets. Madagascar 1988 25 percent of bank loans deemed unrecoverable. Mali 1987-89 Nonperforming loans of largest bank reached 75 percent. Mauritania 1984-93 In 1984 five major banks had nonperforming assets ranging Cost of rehabilitation estimated from 45-70 percent of their portfolios. at 15 percent of GDP in 1988. Mozambique 1987-95? Main commercial bank experienced solvency problems that became apparent after 1992. Niger 1983-? In the mid-1980s banking system nonperforming loans reached 50 percent. Four banks were liquidated and three restructured in the late 1980s. 34 Managing the Real and Fiscal Effects of Banking Crises Systemic Banking Crises Sub-Saharan Africa-continued Economy Scope of crisis Estimated losses or costs Nigeria 1990s In 1993 insolvent banks accounted for 20 percent of banking system assets and 22 percent of deposits. In 1995 almost half the banks reported being in financial distress. Sio Tom6 and Principe 1980s-1990s At the end of 1992, 90 percent of the monobank's loans were nonperforming. In 1993 the commercial and development departments of the former monobank were liquidated, as was the only financial institution. At the same time, two new banks were licensed that took over many of the assets of their predecessors. The credit operations of one new bank have been suspended since late 1994. Senegal 1988-91 In 1988, 50 percent of banking system loans were $830 million, equivalent to nonperforming. Six commercial banks and one development 17 percent of GDP. bank closed, accounting for 20-30 percent of financial system assets. Sierra Leone 1990-present In 1995, 40-50 percent of banking system loans were nonperforming. One bank's license was suspended in 1994. Bank recapitalization and restructuring are ongoing. Swaziland 1995 Meridien BIAO Swaziland was taken over by the Central Bank. The Central Bank also took over the Swaziland Development and Savings Bank, which faced severe portfolio problems. Tanzania Late 1980s; 1990s In 1987 the main financial institutions had arrears amounting to In 1987 implied losses half their portfolios. In 1995 it was determined that the National amounted to nearly 10 percent Bank of Commerce, which accounted for 95 percent of banking of GNP. system assets, had been insolvent since at least 1990. Togo 1993-95 Banking sector experienced solvency problems. Uganda 1994-present Half of banking system facing solvency problems. Zambia 1995 Meridian Bank, which accounted for 13 percent of commercial $50 million (1.4 percent bank assets, became insolvent. of GDP). Zimbabwe 1995-present Two of five commercial banks have high nonperforming loans. Episodes of Systemic and Borderline Banking Crises 35 Systemic Banking Crises East and South Asia Economy Scope of crisis Estimated losses or costs Bangladesh Late 1980s-96 In 1987 four banks accounting for 70 percent of credit had nonperforming loans of 20 percent. From the late 1980s the entire private and public banking system was technically insolvent. China 1990s At the end of 1998 China's four large state-owned commercial Net losses estimated to reach banks, accounting for 68 percent of banking system assets, $428 billion, or 47 percent of were deemed insolvent. Banking system nonperforming loans GDP in 1999. were estimated at 50 percent. Indonesia 1997-present By March 1999 Bank Indonesia had closed 61 banks and Fiscal costs estimated at nationalized 54, of a total of 240. Nonperforming loans for the 50-55 percent of GDP banking system estimated at 65-75 percent of total loans. Korea, Rep. of 1997-present In March 1999, 2 of 26 commercial banks-accounting for Net losses estimated at 12 percent of banking system assets-were nationalized. Five $68 billion, or 20 percent of banks-accounting for 8 percent of banking system assets- GDP in 1999. were closed. Seven banks-accounting for 38 percent of banking system assets-were placed under special supervision. Banking system nonperforming loans are expected to peak at 30-40 percent. Malaysia 1997-present Finance companies are being restructured, and the number of Net losses estimated at finance companies is to be reduced from 39 to 16 through $15 billion, or 21 percent of mergers. Two finance companies were taken over by the GDP in 1999. Central Bank, including the largest independent finance company. Two banks deemed insolvent-accounting for 14 percent of financial system assets-will be merged with other banks. At the end of 1998 nonperforming loans estimated at 25-35 percent of banking system assets. Nepal 1988 In early 1988 the reported arrears of three banks accounting for 95 percent of the financial system averaged 29 percent of assets. Philippines 1981-87 Problems in two public banks accounting for 50 percent of At its peak, central bank banking system assets, six private banks accounting for assistance to financial 12 percent of banking system assets, 32 thrifts accounting for institutions amounted to 53 percent of thrift banking assets, and 128 rural banks. 19 billion pesos (3 percent of GDP). 1998-present Since January 1998 one commercial bank, 7 of 88 thrifts, and Net losses estimated at 40 of 750 rural banks have been placed under receivership. $4 billion, or 7 percent of Banking system nonperforming loans reached 12 percent by GDP in 1999. November 1998, and were expected to reach 20 percent in 1999. 36 Managing the Real and Fiscal Effects of Banking Crises Systemic Banking Crises East and South Asia-continued Economy Scope of crisis Estimated losses or costs Sri Lanka 1989-93 State-owned banks comprising 70 percent of banking system Restructuring cost amounted estimated to have nonperforming loans of about 35 percent. to 25 billion rupees (5 percent of GDP). Taiwan, China 1997-98 Banking system nonperforming loans estimated at 15 percent at In 1999 net losses estimated the end of 1998. at $26.7 billion, or 11.5 percent of GDP. Thailand 1983-87 Authorities intervened in 50 finance and security firms and Government cost for 50 5 commercial banks, or about 25 percent of financial system finance companies estimated assets; 3 commercial banks deemed insolvent (accounting for at 0.5 percent of GNP; 14 percent of commercial bank assets). government cost for subsidized loans amounted to about 0.2 percent of GDP a year. 1997-present Through March 1999 the Bank of Thailand intervened in 70 (of Net losses estimated at 91) finance companies that together accounted for 13 percent of $60 billion, or 42 percent of financial system assets and 72 percent of finance company GDP in 1999. assets. It also intervened in six banks that together had a market share of 12 percent. At the end of 1998 banking system nonperforming loans reached 46 percent. Vietnam 1997-present Two of four large state-owned commercial banks-accounting for 51 percent of banking system loans-deemed insolvent; the other two are experiencing significant solvency problems. Several joint stock banks are in severe financial distress. Banking system nonperforming loans reached 18 percent in late 1998. Latin America and the Caribbean Economy Scope of crisis Estimated losses or costs Argentina 1980-82 More than 70 institutions-accounting for 16 percent of 55 percent of GDP. commercial bank assets and 35 percent of finance company assets-were liquidated or subjected to central bank intervention. 1989-90 Nonperforming assets accounted for 27 percent of the aggregate portfolio and 37 percent of the portfolios of state banks. Failed banks held 40 percent of financial system assets. 1995 Eight banks suspended and three banks collapsed. Through the Direct and indirect costs to end of 1997, 63 of 205 banking institutions were closed or public estimated at 2 percent merged. of GDP. Episodes of Systemic and Borderline Banking Crises 37 Systemic Banking Crises Latin America and the Caribbean-continued Economy Scope of crisis Estimated losses or costs Bolivia 1986-88 Five banks were liquidated. Banking system nonperforming loans reached 30 percent in 1987; in mid-1988 reported arrears stood at 92 percent of commercial banks' net worth. 1994-? Two banks with 11 percent of banking system assets were closed in 1994. In 1995, 4 of 15 domestic banks, accounting for 30 percent of banking system assets, experienced liquidity problems and suffered high nonperforming loans. Brazil 1990 Deposits converted to bonds. 1994-present By the end of 1997 the Central Bank had intervened in or put In 1996 the negative net worth under temporary administration 43 financial institutions, and of selected state and federal banking system nonperforming loans reached 15 percent. banks was 5-10 percent of GDP. By the end of 1997 bank recapitalizations had cost $3 billion for Banco Economico, $3 billion for Bamerindus, $8 billion for Banco do Brazil, and $5 billion for Unibanco. In 1998 public support to private banks cost 1-2 percent of GDP. Chile 1976 Entire mortgage system insolvent. 1981-83 In 1981 the authorities intervened in four banks and four In 1982-85 the government nonbank financial institutions accounting for 33 percent of spent 42 percent of GDP to outstanding loans. In 1983 the authorities intervened in seven resolve the banking crisis. banks and one financiera accounting for 45 percent of financial system assets. By the end of 1983, 19 percent of loans were nonperforming. Colombia 1982-87 The Central Bank intervened in six banks accounting for Restructuring costs were 25 percent of banking system assets. estimated to be about 5 percent of GDP. Costa Rica Several instances In 1987 public banks accounting for 90 percent of banking Implied losses of at least twice system loans were in financial distress, with 32 percent of their the capital plus reserves. loans considered uncollectable. 38 Managing the Real and Fiscal Effects of Banking Crises Systemic Banking Crises Latin America and the Caribbean-continued Economy Scope of crisis Estimated losses or costs Ecuador Early 1 980s Program exchanging domestic for foreign debt implemented to bail out banking system. 1 996-present The authorities intervened in several small financial institufions in late 1995 and early 1996 and in the fifth largest commercial bank in 1996. Seven financial institutions, accounting for 25-30 percent of commercial banking assets, were closed in 1998-99. In March 1999 the authorities declared a one-week bank holiday. El Salvador 1989 Nine state-owned commercial banks had nonperforming loans averaging 37 percent. Jamaica 1994-present In 1994 a merchant banking group was closed. In 1995 a medium-size bank received financial support. In 1997 the Financial Credit Adjustment Company effectively nationalized five of six commercial banks as a result of a sharp deterioration in their asset quality and the erosion of their capital base. Mexico 1981-91 Govemment took over troubled banking system. 1995-present Of 34 commercial banks in 1994, 9 were intervened in and 11 Bank rescue estimated to participated in the loan/purchase recapitalization program. The cost $65 billion by early 1998, 9 intervened banks accounted for 19 percent of financial system or nearly 15 percent of GDP. assets and were deemed insolvent. Nicaragua Late 1980s-96 Banking system nonperforming loans reached 50 percent in 1996. Panama 1988-89 In 1988 Panama's banking system experienced a nine-week banking holiday. The financial position of most state-owned and private commercial banks was weak. As a result 15 banks ceased operations. Paraguay 1995-present The Government Superintendency intervened in two connected By May 1998 the govemment commercial banks, two other banks, and six related finance had spent $500 million, or houses accounting for 10 percent of financial system deposits. 5 percent of GDP. By 1998 the government had intervened in six other financial institutions, including the country's largest public bank and the largest savings and loan institution. By the end of 1998 the government had intervened in most remaining domestic private and public banks and a number of finance companies. Episodes of Systemic and Borderline Banking Crises 39 Systemic Banking Crises Latin America and the Caribbean-continued Economy Scope of cuisis Estimated losses or costs Peru 1983-90 Two large banks failed. The rest of the system suffered from high nonperforming loans and financial disintermediation following the nationalization of the banking system in 1987. Uruguay 1981-84 Affected institutions accounted for 30 percent of financial system The costs of recapitalizing assets; insolvent banks accounted for 20 percent of financial banks were estimated at system deposits. $350 million, or 7 percent of GNP. In 1982-85 the Central Bank's quasi-fiscal losses associated with subsidized credit operations and loan portfolio purchases amounted to 24 percent of GDP. Venezuela, RB 1994-present Insolvent banks accounted for 30 percent of financial system Losses were estimated at deposits in 1994. In 1994 the authorities intervened in 13 of 47 more than 18 percent of GDP. banks that held 50 percent of deposits, and in 1995 in five additional banks. Middle East and North Africa Economy Scope of crisis Estimated losses or costs Algeria 1990-92 Share of nonperforming loans in the banking system reached 50 percent. Djibouti 1991-93 Two of six commercial banks ceased operations in 1991-92; other banks experienced difficulties. Egypt, Rep. of Early 1980s The government closed several large investment companies. Israel 1977-83 Almost the entire banking sector was affected, representing About 30 percent of GDP in 60 percent of stock market capitalization. The stock exchange 1983. closed for 18 days, and bank share prices fell more than 40 percent. Kuwait 1980s An estimated 40 percent of loans were nonperforming by 1986. Lebanon 1988-90 Four banks became insolvent. Eleven had to resort to Central Bank lending. 40 Managing the Real and Fiscal Effects of Banking Crises Systemic Banking Crises Middle East and North Africa-continued Economy Scope of crisis Estimated losses or costs Morocco Early l 980s Banking sector experienced solvency problems. Yemen, Rep. of 1996-? Banks suffered from extensive nonperforming loans and heavy foreign currency exposure. Europe and Central Asia Economy Scope of crisis Estimated losses or costs Turkey 1982-85 Three banks were merged with the state-owned Agriculture Rescue costs totaled Bank and then liquidated; two large banks were restructured. 2.5 percent of GNP. Transition economies Economy Scope of crisis Estimated losses or costs Albania 1992-? After the July 1992 cleanup, 31 percent of "new" banking system loans were nonperforming. Some banks faced liquidity problems due to a logjam of interbank liabilities. Armenia 1994-96 Starting in August 1994, the Central Bank closed half of active banks. Large banks continued to suffer from high nonperforming loans. The savings bank was financially weak. Azerbaijan 1995-? Twelve private banks closed; three large state-owned banks deemed insolvent; one large state-owned bank faced serious liquidity problems. Bosnia and Herzegovina 1992-present Banking system suffers from high nonperforming loans due to the breakup of the former Yugoslavia and the civil war. Bulgaria 1990s In 1995 an estimated 75 percent of banking system loans were By early 1996 the sector had a substandard. The banking system experienced a run in early negative net worth equal to 1996. The government then stopped providing bailouts, 3 percent of GDP. prompting the closure of 19 banks accounting for one-third of sector assets. Surviving banks were recapitalized by 1997. Croatia 1996 Five banks accounting for about half of banking system loans were deemed insolvent and taken over by the Bank Rehabilitation Agency. Episodes of Systemic and Borderline Banking Crises 41 Systemic Banking Crises Transition economies-continued Economy Scope of crisis Estimated losses or costs Czech Republic 1991-? Several banks have closed since 1993. In 1994-95, 38 percent Through 1994, 12 percent of of banking system loans were nonperforming. GDP was spent on bank support. Estonia 1992-95 Insolvent banks accounted for 41 percent of financial system Recapitalization outlays assets. Five banks' licenses were revoked, and two major for new entity totaled bankswere merged and nationalized. Two other large banks 300 million kroon, or were merged and converted to a loan recovery agency. In 1994 1.4 percent of 1993 GDP. the Social Bank, which controlled 10 percent of financial system assets, failed. Georgia 1991-? Most large banks virtually insolvent. About one-third of banking system loans were nonperforming. Hungary 1991-95 In the second half of 1993 eight banks-accounting for Resolution costs estimated 25 percent of financial system assets-were deemed insolvent. to total 10 percent of GDP. Kyrgyz Republic 1990s Some 80-90 percent of banking system loans doubfful. Four small commercial banks closed in 1995. Latvia 1995-present Between 1994 and 1999, 35 banks saw their license revoked, In 1995 the negative net worth were closed, or ceased operations. of the banking system was estimated at $320 million, or 7 percent of GDP. Aggregate banking system losses in 1998 estimated at 100 million lats ($172 million), about 3 percent of GDP. Lithuania 1995-96 In 1995, of 25 banks, 12 small banks were liquidated, 3 private banks (accounting for 29 percent of banking system deposits) failed, and 3 state-owned banks were deemed insolvent. Macedonia, FYR 1993-94 About 70 percent of banking system loans were nonperforming. Costs of banking system The government took over banks' foreign debt and closed the rehabilitation, obligations from second largest bank. assumption of external debt, liabilities regarding frozen foreign exchange, and contingent liabilities in banks together estimated at 32 percent of GDP. 42 Managing the Real and Fiscal Effects of Banking Crises Systemic Banking Crises Transition economies-continued Economy Scope of crisis Estimated losses or costs Poland 1990s In 1991 seven of nine treasury-owned commercial banks- In 1993 recapitalization costs accounting for 90 percent of credit-the Bank for Food Economy, were $750 million for the and the cooperative banking sector experienced solvency seven commercial banks and problems. $900 million for the Bank for Food Economy and the cooperative banking sector, for a total equivalent to 2 percent of GDP Romania 1990-present In 1998 nonperforming loans reached 25-30 percent in the six The Agricultural Bank was main state-owned banks. recapitalized on a flow basis. In 1998 the Central Bank injected $210 million (0.6 percent of GDP) in Bancorex, the largest state bank, and in 1999 another $60 million. Russian Federation 1995 In August 1995 the interbank loan market stopped working due to concerns about connected lending in many new banks. 1998-99 Nearly 720 banks, or half of those now operating, were deemed In 1999 bailout costs were insolvent. These banks accounted for 4 percent of sector assets estimated at $15 billion, and 32 percent of retail deposits. According to the Central Bank, or 5-7 percent of GDP 18 banks holding 40 percent of sector assets and 41 percent of household deposits are in serious difficulties and will require rescue by the state. Slovak Republic 1991-present In 1997 unrecoverable loans were estimated at 101 billion crowns, or about 31 percent of loans and 15 percent of GDP. Slovenia 1992-94 Three banks-accounting for two-thirds of banking system Recapitalizations cost $1.3 assets-were restructured. billion. Ukraine 1997-98 By 1997, 32 of 195 banks were being liquidated, while 25 others were undergoing financial rehabilitation. Bad loans accounted for 50-65 percent of assets even in some leading banks. In 1998 banks were further hit by the government's decision to restructure government debt. Episodes of Systemic and Borderline Banking Crises 43 Systemic Banking Crises Industrial countries Economy Scope of crisis Estimated losses or costs Finland 1991-94 Savings banks badly affected; government took control of Recapitalization costs three banks that together accounted for 31 percent of system amounted to 11 percent of deposits. GDP. Japan 1990s Banks suffered from sharp decline in stock market and real In 1996 rescue costs were estate prices. In 1995 the official estimate of nonperforming estimated at more than loans was 40 trillion yen ($469 billion, or 10 percent of GDP). $100 billion. In 1998 the An unofficial estimate put nonperforming loans at $1 trillion, government announced the equivalent to 25 percent of GDP Banks made provisions for Obuchi Plan, which provided some bad loans. At the end of 1998 banking system 60 trillion yen ($500 billion, nonperforming loans were estimated at 88 trillion yen or 12 percent of GDP) in ($725 billion, or 18 percent of GDP). In 1999 Hakkaido public funds for loan losses, Takushodu bank was closed, the Long Term Credit Bank was bank recapitalizations, and nationalized, Yatsuda Trust was merged with Fuji Bank, and depositor protection. Mitsui Trust was merged with Chuo Trust. Norway 1987-93 The Central Bank provided special loans to six banks suffering Recapitalization costs totaled from the recession of 1985-86 and from problem real estate 8 percent of GDP. loans. The state took control of the three largest banks (with 85 percent of banking system assets, whose loan losses had wiped out capital), partly through a Government Bank Investment Fund (5 billion kroner), and the state-backed Bank Insurance Fund had to increase capital to 11 billion kroner. Spain 1977-85 In 1978-83, 24 institutions were rescued, 4 were liquidated, Estimated bank losses were 4 were merged, and 20 small and medium-size banks were equivalent to about 17 percent nationalized. These 52 banks (of 110), representing 20 percent of GNP. of banking system deposits, were experiencing solvency problems. Sweden 1991-94 Nordbanken and Gota Bank, accounting for 22 percent of Recapitalization costs totaled banking system assets, were insolvent. Sparbanken Foresta, 4 percent of GDP accounting for 24 percent of banking system assets, intervened. Overall, five of the six largest banks, accounting for more than 70 percent of banking system assets, experienced difficulties. 44 Managing the Real and Fiscal Effects of Banking Crises Borderline and Smaller (Nonsystemic) Banking Crises Sub-Saharan Africa Economy Scope of crisis Estimated losses or costs Angola 1991-present Two state-owned commercial banks have experienced solvency problems. Botswana 1994-95 One problem bank was merged in 1994, a small bank was Recapitalizing the National liquidated in 1995, and the state-owned National Development Development Bank cost Bank was recapitalized. 0.6 percent of GDP. Ethiopia 1994-95 A govemment-owned bank was restructured, and its nonperforming loans were taken over by the government. Gabon 1995-? One bank was temporarily closed in 1995. Gambia, The 1985-92 In 1992 a government bank was restructured and privatized. Ghana 1997-present Nonperforming loans increased sharply in 1997, from 16 percent to 27 percent. Two state-owned commercial banks-accounting for 34 percent of the market-are in bad shape. Three banks, accounting for 4 percent of deposits, are insolvent. Kenya 1996-? At the end of 1996 nonperforming loans reached 19 percent. Lesotho 1988-? One of four commercial banks suffered from large nonperforming loans. Mauritius 1996 The Central Bank closed 2 of 12 commercial banks for fraud and other irregularities. Nigeria 1997 Distressed banks accounted for 4 percent of banking system assets. Rwanda 1991-? One bank, with a well-established network, closed. South Africa 1977 Trust Bank experienced problems 1989-? Some banks are experiencing problems. Episodes of Systemic and Borderline Banking Crises 45 Borderline and Smaller (Nonsystemic) Banking Crises Sub-Saharan Africa-continued Economy Scope of crisis Estimated losses or costs Tunisia 1991-95 In 1991 most commercial banks were undercapitalized. During 1991-94 the banking system raised equity equivalent to 1.5 percent of GDP and made provisions equivalent to another 1.5 percent. Thus recapitalization through 1994 required at least 3 percent of GDP. East and South Asia Economy Scope of crisis Estimated losses or costs Brunei Mid-1980s-1990s Several financial firms failed. The second largest bank failed in 1986. In 1991, 9 percent of loans were past due. Hong Kong, China 1982-83 Nine deposit-taking companies failed. 1983-86 Seven banks or deposit-taking institutions were liquidated or taken over. 1998 One large investment bank failed. India 1993-present Nonperforming assets reached 11 percent in 1993-94. Nonperforming assets of the 27 public banks estimated at 20 percent in 1995. At the end of 1998 nonperforming loans estimated at 16 percent. Indonesia 1994 Nonperforming assets equal to more than 14 percent of banking Recapitalization costs for system assets, with more than 70 percent in state banks. five state banks amounted to nearly 2 percent of GDP. Lao People's Dem. Rep. Early 1990s Some banks experienced problems. Recapitalization of state-owned commercial banks amounted to 1.5 percent of GDP. Malaysia 1985-88 Insolvent institutions accounted for 3 percent of financial system Reported losses equivalent deposits; marginally capitalized and possibly insolvent to 5 percent of GNP. institutions accounted for another 4 percent. Myanmar 1996-? The largest state-owned commercial bank reported to have large nonperforming loans. 46 Managing the Real and Fiscal Effects of Banking Crises Borderline and Smaller (Nonsystemic) Banking Crises East and South Asia-continued Economy Scope of crisis Estimated losses or costs Papua New Guinea 1989-? Some 85 percent of savings and loan associations have ceased operations. Singapore 1982 Commercial banks' nonperforming loans rose to about $200 million, or 0.6 percent of GDP. Taiwan, China 1983-84 Four trust companies and eleven cooperatives failed. 1995 Failure of credit cooperative Changua Fourth in late July sparked runs on other credit unions in central and southem Taiwan. Latin America and the Caribbean Economy Scope of crisis Estimated losses or costs Costa Rica 1994? One large state-owned commercial bank was closed in December 1994. The ratio of overdue loans (net of provisions) to net worth in state commercial banks exceeded 100 percent in June 1995. Guatemala 1990s Two small state-owned banks had high nonperforming assets; these banks discontinued operations in the eary 1990s. Trinidad and Tobago 1982-93 In the early 1980s several financial institutions experienced solvency problems, resulting in the merging of three govemment- owned banks in 1993. Venezuela, RB Late 1970s and Notable bank failures included Banco Nacional de Descuento 1980s (1978), BANDAGRO (1981), Banco de los Trabajadores de Venezuela (1982), Banco de Comercio (1985), BHCU (1985), BHCO (1985), and Banco Lara (1986). Middle East and North Africa Economy Scope of crisis Estimated losses or costs Egypt, Rep. of 1991-95 Four public banks were given capital assistance. Episodes of Systemic and Borderline Banking Crises 47 Borderline and Smaller (Nonsystemic) Banking Crises Middle East and North Africa-continued Economy Scope of crisis Estimated losses or costs Jordan 1989-90 The third largest bank failed in August 1989, The central bank provided overdrafts equivalent to 10 percent of GDP to meet a run on deposits and allow banks to settle foreign obligations. Europe and Central Asia Economy Scope of crisis Estimated losses or costs Turkey 1994 Three banks failed in April 1994. Through June 1994 the authorities spent 1 percent of GDP. Transition economies Economy Scope of crisis Estimated losses or costs Belarus 1995-? Many banks undercapitalized; forced mergers burdened some banks with poor loan portfolios. Estonia 1998 Three banks failed in 1998: Maapank (Agricultural Bank), which Maapank's losses reached accounted for 3 percent of banking system assets, and two $500 million. smaller banks, EVEA and ERA. Tajikistan 1996-? One of the largest banks is insolvent, one small bank has been closed, and another (out of 17) is in the process of liquidation. Industrial countries Economy Scope of crisis Estimated losses or costs Australia 1989-92 Two large banks received capital from the government to cover Rescuing state-owned banks losses. Nonperforming loans rose to 6 percent of assets in was estimated to cost 1991-92. 2 percent of GDP. Canada 1983-85 Fifteen members of the Canadian Deposit Insurance Corporation, including two banks, failed. 48 Managing the Real and Fiscal Effects of Banking Crises Borderline and Smaller (Nonsystemic) Banking Crises Industrial countries-continued Economy Scope of crisis Estimated losses or costs Denmark 1987-92 Cumulative loan losses over 1990-92 were 9 percent of loans; 40 of the 60 problem banks were merged. France 1994-95 Credit Lyonnais experienced serious solvency problems. According to unofficial estimates, losses totaled about $10 billion, making it the largest bank failure up to that time. Germany Late 1970s So-called Giroinstitutions faced problems. Greece 1991-95 Localized problems required significant injections of public funds into specialized lending institutions. Iceland 1985-86 One of three state-owned banks became insolvent and was eventually privatized in a merger with three private banks. 1993 The government was forced to inject capital into one of the largest state-owned commercial bank after it suffered serious loan losses. Italy 1990-95 During 1990-94, 58 banks (accounting for 11 percent of lending) were merged with other institutions. New Zealand 1987-90 One large state-owned bank accounting for one-quarter of The bank required a capital banking assets experienced serious solvency problems due injection equal to 1 percent of to high nonperforming loans. GDP. United Kingdom 1974-76 'Secondary Banking Crisis." 1980s and 1990s Notable bank failures included Johnson Matthey (1984), Bank of Credit and Commerce International (1991), and Barings (1995). United States 1984-91 More than 1,400 savings and loan institutions and 1,300 Cleaning up savings and loan banks failed. institutions cost $180 billion, or 3 percent of GDR Source: World Bank data and staff; Sheng 1995; World Bank 1989; Baer and Klingebiel 1995; Vittas 1992; Sundarajan and Balino 1991; Rodriguez 1994; Morris and others 1990; Blass and Grossman 1995; Fleming and Talley 1996; Lindgren, Garcia, and Saal 1996; Fleming, Chu, and Bakker 1996. Episodes of Systemic and Borderline Banking Crises 49 References Baer, Herbert, and Daniela Klingebiel. 1995. "Systemic Risk When Depositors Bear Losses: Five Case Studies." In G. G. Kaufhnan, ed., Research in Financial Services: Private and Public Policy. vol. 7. Greenwich, Conn.: JAI Press. Blass, Asher A., and Richard S. Grossman. 1995. "A Costly Guarantee? The 1983 Israel Bank Shares Crisis Revisited." Discussion Paper 95.05. Maurice Falk Institute for Economic Research in Israel, Jerusalem. Caprio, Gerard, and Daniela Klingebiel. 1996. "Bank Insolvencies: Cross-country Experience." Policy Research Working Paper 1620. World Bank, Washington, D.C. 1997. "Bank Insolvency: Bad Luck, Bad Policy, or Bad Banking?" In Michael Bruno and Boris Pleskovic, eds., Annual World Bank Conference on Development Economics 1996. Washington, D.C.: World Bank. Fleming, Alex, and Samuel Talley. 1996. "The Latvian Banking Crisis: Lessons Learned." Policy Research Working Paper 1590. World Bank, Washington, D.C. Fleming, Alex, Lily Chu, and Marie-Renee Bakker. 1996. "The Baltics-Banking Crises Observed." Policy Research Working Paper 1647. World Bank, Washington, D.C. Lindgren, Carl-Johan, Gillian Garcia, and Matthew I. Saal. 1996. Bank Soundness and Macroeconomic Policy. Washington, D.C.: International Monetary Fund. Morris, Felipe, Mark Dorfmian, Jose Pedro Ortiz, and Maria Claudio Franco. 1990. Latin America s Banking Systems in the 1980s. World Bank Discussion Paper 81. Washington, D.C. Rodriguez, Carlos Alfredo. 1994. "Argentina: Fiscal Disequilibria Leading to Hyperinflation." In William Easterly, Carlos Alfredo Rodriguez, and Klaus Schmidt-Hebbel, eds., Public Sector Deficits and Macroeconomic Performance. New York: Oxford University Press. Rojas-Suarez, Liliana, and Steven Weisbrod. 1995. "Banking Crises in Latin America: Experiences and Issues." In Ricardo Hausmann and Liliana Rojas-Suarez, eds., Banking Crises in Latin America. Baltimore, Md.: The Johns Hopkins University Press. Sheng, Andrew, ed. 1996. Bank Restructuring: Lessons from the 1980s. Washington, D.C.: World Bank. Sundarajan, Vasudevan, and Tomas Jose T. Balino, eds. 1991. Banking Crises: Structural Weaknesses, Support Operations, and Economic Consequences. Washington, D.C: International Monetary Fund. Vittas, Dimitri, ed. 1992. Financial Regulation: Changing the Rules of the Game. EDI Development Studies. Washington, D.C.: World Bank Economic Development Institute. World Bank. 1989. World Development Report 1989: Financial Systems and Development. New York: Oxford University Press. Recent World Bank Discussion Papers (continued) No. 384 Integrating Social Concerns into Private Sector Decisionmaking: A Review of Corporate Practices in the Mining, Oil, and Gas Sectors. Kathryn McPhail and Aidan Davy No. 385 Case-by-Case Privatization in the Russian Federation: Lessonsfrom International Experience. Harry G. Broadman, editor No. 386 Strategic Managementfor Government Agencies: An Institutional Approachfor Developing and Transition Economies. Navin Girishankar and Migara De Silva No. 387 The Agrarian Economies of Central and Eastern Europe and the Commonwealth of Independent States: Situation and Perspectives, 1997. Csaba Csaki and John Nash No. 388 China: A Strategyfor International Assistance to Accelerate Renewable Energy Development. Robert P. Taylor and V. Susan Bogach No. 389 World Bank HIV/AIDS Interventions: Ex-ante and Ex-post Evaluation. Julia Dayton No. 390 Evolution of Agricultural Services in Sub-Saharan Africa: Trends and Prospects. V. Venkatesan and Jacob Kampen No. 391 Financial Incentivesfor Renewable Energy Development: Proceedings of an Interntational Workshop, February 17-21. 1997, Amsterdam, Netherlands. E. Scott Piscitello and V. Susan Bogach No. 392 Choices in Financing Health Care and Old Age Security: Proceedings of a Conference Sponsored by the Institute of Policy Studies, Singapore, and the World Bank, November 8,1997. Nicholas Prescott, editor No. 393 Energy in Europe and Central Asia: A Sector Strategyfor the World Bank Group. Laszlo Lovei No. 394 Kyrgyz Republic: Strategyfor Rural Growth and Poverty Alleviation. Mohinder S. Mudahar No. 395 School Enrollment Decline in Sub-Saharan Africa: Beyond the Supply Constraint. Joseph Bredie and Girindre Beeharry No. 396 Transforming Agricultural Research Systems in Transition Economies: The Case of Russia. Mohinder S. Mudahar, Robert W. Jolly, and Jitendra P. Srivastava No. 398 Land Reform and Farn Restructuring in Moldova: Progress and Prospects. Zvi Lerman, Csaba Csaki, and Victor Moroz No. 400 Russian Enterprise Reform: Policies to Further the Transition. Harry G. Broadman, editor No. 401 Russian Trade Policy Reformfor WTO Accession. Harry G. Broadman, editor No. 402 Trade, Global Policy, and the Environment. Per G. Gredriksson, editor No. 403 Ghana: Gender Analysis and Policymakingfor Development. Shiyan Chao, editor No. 404 Health Care in Uganda: Selected Issues. Paul Hutchinson, in collaboration with Demissie Habte and Mary Mulusa No. 405 Gender-Related Legal Reform and Access to Economic Resources in Eastern Africa, Gita Gopal No. 406 The Private Sector and Pow7er Generation in China. Energy and Mining Sector Unit, East Asia and Pacific Region, World Bank No. 407 Economic Growth With Equity: Ukrainian Perspectives. John Hansen, editor No. 408 Economic Growth with Equity: Which Strategyfor Ukraine? John Hansen and Diana Cook No. 409 East Asian Corporations: Heroes or Villains? Stijn Claessens, Simeon Djankov, and Larry H. P. Lang No. 411 Making the Transition Workfor Women in Europe and Central Asia. Marnia Lazreg, editor No. 412 Intellectual Property Rights and Economic Development. Carlos A. Primo Braga, Carsten Fink, and Claudia Paz Sepulveda No. 413 Management and Resolution of Banking Crises: Lessonsfrom the Republic of Korea and Mexico. Jose De Luna-Martinez No. 414 Liquefied Natural Gas in China: Optionsfor Markets, Institutions, and Finance. Dean Girdis, Stratos Tavoulareas, and Ray Tomkins No. 416 Fostering Competition in China's Power Markets. Noureddine Berrah, Ranjit Lamech, and Jianping Zhao No. 417 Hungary: Modernizing the Subnational Government System. Mihaly Kopanyi, Samir El Daher, Deborah Wetzel, Michel Noel, and Anita Papp No. 419 Measuring and Apportioning Rentsfrom Hydroelectric Power Developments. Mitchell Rothman No. 420 Financing of Private Hydropower Projects. Chris Head No. 421 Free Trade Area Membership as a Stepping Stone to Development: The Case of ASEAN. Emiko Fukase and Will Martin No. 422 Environmental Health: Bridging the Gaps. James A. Listorti and Fadi M. Doumani Ru.') THE WORLD BANK 1818 H Street, N.W Washington, D.C. 20433 USA Tlelephone: 202-477-1234 Facsimile: 202-477-6391 I nternct: se -%Nvwor1dbank.org E-mail: feedback(FNvorldbank.org 051 116 ISBN 0-8213-5056-0