WORLD BANK DISCUSSION PAPER NO. 429 WDP429 Work in progress for Public discussion January 2002 Structural Adjustment in the Transition Case Studies from Albania, Azerbayjan, Kyrgyz Republic, and Moldova Paul J. Siegoelbaum Khaled Sherif iLic/hael Borish George Clarke Recent World Bank Discussion Papers No. 351 From Universal Food Subsidies to a Self-Targeted Program: A Case Study in Tunisiani 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 Serv6n 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 Nlon-Bank Financial Instit utions: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 Conferenice 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 Economtiy. 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 Europeani 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-Counltry Study. Judy L. Baker No. 367 Easing Barriers to Movement of Plant Varieties for Agricultural Developmenit. 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 Microfintanice 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 Founldation for a Socialist Economy. Jiwei Lou No. 375 Design of Social Funds: Participation, Demand Orientationz, 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 Transitioni 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 Instruments for Environmental Policymaking in Latin America and the Caribbean: Lessons from Eleven Countries. Richard M. Huber, Jack Ruitenbeek, and Ronaldo Seroa da Motta. No. 382 Public Expenditure Reform under Adjustment Lending: Lessons from 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. 429 Structural Adjustment in the Transition Case StudiesfromAlbania, Azerbay/an, Kyrgyz Republic, and Moldova Paul J. Siegelbaum Khaled Sherif Michael Borish George Clarke The World Bank Washington, D. C Copyright © 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 comrnment 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: 0-8213-5065-X ISSN: 0259-210X Paul J. Siegelbaum is a Sector Director in the Private and Financial Sectors Development Sector Unit of the World Bank. Khaled Sherif is a Sector Manager in the Private and Financial Sectors Development Sec- tor Unit of the World Bank. Michael Borish is a Consultant in the Private and Financial Sectors Develop- ment Sector Unit of the World Bank and George Clarke is an Economist in the Development Research Group of the World Bank. Cover Photo: Agriculture, Uzbekistan, 1994, Curt Camemark, World Bank. Library of Congress Cataloging-in-Publication Data has been applied for. TABLE OF CONTENTS PREFACE ..................................................... V ACKNOWLEDGMENTS ................................... X ABBREVIATIONS AND ACRONYMS .................hl............. XI EXECUTIVE SUMMARY ............................................................... I CHAPTER ONE: BRIEF PROFILE OF THE COUNTRIES SELECTED 13 CHAPTER TWO: FINANCIAL SECTOR CHALLENGES ................... 17 The Structural Environment for Banking ............................................................... 23 The Financial Condition of the Banking Sector ............................................................... 36 Banking Sector Restructuring and Recapitalization ........................................................... 57 The Non-Bank Financial Sector ............................................................... 59 Summary of Financial Sector Challenges, Issues and Risks ........................ ...................... 67 CHAPTER THREE: PRIVATE SECTOR DEVELOPMENT AND PRIVATlIZA0TION ...........................................mm ... ...m .. m7 3 Progress with Privatization ............................................................... 74 Approaches to Privatization ............................................................... 76 Characteristics of State-Owned and Private Enterprises .................................................... 78 Enterprise Reform ............................................................... 82 The Business Environment and Constraints on Enterprise Operations ............. ................. 86 Foreign Direct Investment ............................................................... 100 iii CHAPTER FOUR: INTER-SECTORAL ARREARS AND FINANCING ISSUES .3........................................................ 1 3 Soft Budget Constraints in the Transition Economies ...................................................... 114 Arrears ................................................................... 117 Barter ................................................................... 128 Arrears, Costs and Bank Losses ................................................................... 130 CHAPTER FIVE: LESSONS FROM THE EU ACCESSION CANDIDATES ........................................................m1. .................. 35 General Overview ................................................................... 135 Financial Sector Lessons ................................................................... 139 Private Sector Lessons ................................................................... 147 CHAPTER SIX: RECOMMENDATIONS AND CONCLUSIONS ......... 151 Financial Sector Development Recommendations ........................................................... 151 Private Sector Development Recommendations ............................................................... 158 CHAPTER SEVEN: OVERARCHING THEMES OF THE TRANSITION EXPERIENCE .........................................................1 63 BIBLIOGRAPHY ...................*.m .u.u.mmuuuuuuuu 1 75 iv PREFACE Ten years have now passed since countries of Central and Eastern Europe (CEE) and the former Soviet Union (FSU) began the transition from central planning to economies based, more or less, on market principles. During that period, there have been some impressive results and encouraging trends. In several countries, confidence in the future is strengthening, and there is cautious optimism that the worst of economic destabilization is over, that purchasing power is rising, and that productive employment and sustainable growth are beginning to materialize. This scenario would characterize a number of the Central European countries and the three Baltic states. This is not restricted to Central Europe and the Baltics, as developments in the Russian Federation itself in the last year have also been encouraging, as have some of the economic measures in a handful of other non-CIS countries. However, in most of the transition economies, including the Balkan states and the most CIS countries, conditions remain fragile. The economies of most of these countries are weak, public confidence is low, and political instability remains a major risk in several countries. These all undermine the development of well functioning civil society, a viable fiscal base, and stable markets. The Balkans have contended with war for most of the 1990s, while cross- border conflicts have affected several of the former Soviet states. More broadly, in most of the transition countries, corruption remains a rampant and costly problem. Direct investment is low, reflecting the unwillingness of foreign investors to commit resources in an environment of high political and commercial risk. Short-term patronage concerns have slowed or distorted the privatization process, causing unnecessary delays and adding costs to the economy in the form of budget-draining subsidies, as well as significant opportunity costs in the form of foregone investments, productivity and employment. As macroeconomic and structural weaknesses have constrained opportunities in the formal private sector, the weak fiscal base and limited technical capacity of public administration have rendered government poverty alleviation programs wholly ineffective, making the transition process more painful for the most vulnerable members of society. These conditions have been exacerbated by a weak judiciary with often anachronistic or unsynchronized laws and limited implementation capacity. Against this backdrop, banks and other intermediaries have found it difficult to assume risk when a stricter prudential framework has been introduced as a basis for establishing financial sector stability. Purpose and History of the Study This study presents recommendations on how countries that have struggled with the transition process in the areas of private and financial sector development might be able to shift gears and more rapidly achieve objectives. This study focuses on Albania, Azerbaijan, v the Kyrgyz Republic and Moldova (referred to in this study as the "subject countries"), which were selected because they are united by a number of characteristics which have severely impaired their ability to overcome structural and institutional obstacles to realize sustainable growth and poverty reduction. The four subject countries are among the poorest countries of the region and the world. All are low in tenns of real GDP, pointing to the intractability of many of the challenges they face. They are small in population, and were isolated until recently. All four countries total only 20 million people, and they were largely detached from the international marketplace until the transition process began. All four have faced at least one severe, dislocating crisis or shock, such as war, internal civil conflict, the economy-threatening collapse of pyramid schemes, contested borders, and/or banking crises. Most have experienced several of these shocks. Yet, as a group, these countries present a broad enough range of characteristics in terms of history, resource endowment, and proximity to markets to permit extrapolation of lessons learned to other countries and regions. They are typically viewed as either "late reformers" or "late performers" in terms of economic development and general competitiveness. While they have moved forward with numerous policy reforms, economic development levels are still frustratingly low. In general, it is likely to take many years of investment, training and policy reform for these four countries to achieve the level of comfort and stability that many other transition countries are expected to achieve by the end of this decade. By reviewing these four countries' reform efforts, the study seeks to identify lessons learned that could benefit existing and future reform efforts in these four countries and in other transition economies. The performance of the four subject countries is assessed on its own merits, with comparative references to the experience of countries of the Commonwealth of Independent States (CIS), the five first-tier EU accession candidates from the transition world' (referred to in this study as the "EU accession candidates") and, in some cases, with OECD G-24 countries. Most of the comparative assessment is relative to CIS and other transition country indicators. The study looks at results in the four subject countries and makes judgments about where improvements might be considered for future reform efforts. In some cases, it is fair to say that reforms proposed to date have been overly ambitious, poorly timed, or inadequately sequenced. In some cases, reforms may have been insufficiently tough, and in other cases wholly unrealistic. As noted earlier, the failure to achieve goals and objectives has numerous contributing factors. There are clear links in the chain that make the challenge greater due to the inter-sectoral nature of the problems studied. For example, bank lending has often been undermined by the weak environment for enforcing creditors' rights, including numerous These are the Czech Republic, Estonia, Hungary, Poland and Slovenia. Cyprus, the sixth first-tier candidate, is excluded from the analysis. vi obstacles to effective implementation of bankruptcy and collateral legislation. As this is a long and time-consuming process that requires a major overhaul of the legal framework and massive investment in new judicial capacity, it is possible that more targeted reforms - land privatization, pledge registries, specialized arbitration vehicles - would have been more practical, while broader and more ambitious challenges, such as the transformation of the legal incentive framework, proceeded over a longer period of time. At the same time, it would have been a mistake to call for an easing of standards in the financial sector simply to increase intermediation rates. The evidence shows that weak prudential standards leads to portfolio erosion, higher costs of intermnediation and, ultimately, a lack of access to resources. While the study seeks to generate useful lessons based on the diagnostics conducted, the authors do not presume that these lessons can be applied uncritically or mechanically to all other transition countries as well. Although the economic starting points for many transition countries may have been similar, divergence, based on cultural, geographic, economic and political factors began immediately after the collapse of socialism. Ten years later, there are noticeable and growing differences in results. While the five EU accession candidates are the principal external comparators for the four subject countries, this does not imply that they are perfect models. Divergence, as noted above, makes parallels difficult to draw. In some cases, other countries might have served as equally valid reference points for comparison. Nonetheless, the five EU accession candidates have been selected because of the progress they have made, and the relevance of their reforrn efforts to the efforts of other transition countries as they seek to achieve stable and competitive economies. Notwithstanding a host of mistakes made during the transition process, these five countries have been the most broadly successful of all the transition countries in moving towards international norms and economic performance. This should position them to assume the obligations of EU membership later in the decade. Considering where they started a decade ago, this constitutes success on many fronts. Thus, there are many valuable lessons to extract from the experiences of these countries for our four subject countries. Methodology Primary data were used as the main tool in assessing private sector development issues, while secondary data were used for an evaluation of banking and financial services. To the extent that macroeconomic issues enter the analysis, these data were primarily sourced from the World Bank, IMF and EBRD. In the various private sector assessments, the World Business Environment Survey (WBES)2 provided primary data to review the business and investment climate, enterprise performance and restructuring, the role of the state, constraints to financing, and general conditions influencing competition. 2 The World Business Environment Survey was a World Bank project conducted in partnership with other regional development institutions. For a more complete description of the survey, see Box 3.3. vii Data on the private and financial sectors are derived from three main sources: * Official and externally distributed publications by the World Bank (ECA Privatization and Restructuring Database and various country reports for Albania, Azerbaijan, the Kyrgyz Republic and Moldova); IMF (IFS and country reports); EBRD (Transition Reports for years 1998-2000); European Union (EU-TACIS Country Economic Trends quarterlies for Azerbaijan and Moldova); UNDP (annual Human Development Reports for Azerbaijan, the Kyrgyz Republic and Moldova); bank rating agencies (Thomson Bankwatch for Azerbaijan, and Moody's for Albania); and OECD (annual reports). * Publications and websites of government agencies, including the central banks of Albania, the Kyrgyz Republic and Moldova (annual, quarterly, and monthly reports on the financial sector); the Privatization Authorities of Azerbaijan, the Kyrgyz Republic and Moldova; and the State Statistical Committees of Azerbaijan and Kyrgyz Republic. * Internal World Bank data on the private and financial sectors of Europe and Central Asia. All regional macroeconomic and financial indicators have been calculated as either simple arithmetic averages, or sums of country indicators depending on the nature of the indicator (e.g., percentage, dollar value per bank, total dollar value). The averages have been calculated using all data that were available. When the countries included in the same indicator for different years were different due to data availability, footnotes have been provided. Amounts presented in US dollar-equivalent terms are signified by the term "US$" and have been converted on the basis of year-end dollar exchange rates for "stock" figures and average exchange rates for "flow" figures, unless already available in US dollars (i.e., balance of payments figures). Despite efforts to create comprehensive data sets for 1992-2000, gaps still remain. These particularly relate to private sector indicators and for the early years of transition, due to the tendency of the transition counties not to maintain data of particular relevance to this topic until several years into the transition. For these reasons, the sections of the study on arrears and the business environment tend to rely on the enterprise survey data, whereas financial sector analyses tend to rely more on the sources noted above. Finally, survey data were utilized for the private sector sections of the report to allow for cross-country comparisons. In the end, recognizing the substantial limitations of surveys, the findings and results were generally cross-checked with adequate and reliable secondary sources (e.g., IMF, World Bank, EBRD, EU-TACIS) and academic papers. viii Organization of the Study The study consists of seven chapters. Chapter 1 provides a basic profile of the four subject countries and where they fit relative to other transition countries and the reform process in general. Chapter 2 focuses on the financial sector, including legal and regulatory issues, privatization and competition, trends in capital structure and intermediation, earnings and asset quality, restructuring and recapitalization, and trends in non-bank financial services. Chapter 3 reviews private sector and privatization trends, including ownership, governance, performance and competitiveness. Chapter 4 assesses subsidies, transfers, arrears, and the role of soft budget constraints in undermining economic performance and financial intermediation. Chapter 5 provides a composite of lessons learned from the five EU accession candidates, and some recommendations on how these lessons can contribute to the efforts of the four subject countries. Chapter 6 summarizes key findings and provides recommendations for the four subject countries on how to move forward in the future with more effective reforms. Chapter 7 articulates a number of development and strategic themes that have emerged from the study that are applicable to a considerably broader audience than the four subject countries. ix ACKNOWLEDGMENTS The authors would like to thank Paul J. Siegelbaum, Director, ECSPF, for his enthusiastic support and encouragement to produce a cross-country and cross-regional study focusing on some of the poorer transition countries, and his substantive insights and critique on findings and recommendations. He has served as editor of this piece. Alexander Pankov and Anahit Adamyan provided research assistance and compiled comprehensive data sets for this study. The authors would not have been able to produce this report without their thorough research, effective output, and tireless effort. Valuable comments and data were provided for Albania by Hormoz Aghdaey, Greta Minxhozi and Sylvia Minotti. The team received helpful comments on Azerbaijan from Judy O'Connor, Nicolas Mathieu, Tunc Uyanik and Carlo Segni. Marina Prodan provided useful data and information on Moldova. Neesham Kranz was helpful in providing input from the Financial Sector Department on a range of information needs. Alon Laniado helped with collecting the data for OECD economies. The team wishes to thank them all for their input, their spirit of cooperation, and their cordial and constructive assistance, all of which made a major contribution to the quality of the study. Work on the private sector-related sections benefited from the provision of data from the WBES. For this, the team wishes to thank Andrew Stone and Luke Haggarty for their constructive and generous assistance. x ABBREVIATIONS AND ACRONYMS ADR - American Depository Receipts BIS - Bank for International Settlements CAR - Capital Adequacy Ratio CEB - Central Europe and the Baltics (i.e., the Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland, Slovenia, and the Slovak Republic) CEE - Central and Eastern Europe CIS - Commonwealth of Independent States (Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, the Kyrgyz Republic, Moldova, the Russian Federation, Tajikistan, Turkmenistan, Ukraine and Uzbekistan). CPI - Consumer Price Index Early applicants to the - Czech Republic, Estonia, Hungary, Poland, and EU Slovenia. EBRD - European Bank for Reconstruction and Development ECA - Europe and Central Asia Regional Office EMU - European Monetary Union EU - European Union FDI - Foreign Direct Investment FSU - Former Soviet Union FYR Macedonia - Former Yugoslav Republic of Macedonia. GDP - Gross Domestic Product GDR - Global Depository Receipts GNP - Gross National Product IAIS - International Association of Insurance Supervisors IAS - International Accounting Standards IBA - International Bank of Azerbaijan ICRG - International Country Risk Guide IFS - International Financial Statistics IMF - International Monetary Fund INSIG - Albanian Insurance Company IOSCO - International Organization of Securities Commissions IPF - Investment Privatization Funds ISA - International Standards of Audit KSE - Kyrgyz Republic Stock Exchange LSMS - Living Standards Measurement Study MEBO - Management-Employee Buyouts NBFI - Non-Bank Financial Institutions NBK - National Bank of the Kyrgyz Republic NBM - National Bank of Moldova NCB - National Commercial Bank (of Albania) xi NGO - Non-Governmental Organizations NPL - Non-Performing Loans PAYG - Pay As You Go PRGF - Poverty Reduction and Growth Facility PSA - Production Sharing Agreement ROA - Return on Assets ROE - Return on Equity RTGS - Real Time Gross Settlement OECD - Organization for Economic Cooperation and Development OTC - Over-the-Counter Exchange SEE - South Eastern Europe (Albania, Bulgaria, Croatia, FYR Macedonia, and Romania; sometimes includes Bosnia and Herzegovina and Federal Republic of Yugoslavia, depending on the source) SME - Small- and Medium-sized Enterprises SOCAR - State Oil Company (Azerbaijan) SOE - State-owned Enterprises TACIS - Technical Assistance to Commonwealth of Independent States UNDP - United Nations Development Program WBES - World Business Environment Survey xii EXECUTIVE SUMMARY This study reviews the performance of four transition countries - Albania, Azerbaijan, the Kyrgyz Republic and Moldova - in the areas of private and financial sector development. It evaluates progress in the subject countries, identifies notable achievements as well as remaining challenges, and extracts lessons from other transition countries that could benefit future reform efforts in these four countries and others in similar positions. Importantly, this study seeks to provide suggestions for alternative approaches that may help to put these four countries on the path to effective, sustainable reform. Some of these recommendations might also be useful for other transition countries, including most countries in the Balkans and the CIS. Albania, Azerbaijan, the Kyrgyz Republic and Moldova were selected for this study for several reasons. They are among the poorest countries in the region, and many of the challenges that they face appear intractable. They are small in population, with only a total of 20 million for all four countries. All four countries were largely detached from the international marketplace until the transition began. All four countries have faced at least one severe, dislocating crisis or shock, such as war, internal civil conflict, the economy- threatening collapse of pyramid schemes, contested borders and/or banking crises. Most have faced several of these shocks. Yet, as a group, they present a broad enough range of characteristics in terms of history, resource endowment, and proximity to markets to allow lessons to be extrapolated to other countries and regions. They are typically viewed as either 'late reformers" or "late performers" in terms of economic development and general competitiveness. While they have moved forward with numerous policy reforms, economic development levels are still frustratingly low. Several themes have emerged from the study that point to the factors that have contributed to these countries' weak performances. These are summarized below. Selectivity. Even a cursory review of the findings set out in this study on the financial and enterprise sectors of the four subject countries indicates how important selectivity will be in shaping their future reform strategies. Literally hundreds of policy, implementation and institutional actions can be identified in these two sectors alone, even before considering the equally complex and daunting problems in other sectors of their economies. However, first among the generic problems identified in the four subject countries is the lack of capacity to implement these reforms. Even in the most advanced political and economic systems, it is unrealistic to expect fundamental change in more than two or three areas in a given year. As a result, the first challenge of shaping the reform strategy for the near and medium-term will be to decide on priorities and the allocation of time, effort and resources for what is most critical. In this regard, selectivity has a number of dimensions. First, there must be a decision as to which sectors of the economy are the most promising, and the most likely to generate the greatest positive response to reform. It is risky to recommend this approach, given the failed I efforts at the "industrial strategies" which were popular in developing countries in the 1 970s and early 1980s. Nevertheless, most countries simply do not have the capacity to do everything in all sectors at the same time, and such an ambitious approach can undermine prospects for other promising initiatives. Second, reform must be phased, so that change moves from the most basic to the more complex, without skipping intermediate steps. One good example is housing finance, which is virtually impossible to develop until functioning banking sectors, securities markets, land markets, and land use regulations are all reasonably in place. Similar arguments can be made about expending energy on securities markets while the banking system cannot support them. Third, difficult judgments must be made as to the political economy of alternative reform possibilities. If land privatization is simply not achievable, and alternatives cannot be found to unlock the collateral value of real estate, policy makers must carefully consider the utility of agricultural and rural reform and the risk of failure under such policy constraints. Critical Mass and Regional Association. There is little chance that the four subject countries will become competitive and commercially viable on their own, even in the long run. All of them face a multitude of problems that place them in a poor competitive position by global standards. Examples of international competitive There is little advantage are limited and vulnerable. For instance, the Kyrgyz Republic is often cited for the potential of its gold chance that the four sector, yet the output and earnings from this source are countries will expected to decline by 2004 and virtually end by 2008. become competitive Azerbaijan is probably the lone exception with its petroleum resources. However, there is little evidence to date that and commercially Azerbaijan's oil wealth has effectively benefited other sectors viable on their own. of the economy, aside from construction, a shortcoming which is common in other countries with similar resource endowments in an unbalanced economy. More positively, all four countries are fortunate to have viable rural sectors as a social safety net. In addition, Albanian residents benefit from high remittance flows from family members working abroad, which, at a minimum, provides a modest cushion that other countries lack. All four countries also have trade ties to neighboring countries that offer the potential to increase economic growth prospects over time. For instance, the number of Albanians in Italy, Greece and other European countries is high. The Azeri population in Iran is reported to be larger than the 8 million in Azerbaijan. Moldovan workers are currently welcome in the Russian Federation, and would presumably benefit if Romania made progress with its reform efforts. Only the Kyrgyz Republic appears to have little cordiality in its relations with larger neighbors, many of which have repeatedly put up trade barriers to stem the flow of cheap Kyrgyz Republic goods into their domestic markets. One of the main lessons from the five "first-wave" EU accession candidates is that the prospect of EU membership can provide an incentive to identify and adopt difficult policies that might otherwise have been too controversial. The same logic may apply to the four 2 countries studied, as well as to other countries in the Balkans and CIS. Promoting efforts to establish larger trade, customs, and even currency unions might help to enlarge markets in the limited areas where these countries are potentially competitive. While the currency issue would be more complex, forming trade and customs areas could increase the flow of goods, reduce transactions costs and barriers that are conducive to rent-seeking behavior, and provide opportunities for economies of scale and rationalization. Movement in this direction would require agreed criteria and joint monitoring and supervision. More importantly, these regional efforts could parallel norms adopted in other free trade areas, with an eye towards eventual linkage over time. For instance, efforts to link Albania with other Balkan markets should be structured to enhance legal trade that already flows through the region, largely on an informal basis. The Role of Government, Arrears, and Soft Budget Constraints. A core challenge of the transition is to reduce the role of the government from an all-encompassing presence towards a more professionally managed model in which it rationalizes its role, provides a high level of service delivery, restores public confidence in civil institutions, and plays an effective regulatory role in what is a predominantly market economy. To accomplish this transformation, the governments of the four countries covered by this study need to successfully resolve several interrelated problems. First, A core challenge of they need to work to improve financial discipline throughout their economies. Second, they need to operate the transition is to within a reasonable fiscal envelope, while addressing long- reduce the role of the term structural issues, providing needed services and governmentfrom an protecting vulnerable groups. Third, they need to operate professionally so that the public has more trust in these all-encompassing institutions, and are willing to pay taxes at reasonable rates presence to a for ongoing operations and the provision of needed public professionally goods. Each country has made progress in some of these areas, but not enough yet to trigger wholesale confidence in managed model the future. Enterprise arrears and soft budget constraints have been a significant problem in many transition economies, including Azerbaijan, the Kyrgyz Republic and Moldova. More often than not, this has recently been manifested more in enterprise arrears as some fiscal tightening has occurred to offset earlier soft budget constraints. In effect, this has served as a transfer from budget to off-budget loosening for uncompetitive state enterprises that have failed to restructure sufficiently to compete. As well as undermining the functioning of the market economy, poor financial discipline, which is reflected in the use of arrears and non-cash payments, has also affected government firlan By obscuring market transactions and prices and reducing enterprise liquidity, these practices make tax collection more difficult and tax evasion easier. Although arrears and the use of non-cash payments reflect poor financial discipline in the enterprise sector, they are a symptom of underlying problems rather than a direct problem in and of themselves. Consequently, some government actions designed to reduce stocks of arrears, such as netting 3 operations, might worsen, rather than relieve, problems by signaling that enterprises can expect bailouts rather than the imposition of hard budget constraints in response to weak financial discipline. Instead, governments need to enact policies that limit the accumulation of enterprise arrears, including privatizing state-owned enterprises and utilities, and forcing enterprises to adjust to the new conditions around them. Although improving financial discipline will improve the governments' fiscal situations in the long-term, additional steps will be needed to improve government finances in the short- and medium-term. One critical area that needs attention is improving public administration and increasing the public's confidence in civil institutions, which has been reduced by cynicism about the political process and corruption. The ability to resolve these issues is directly related to the ability of these countries to rationalize the role of government, to leverage off of partnerships with the private sector and the NGO community, to effectively prioritize services, and to improve methods of delivery. This also points to the need to privatize remaining state enterprises to reduce losses, improve financial discipline, possibly generate revenues, and allow the needed restructuring and professionalization of the civil service to proceed. This challenge particularly exists in the three CIS countries. It also remains important in Albania, where civil disorder prevailed just three years ago, and where public administration is still in the process of reformn and professionalization. (To its credit, Albania has privatized most of its state enterprises and banks at this point.) In this area, there is a clear role for donors to work closely in partnership with governments and with civil society at large, including NGOs, business associations, and other parties at all levels. Donors should mainly act as advisors, providing advice and counsel on what the return on expenditure would be under different scenarios, and to help recipients prioritize spending needs. While there is always a risk that excess borrower control will be overly politicized and used for patronage purposes, broadening the discussion to include non- government groups should help dilute the risk. Privatization and Foreign Investment. After more than 10 years of transition marked by an almost compete absence of capital investment, technological renewal or even maintenance, there is little hope that most remaining state enterprises can be privatized unless they are virtually given away. Few of the remaining "major" state enterprises in the four countries should be viewed as "strategic." More significantly, the retention of even minority interests distracts from the proper role of government and allows for anti-competitive interference by bureaucrats, the potential for future calls on overstressed state budgets to fund direct and indirect subsidies, the selective forgiveness of taxes, and the provision of other assistance. Beyond the direct costs and distractions is an equally important opportunity cost. By now, it is abundantly clear that private enterprises are more efficient and more profitable than state-owned firms in virtually all circumstances. By deferring privatization, these countries lose the potential for production and growth locked away in state enterprises, something that they can ill-afford to do. 4 Thus, large-scale privatization, including the privatization of infrastructure, needs to be accelerated, especially in Azerbaijan. When possible, privatization methods that promote concentrated outsider ownership and foreign participation should be favored over other methods. In the countries that have started to focus on cash privatization as their main form of privatization, overly Privatization methods optimistic assessments of enterprise value should not be that promote allowed to slow the privatization process. Privatization of concentrated outsider land is a critical goal to unlock perhaps the single most ownership and valuable source of wealth remaining in pubic ownership. Moreover, the disposition of residual enterprise interests foreign participation held by the state - leftovers from previous privatization should be avored programs - should be accelerated, even if that means the O simple cancellation and surrender of state shares. The best over other methods approach for the four governments to take, as Albania appears to be doing, is to quickly divest less valuable assets, while mapping out a clear timeline with transparent procedures for enterprises with some residual value. For enterprises and assets that nobody wants, liquidation rather than restoration is the recommended path. Given the limited resources of the domestic business community, foreign investment is critical to turn these economies around. In addition to providing direct financial inflows, foreign investment can benefit domestic economies by upgrading management skills, introducing new systems, and improving trade links. While there is a threat to protected businesses, FDI is uniformly considered a useful injection of skills that increase competitiveness and financing. Azerbaijan's oil and gas sector is the best example of a sector potentially attractive to foreign investment among the four countries. Azerbaijan's banking sector may also be attractive at some point, although this is currently limited to one bank. In Albania, the benefits of foreign investment have already been recognized in the banking sector and several other important industries. When the Savings Bank is privatized later in 2001-02, the banking sector will be 100 percent private, with foreign participation in all 13 institutions. Foreign investment in this sector has provided customers with better electronic access for payments and transfers, and in some cases, trade finance. The Kyrgyz Republic and Moldova have also become more open to foreign investment, with the latter registering some major successes in the last two years. However, important as foreign investment is, in practice, prospects for large inflows are generally limited. For example, in 2000, Albania reached its best FDI level to date, equivalent to US$143 million, or about US$40 per capita - far lower than in the leading economies in Central Europe and the Baltics. In practice, the four countries covered in this study are unlikely to attract much more than this in the near future, apart possibly from Azerbaijan. Consequently, foreign investment will not dramatically influence GDP or purchasing power in the short term. 5 Investment opportunities are often limited, even in strategic sectors. For example, no banks in Moldova or the Kyrgyz Republic can be viewed as "strategic." Further, although there are opportunities for foreign involvement in telecommunications, these are limited in value due to the amount of investment required and the potential for resistance to liberalized tariff structures. Electricity utilities also offer opportunities, although market prospects are hampered by resistance to liberalized tariff structures, subsidization of rates, the common culture of nonpayment, the virtual non-existence of local capital markets, and limited interest in foreign markets for securities of utilities companies from these and many other transition countries. For these countries to attract more than their relatively low per capita levels of FDI observed to date, they will need to rationalize the role of the state sector, provide a favorable business environment, reduce corruption and crime, and increase political stability. In addition, they will need to forge ahead with public sector reform, making their public sectors more professional and service-oriented, less bureaucratic, and less corrupt. Reducing crime and political instability is also important, especially in Albania, where enterprises rated them as the most important constraints to performance and growth. Reducing corruption is important in all the countries, but appears to be most important in Azerbaijan, where corruption is reported to be the most pervasive. The four subject countries will also have to ensure that requisite infrastructure is in place to support private businesses, including power, telecommunications, roads, air travel, and hotels. With small markets, their focus should be on outperforming the norm to compensate for competitive disadvantages. In other words, the model should be Singapore, rather than their next-door neighbors. While FDI will remain low by global standards, the impact on the economic performance of these countries could be significant if they take the steps required to encourage inflows. Financial Intermediation and the Prudential Framework. Given the difficult enviromnent, it has been hard for banks to expand lending to the real sector in all the subject countries. The culture of default has been evident in the formal enterprise sector for years, largely due to earlier incentives that did not require market-based discipline on the part of companies. The legal framework provided banks with little or no protection for secured transactions, while unsecured transactions were more difficult to contemplate or too costly for enterprises to assume. Where the state stayed involved in banking, loan portfolios deteriorated and generated losses for the economy to absorb through fiscal bailouts, higher interest rates, and foregone competition. In Albania and Azerbaijan, the loan portfolios of state banks have added significant costs to intermediation. In response, Albania is moving forward with complete banking sector privatization after having frozen state bank lending three years ago as part of the effort to clean up the large non-performing loan portfolio and prevent new bad loans from being made. Azerbaijan has taken a more gradualist approach, opting to consolidate and restructure prior to definitive privatization. Meanwhile, Moldova has made progress in building up capacity, yet it has no major banks, and many of the existing banks have problems of illiquid and non-eaming assets. The Kyrgyz Republic has 6 also experienced numerous problems in its banking sector, largely reflecting the fragility of its economy at large. Efforts to ensure There have been two salient features that have pointed to the financial discipline insufficiency (notwithstanding their necessity) of reform efforts to date. First, positive efforts to ensure financial discipline and and proper proper management in banks have not been matched in the real management in sector, particularly in weak and fragile economies that have banks have not mobilized limited foreign investment and, other than Albania, have low levels of trade with OECD countries. This is partly been matched in because there is no centralized regulatory authority for the the real sector enterprise sector as there is for banking, with minimum capital requirements, capital adequacy tests, liquidity management requirements, and fundamental reporting requirements. The only time such comparable discipline applies to the real sector is when companies are large enough to list on a securities exchange, seek international depository receipt financing, or obtain credit from the international syndicated loan market. Apart from that, accounting and other governance- related requirements are tax-oriented rather than focused on larger issues of stability. This is a gap that will remain, and points to other needs that include training, systems, and fundamental exposure to advanced market methods. This gap, which is evident in all four countries, also points to a second weakness. Reform efforts throughout the decade have focused on formal institutions that cater to larger enterprises. While banks obviously mobilize household deposits and provide other fundamental services, this role has been limited from a lending standpoint because of the risks mentioned above. Any effort to ignore prudential requirements is perilous, either in the form of regulatory penalties or in the form of poor portfolio performance. Other more advanced forms of financing, such as developed securities markets, have been tried in places like the Kyrgyz Republic and Moldova without any real impact. What is not sufficiently developed are more focused forms of financing, reflecting the scale of enterprises, income, investment, and management capacity that prevail in these economies. Progress has been made in Albania, the Kyrgyz Republic and Moldova with micro-finance and village credit groups. These have received donor support, and are far more suitable for the financing needs of most enterprises. Meanwhile, the more direct contact of these groups with borrowers and the reduced formality of the process lead to better performance. However, these institutions alone are not enough to bring savings into the system and to achieve OECD-level intermediation figures. Banks have had to contend with a complete shift in incentives for themselves and for their prospective customers since the transition began. Rather than banks serving as vehicles of unchecked financial flows for political patronage and production targets, they have been assigned the responsibility of managing risk and becoming a source of financial discipline in the economy. In some cases, such as in Moldova, banks have simply curtailed lending. This has made it more feasible for banks to meet prudential requirements regarding capital 7 adequacy, and to curtail their losses to weak or undisciplined borrowers. In other cases, lending continued through the state banks (as in Albania through 1997, and Azerbaijan), only to run up major losses and experience major portfolio erosion. More recently, efforts have been made to alter this, particularly in Albania where state banks were put under a lending moratorium. In the Kyrgyz Republic, as elsewhere, public and private banks have generated poor results from lending. At the same time, banks have been criticized for not lending, leaving behind many projects that were sensible and depriving the real sector of funds needed for growth. Banks' preference for investing in government securities as a safe way to recapitalize leads to further criticism of their role as a source of quasi-fiscal financing, rather than true intermediaries. Banks' responses are that they are restricted in their activities and risk assumption by regulatory authorities. Such restrictions appear to be justified considering the impact of pyramid schemes in Albania (1997) and Azerbaijan (1994), and the culture of loan default that has permeated all four banking systems (as reflected in high levels of non-performing loans). However, where Allfour central problems appear to have been suppressed, as in Moldova, this banks have has only translated into low levels of internediation from the funding side as well, and a greater reliance on arrears than successfully might otherwise have materialized. reduced inflation ... All four central banks have successfully reduced inflation, they have also although it remains moderately high in the Kyrgyz Republic managed to and Moldova. In most cases, they have also managed to stabilize exchange stabilize exchange rates. Whereas the Kyrgyz Republic's currency depreciated significantly after the Russian Federation rates crisis, the regulatory authorities in Azerbaijan and Moldova succeeded in limiting exposure to the Russian market, shielding their banks from adverse effects. Albania has long shown extraordinary stability with inflation rates and exchange rates, partly due to high remittance flows and the cushion of informal trade. However, from the perspective of financial sector stability, there are questions about whether a radical shift to market conditions and risk-based banking supervision is the most appropriate route to pursue. While this study does not provide conclusive evidence, it is clear that significant time is required to build up adequate regulatory capacity as well as management and board capacity at banks. Furthermore, a risk-based system is not able to function if the legal environment for lending and recovering on exposures is flawed, as it has been in all four countries. Meanwhile, all four countries have had major problems with portfolio quality, that these problems have added to the costs of intermediation, undermined confidence in the banking system, and deterred funds mobilization into the banking system. These negative trends are also positively correlated with state banks playing a major role in the sector. In Albania, which is now finalizing its privatization of banks, and Azerbaijan, which still has two large state banks, non-performing loans in the state banks were high, cost 8 the budget substantial sums, and slowed competition in the marketplace. Albania is correcting the problem. Azerbaijan is taking time to restructure. Banking appears to be one sector (along with a few others where foreign expertise is recognized as vital) where the public's confidence is enhanced by the presence of investment grade banks from abroad, as opposed to traditional domestic banks. In several countries, there has already been substantial foreign investment in the The prudential banking sector. Most notably in Albania, the Savings Bank has framework needs been restructured and will likely be privatized in 2002 after to be disciplined selecting a privatization advisor in 2001, eliminating state ownership from the system. These changes, following strict efforts and soundfor to privatize the last two state banks and to open the system up to long-term safety foreign-controlled operations, is coinciding with an impressive and proper risk increase in deposit mobilization. management What must be accepted is that the prudential framework needs to be disciplined and sound for long-term safety and proper risk management. When such a foundation is absent or abandoned, it takes years to restore performance and public trust. However, it is possible that a more objective strategy, in which the framework evolves in stages from "rules-based" to "risk-based" standards, would be more appropriate for small economies in situations like those facing the four subject countries. At a minimum, this would allow for needed training and adaptation by banks and regulators, and perhaps may not have led to such low levels of lending. In the future, there may be a rationale for a customized financial framework in which credit unions, savings and credit associations, micro-finance and village financing groups and other small-scale operators are viewed as a priority. Here, the establishment of customized needs could be introduced, such as pledge laws, property registries, and specialized courts on a decentralized basis. Credit information bureaus that document loan performance could be developed to ensure that performance (good and bad) is recognized, and more seamlessly introduced into the banking system when financing requests exceed the levels provided by the smaller institutions. There may be some room for the easing of requirements on these smaller institutions, as the demise of individual institutions would not be as devastating to financial sector stability as Donors should not the demise of a major deposit-taking bank. encourage premature What this approach also suggests is that donors should development oJ focus on the smaller scale aspects of financial "universal" intermediation at least as much as they focus on securities institutions until markets, contractual savings, and more complex, secondary forms of financing. These should certainly not adequate institutional be ignored, particularly as pension reform remains a and market capacity is critical macroeconomic and structural need and insurance is necessary to foster investment. Their development also in place 9 helps to reduce the emphasis on banks for all financing needs, and allows a more appropriate segmentation of the market. However, there is also an argument that these should evolve over time in a more "organic" way based on market demand, and that the phasing of reforms should be parallel to the structure of the economy at large. That would translate into a sizable portion of donor efforts being in the small-scale domain, with an eventual view to graduation to more formal and comparatively larger institutions (e.g., banks) that have higher fixed costs and more complex regulatory, managerial and administrative requirements. At a minimum, donors should not encourage premature development of "universal" institutions until adequate institutional and market capacity is in place. Legal Framework. The issue of transforming the legal framework with appropriate incentives is a challenge that is likely to take a generation, at least. In this regard, judicial capacity efforts have not been as effective as needed. Accommodating earlier legal structures with the requirements of a modem economy is an ambitious endeavor that requires a comprehensive framework with appropriate legislation and, above all, trained judges, bailiffs, administrators, and elected officials. Instead, what has occurred is a frenzy of legislation and regulation driven by different pressures, combined with modest levels of training and assistance to support the effort. Importantly, this assistance has not been demand-based, so it has tended to reflect the views and capacity of specific donors - and sometimes even national commercial interests in donor countries - rather than genuine local demand arising from local conditions. As in all societies, it takes decades for bodies of law, decisions, precedents, and general capacity to take shape. Given that Albania was among the most isolated of communist countries for more than four decades and that Azerbaijan, the Kyrgyz Republic and Moldova have all had to revise and unwind Soviet-era legislation, legislation alone has taken time. Unlike in several other Central European and Baltic states, there was virtually no commercial tradition that was actively exercised through courts in any of the four countries. Thus, the introduction of modem norms for bankruptcy, liquidation, and secured transactions is a radical development that will take several decades before the reforms become effective. These weaknesses have shown up in all four countries with respect to financial intermediation, and represent a major factor in banks being unwilling to lend and take risk. Even with secured assets, the time and cost involved in seizing collateral makes such prospects unattractive. The corruption of court personnel due to low pay and other factors adds to the risk. This translates into low levels of lending, and this may well remain the case until a more creditor-friendly environment is in place. This is beginning to happen in Albania, and has actually been partly in place in the Kyrgyz Republic. However, in the latter case, enterprises have been able to avert bankruptcy by running up fiscal and utilities payment arrears. Under such circumstances, banks have had limited recourse through the courts. These are the kinds of practical problems that need to be resolved for banks to ultimately feel confident about lending to the enterprise sector. 10 Under the circumstances, a strategy that focuses on the "batching" of legislation, transactions and specialized in-court and out-of-court capacity might be more effective in solving some of the problem. The objective here would be to streamline the focus so that commercial transactions could be implemented within the context of a narrower yet effective legal and judicial framework. This is a lesson that can partly be drawn from Thefour Poland in 1993-94, where exceptions were permitted to the civil code for 15 months to allow out-of-court restructuring to proceed as countries would a basis for financial restructuring of debtors' and banks' bad loan benefit from portfolios. alternative While not exactly the same situation, the four subject countries forms of dispute would benefit from alternative forms of dispute resolution, such as resolution out-of-court arbitration and specialized courts for common commercial transactions. This approach could be applied with specifically tailored legislation defining clear rights and responsibilities of parties, all the way to providing standardized contracts. Automated and mandatory pledge registries and properly notarized contracts would clarify rights and responsibilities associated with secured assets as adjuncts to legal credit agreements. International standards of asset valuation combined with clear guidelines regarding quality control and substitutability would provide some modicum of protection to those receiving pledges. Standardized and statutorily approved contract language could be adopted to serve as basic norms to be followed according to specific types of transactions, much as lease agreements and Powers of Attorney are standardized in Western markets. The use of such forms and consequent disputes would entitle the parties to access to specialized "courts," with specially trained judges, to benefit from expedited procedures under the direction of trained arbitrators. While this approach would not solve all problems, it would be effective at addressing some of the impediments to lending and basic dispute resolution. In addition, land privatization, to the extent permissible based on local cultural standards and traditions, is an important catalyst for secured transactions. Innovative land-lease arrangements and the development of commercial property markets based on clear claims and sound mortgage legislation could also help overcome some of these obstacles. CHAPTER ONE: BRIEF PROFILE OF THE COUNTRIES SELECTED Albania, Azerbaijan, the Kyrgyz Republic, and Moldova were selected for the study for several reasons. First, they are among the poorest countries of the region, with per capita incomes ranging from US$2,218 to US$2,797. On an international scale, they range from 144 to 175 out of 210 countries in terms of per capita incomes (see Table 1.1).3 They are also below averages achieved in most CEE countries, and well below the five countries that are the leading candidates to join the European Union. At US$2,416 on average, they are about two-thirds the average of CIS countries. This is less than a quarter the per capita income levels of the five EU accession candidates, and less than 10 percent of G-24 OECD averages. Table 1.1 A Brief Profile of the Case Study Countries Albania Azerbaijan Kyrgyz Moldova _____________ ... ___________ R epublic Population (mn) 3 8 5 4 GNP$-billion) 2.7 3.9 1.6 1.8 GNP Global Rank 137 125 162 158 GNPpcapita (p) 8 10 490 350 410 GNP pc GlobalRank _44 161 175 167 Notes: data from 1998 Source: World Development Report, 2000 1 Second, they have shown varying degrees of success relative to previous years, yet they are all low in terms of real GDP. For instance, Albania's 1999 output exceeded its pre-transition levels, one of only four transition countries to do so. (The others are Poland, the Slovak Republic, and Slovenia.) Yet, the other three countries studied are still below their output performance when they embarked on the road to market reform. This is not altogether unusual, as non-Baltic transition economies from northern Europe have generally restored pre-transition output levels, while countries of the former Soviet Union are only at. output levels of about 60-70 percent of the pre-transition period.4 Table 1.2 compares current GDP and output to levels one year before the transition process began, providing a rough indication of the relative progress of the various transition economies a decade into the transition. 3 Statistics sourced from World Development Report (World Bank), 1999/2000. 4 See Fischer and Sahay, "Economies in Transition: Taking Stock", Finance & Development, September 2000. 13 Table 1.2 Output and Income Levels Among Transition Countries GDP-PPP FPer Capita GNP- Ratio of 1999 Ratio of Real Basis in I Purchasing Output to One Level of Current UTS$ Power Parity in Year Before GDP: 1998 ________________ billions US$ Transition vs. 1989 Albania 9.3 2,797 1.03 0.86 Armenia 8.4 [ 2,215 0.56 0.41 Azerbaijan 19.1 2,398 0.53 0.44 Belarus 62.3 I 6,088 0.83 0.78 Bulgania 40.6 4,936 0.74 0.66 Croatia 30.6 6,790 0.80 0.78 Czech Republic 127.0 12,331 0.95 0.95 Estonia 11.1 7,726 0.87 0.76 Georgia 28.6 5,239 0.46 0.33 Hungary 108.9 10,814 0.99 0.95 Kazakhstan 67.5 4,372 0.70 0.61 Kyrgyz Republic 10.6 2,252 0.68 0.60 Latvia 15.2 1 6,235 0.64 0.59 Lithuania ___ ___23___3 __ 6,292 0.70 1 0.65 FYR Macedonia 8.9 4,405 0.59 0.72 Moldova 9.5 2,218 0.38 j 0.32 Poland 308.8 7,980 1.28 1.17 Romania 127.2 5,664 0.80 0.76 The Russian 918.8 6,271 0.59 0.55 Federation _______ Slovak Republic 55.4 10,259 1.01 1.00 Slovenia 31.2 15,749 1.05 1.04 Tajikistan 5.8 i 949 0.46 0.42 Turkmenistan 12.6 2,664 0.66 1 0.44 Ukraine ] 159.9 3,178 0.41 1 0.37 Uzbekistan 47.3 1,965 0.93 1 0.90 CEE average 8,006 0.90 J 0.95 EU-5 avelrage 1 __10,567 1.03 0.97 FSU aver~age ___3,786 0.60-0.71 k 0.53 Notes: GDP and per capita income from 1998-99; EU-5 are unweighted averages for Czech, Estonia, Hungary, Poland, and Slovenia; FSU averages include Baltic countries for 1999 output ratio range; for real estimates of 1998 GDP, base of 1.0 is for 1989; for real estimTates of 1998 GDP, CEE includes Baltics, and FSU is CIS (excludes Baltics) Sources: World Development Report (World Bank), 1999/2000; EBRD Transition Report, 1999; Finance & Development, September 2000 ______ ___ 1 4 Third, these countries are often viewed as either "late reformers" or "late performers" in terms of economic development and general competitiveness. While they all have moved forward with policy reforms, economic development levels are still relatively low when measured by per capita income, direct investment, and other indicators that point to rising incomes and future growth. This suggests that policy reform alone is not sufficient for sustainable economic growth. These countries Policy reform demonstrate that structural reforrn, and possibly other changes, are alone is not necessary conditions for long-term growth. This is a sobering sufficient for conclusion, because structural reform itself is a time-consuming process, involving years of investment and effort. Table 1.3 presents a sustainable composite of reform progress, based on the EBRD's Transition economic Indicator rankings, showing that the four countries have moved growth forward with policy reforms, but are still lagging behind many other transition countries. Table 1.3 EBRD Transition Country Indicators r1EU 5 CIS Albania Azerbaijan j Kyrgyz{ IRepublici Moldova Legal Commercial law 4- 3- 2+ 3- 3 3 Bankruptcy law 4- 2+ 1 3+ 4- 3 Secured transactions law Mix Mix Yes Restricted Yes Yes Private Sector Development Private Sector % GDP 72% 50% 75% 45% 60% 50% Large-scale Privatization 4- 3- 2 2- 3 3 Small-scale Privatization 4+ 3+ 4 3+ 3 3 Governance/Restructuring 3 2- 2 2 2 2 Financial Sector Development Banking reform & interest rate liberalization 4- 2 2+ 2 2+ 2+ Securities markets/NBFIs 3+ 2- 2- 2 2+ 2+ Financial regulations 4 2+ 2- 2 3 2 Deposit insurance Yes Mix* No No No No Private pension funds Yes Mix No No No No Notes: Scores range from 1-4, with 1 being low and 4 being high; scores are latest available from EBRD (World Bank modifications: 0.3 added/subtracted as proxy for +/- system of EBRD); EU5 and CIS are average unweighted scores; in the CIS, only Belarus, Kazakhstan, and Ukraine have deposit insurance Source: EBRD Transition Reports, 1999 and 2000 All four countries have also faced at least one crisis or shock of some sort-war, pyramid scheme collapse, contested borders, banking crisis. This again suggests that policy reform is necessary for sustainable growth, yet other factors, including events beyond local control, enter into the equation and undermine the effectiveness of reforms. The shocks and crises 15 likewise add to the time required for effective implementation of a new incentive structure, diverting attention and resources from needed institutional capacity-building and market development. Allfour countries Finally, the four countries represent a range of characteristics in were largely terms of history, resource endowment, and proximity to isolatedfrom the markets. Only one (Albania) is not a member of the former international Soviet Union, yet it was as isolated as any country in Europe until the early 1990s. Thus, all four countries were largely marketplace until isolated from the international marketplace until recently. None recently has a population of more than 8 million and, combined, they only represent about 20 million people. Thus, as long as their economies remain noncompetitive, depending on agriculture and low value-added services, there is little chance they will experience a marked increase in incomes and purchasing power. In this case, Azerbaijan represents an exception, to the extent it can harness its energy sector for broader growth. Comparative advantage has also been identified for labor and skills in Albania, some resources (e.g., gold, cotton) in the Kyrgyz Republic, and agriculture and labor in Moldova5. However, only one of these countries has mobilized significant foreign investment (Azerbaijan in the oil sector), although two have benefited from significant remittance inflows (Albania and Moldova) due to workers living abroad. The Kyrgyz Republic has also attracted helpful foreign investment in its gold sector, although production at its main mine is expected to terminate in 2008. In general, apart from Azerbaijan, direct investment figures are low. All four of the subject countries have high levels of informal economic activity, and all have sufficient agriculture to provide subsistence as part of the safety net. This is beneficial in terms of flexibility and entrepreneurship, but has made it difficult to establish a viable fiscal base for governnents to provide meaningful services and investment, and to develop civil institutions that inspire confidence in the public. 5 See EBRD Transition Report, 1999. 16 CHAPTER TWO: FINANCIAL SECTOR CHALLENGES There are numerous features common to the four countries' banking systems. Table 2.1 provides comparative statistics for the banking systems and selected economic indicators for the four subject countries. Table 2.1 Overview of Subject Country Economies and Banking Systems Alania jAzerbaijan Kyrgz I___________ Republic Moldova M3 (in US$, 2000) $2,250 million $828 million $150 million $228 million M3/GDP (1995-2000) 47.8%-59.8% 11.1%-18.7% 11.9%-17.1% 16.2%-20.6% Peak Year (1995-2000) 2000 2000 1995 1999 Bank Credit (in US$, 2000) $172 million $521 million $58 million $192 million Bank Credit/GDP (1995- 8.9%-15.9% 11.8%-14.5% 4.5%-13.6% 14.6%-18.1% 2000) Peak Year (1995-2000) 1996 1995 1995 1997 US$ Deposits $1,637 million $567 million $67 million $132 million US$ Deposits per capita $481 $70 $14 $36 US$ Capital (net) $242 million $269 million $8 million $76 million US$ Capital per capita $71 $33 $2 $21 US$ Assets $1,934 million $1,050 million $93 million $262 million US$ Assets per capita $569 $130 $19 $72 Pyramid Schemes 1997 1994 N/A N/A Banking Crises 1997 1994-95 1993-94; 1992-93 1998-99 Peakyear for NPLs 1997 1999 1995 1996 Hyperinflation Years 1992 1991-95 1991-93 1992-94 Peak Inflation Rates 237% 1,788% 1,363% 2,198% X-rate: US$ (1995) 94.5 4,440 11.2 4.5 X-rate: US$ (2000) 145.0 4,525 49.0 12.5 Notes: M3/GDP and Credit/GDP figures are ranges, from low to high; deposit, capital and asset figures are for 2000; population figures for Moldova are net of Transnistria; hyperinflation years are inclusive of all years in which average CPI exceeded 100 percent; peak inflation is highest CPI at year end; banking crisis in Albania applies to recognition of the degree of bad loans, subsequent actions to freeze lending by the major (state) banks, and general instability resulting from pyramid schemes, as banks actually benefited in some ways such as in the form of increased deposits Sources: IMF, EBRD, central banks 17 The three CIS countries have low levels of deposits and capital, and for that reason, low levels of financial Allfour countries have intermediation as measured by bank credit to GDP. low levels of deposits Albania is a bit different on the funding side, having and capital, and low attracted a significantly higher ratio of per capita deposits. However, due to the legal environment and bank credit to GDP other weaknesses, Albania is similar to the three CIS countries in having very low levels of lending. Meanwhile, all have faced banking sector crises at some point in the last seven or eight years, often at least twice. This has been brought on by differing factors, including hyperinflation in the CIS countries, pyramid schemes where cash holdings have been high and banking systems weak (Albania in 1997, and Azerbaijan in 1994), and vulnerability to weaknesses in other markets in the absence of sound bank governance, management and supervision (Kyrgyz Republic exposure to the Russian market). In fairness to Albania, its banks actually benefited as a consequence of the pyramid schemes, increasing their share of deposits, almost entirely from the household sector. Nonetheless, the collapse of the pyramid schemes, high levels of non-performing loans, and general breakdown of institutions triggered strong measures to stabilize the system. This included a lending moratorium on the banks, eventually eased on the private banks but kept in effect on the state banks until privatized. Until recently, all four countries have had high levels of concentration in the hands of state- owned banks, or "pocket" bank behavior in which privatized banks or new private banks lent on behalf of related parties and shareholders, resulting in large losses. The end result has been lagging public confidence in the early years of reform, exacerbated by the need to rebound from a major crisis. Only Moldova has followed an effective long-term strategy, although its outcome has been hampered by problems in the real sector. More recently, Albania appears to have articulated a focused strategy with the intention of restoring household confidence in banks and privatizing the entire system. Albania distinguishes itself among the four subject countries for having reasonably high broad money to GDP figures and far higher per capita ratios. This partly reflects the more moneyed bias of non-CIS countries, as well as Albania's high remittance flows from relatives and friends working abroad. However, even there, credit-to-GDP from the banks has not exceeded 16 percent, and that was in 1996. Comparable ratios are 37-65 percent in the EU accession candidates6 (43 percent on average), and an average 117 percent in OECD (G-24) 6 Data based on 1998-99 figures. 18 countries.7 Thus, per capita balance sheet figures in Albania reflect better performance than in CIS countries, yet weak performance relative to EU accession candidates. Apart from the Kyrgyz Republic, the subject countries are improving broad money fundamentals to GDP over time. For example, Albania and Azerbaijan show broad money- to-GDP to be higher in 2000 than in any year since 1995, and Moldova's peak year was 1999. Only the Kyrgyz Republic has shown a decline since 1995. This bodes well from the funding side, as Albania and Moldova have experienced deposit growth. In the case of Albania, this is partly driven by high levels of remittances from offshore workers. To a lesser extent, this is also true in Moldova, where workers are benefiting from improved conditions in the Russian Federation, its major trade partner. In Azerbaijan, M3 trends are more dependent on the oil sector. About 85 percent of the major increase in foreign currency deposits held by the banks in 2000 was related to the Oil Fund, and placed in one bank. Thus, while Azerbaijan shows improving aggregates and ratios in this regard, these are not well distributed throughout the banking system. The Kyrgyz Republic continues to suffer from low levels of liquidity and capital throughout the economy, and as such has the lowest indicators of the four banking systems and economies reviewed. While three of four countries are showing progress in mobilizing funds, their credit-to-GDP ratios are generally weak. The three CIS countries all faced hyperinflation in the early 1990s, and then tightened monetary policy with the introduction of new currencies. This led to a taming of inflation rates and, for a period of a few years, underlying exchange rate stability. In fact, the exchange rate strengthened against the U.S. dollar in Azerbaijan from 1996-97 and remained stable in 1998. It was fairly stable in Moldova in 1996-97, and was likewise relatively stable in the Kyrgyz Republic in 1996-97.8 However, tight money has also translated into low levels of lending. Albania likewise had to apply tight monetary policy after inflation rates in 1992 reached hyperinflationary levels. In fact, for an interim period, the low levels of credit to GDP may reflect positive developments, rather than the negative view commonly associated with declining credit to GDP. The reduction in credit-to-GDP reflects a number of reforms that should have long-term benefits for economic growth, such as more stable inflation and exchange rates as a result of tightened monetary policy, better credit management based on modernized prudential regulations, and a period of stabilization in which banks are trying to increase capital prior to any major resumption of lending. 7 When OECD figures are cited, these are generally restricted to North America, Western Europe. and East Asia/Pacific net of Korea. The Czech Republic, Hungary, Korea, Mexico and Poland are not included in these cited figures. 8 Year-end exchange rates against the US$ were manat 4,100 at end 1996, and manat 3,890 at end 1997-98 in Azerbaijan. In the Kyrgyz Republic, rates were som 16.7-17.4 in 1996-97, as compared with 11.2 in 1995. In Moldova, rates were lei 4.7 in 1996-97, as compared with 4.5 in 1995. All of these coincided with periods of declining CPI rates in 1995-97. 19 Not lending to the While banks in the four subject countries are routinely real sector has criticized for not lending to the real sector and assuming ofiten been an enterprise risk, this has often been an appropriate response to often been an the weak creditworthiness of the enterprise sector. There are appropriate many reasons behind this, not the least of which are low levels response to weak of reported enterprise capital, the absence of a reliable legal creditworthiness infrastructure supporting collateral for secured transactions (a legal constraint, not an asset issue), limited debt capacity, of enterprises inadequate cash flow, and the poor quality of information. In most countries, the enterprise sector often operates outside formal financial channels to avoid tax liabilities and official rent seeking. Estimates of informal activity are particularly high for Albania and Azerbaijan. For this reason, the enterprise sector shows low levels of deposits in banks. Consequently, they receive limited credit in return. This is particularly the case in CIS countries, and partly the case in Albania.9 When prudential norms were introduced for sound banking, the result was a risk-averse position on the part of the banks, hence limited intermediation. Thus, as part of the equation, the real sector will need to improve its capital structure and demonstrate its repayment/debt service discipline to generate a documented track record of creditworthiness. This applies to both public and private enterprises. NPLs have been problematic with all types of enterprises, often because privatized enterprises are still run according to earlier and unsound practices, and new enterprises are handicapped by their incomplete management, business and accounting skills. This keeps deposit levels low, and makes it difficult for banks to attract debt and equity funding. A weak securities market limits both borrowing and investment opportunities. This also complicates the liquidity situation of banks, with fewer instruments available for them to manage their resources and increase their earnings. In all four countries, there was growth in the mobilization of deposits in 2000. However, liquidity remains tight for households in the CIS countries, and in Azerbaijan, the deposit increases are primarily related to oil sector developments, not broad-based deposit placements from households or SMEs. In Albania, there has been an increase in general household liquidity, but the availability of banking services in rural areas is limited. Meanwhile, NPLs have significantly reduced the overall earning asset portion of banks' balance sheets, triggering higher interest rate charges on loan products to compensate for tighter cash liquidity and toughening provisioning and capital adequacy requirements. Net loan-to-deposit spreads of 8-10 percent on a nominal basis and 15 percent on a real basis are not uncommon in countries where NPLs have been problematic. The Kyrgyz Republic is showing these signs as they have been the most recent of the four countries to face a banking crisis. In the CIS countries, tight liquidity has also led 9 Albania's household sector differentiates itself from that of the CIS in this regard, as households account for a major share of bank deposits in Albania. However, Albania and CIS are alike in that the enterprise sector shows fairly low levels of deposits placed with banks. 20 to a run-up of inter-enterprise arrears, which remain a significant drag on economic development. As a reflection of past instability, all four countries have faced at least one banking crisis or related financial sector Allfour countries shock in the last decade, and all have found it difficult to have faced at least build public confidence in the aftermath. Albania introduced one banking crisis or important reforms in the early 1990s after running up high g inflation rates in 1992. However, a weak legal and shock in the last institutional environment and poor lending practices led to decade, and all have major portfolio erosion. Meanwhile, high levels of cash led found it difficult to to speculation in pyramid schemes, which were partly brought on as a result of a weak banking system. These buildpublic collapsed in 1997, bringing the country to the brink of civil confidence in the war. Since then, Albania has moved forward with aftermath institutional reform and privatization in the banking sector, and this has helped to restore some confidence. Azerbaijan also faced the collapse of pyramid schemes in 1994, which likewise destabilized the environment for banking. It is more than coincidental that Azerbaijan's CPI averaged 1,664 percent in the year in which pyramid schemes collapsed. This ultimately contributed to monetary policy tightening and banking sector reforms in 1994-95. Neither Moldova nor the Kyrgyz Republic have faced the issue of pyramid schemes of any major magnitude, although they remain vulnerable due to high levels of currency held outside the banking system. In all four cases, the subject countries have economies that are vulnerable to some form of cross-border risk that could adversely affect the condition of the banking system. In the case of Azerbaijan, this is clearly related to commodity and pricing risk in the oil sector. In Albania, the risk flows from political issues, such as potential restrictions on Albanians working abroad who send back substantial sums to Albania in the form of remittances. In the Kyrgyz Republic, market exposure to the Russian Federation and other CIS countries contributed to a loss of confidence in the exchange rate, triggering another banking sector crisis in 1998-99. Moldova avoided this fate by shielding its banks from excess exposure to the Russian Federation. However, in the process, Moldova's banking system has failed to play a meaningful intermediary role, while high levels of arrears have weakened the economy overall. In general, the subject countries are similar in their need to shore up institutional capacity in the banking system for meaningful intermediation to be achieved. While laws and regulations are generally adequate and the central banks are applying discipline, this has not been enough to establish high levels of confidence among depositors, international creditors, and investors. Recent improvements in M3 statistics suggest conditions may be improving in all countries but the Kyrgyz Republic. Even the Kyrgyz Republic is moving forward with a program to create more stable conditions and infrastructure for banking. However, banking supervision remains weak, as does bank governance and management. Box 2.1 summarizes current priorities and prospects for the immediate future of the banking systems of the four subject countries. 21 Box 2.1 CURRENT Focus AND PROSPECTS FOR THE FOUR SUBJECT COUNTRIES' ECONOMIES AND BANKING SYSTEMS ALBANIA Current Focus: Albania is currently focused on bank privatization, development of a deposit insurance scheme, modernization of the payments system, introduction of a credit information bureau, alignment of accounting standards with international norms, and liberalization of the insurance market. Prospects: High GDP growth rates of 7-8 percent are projected for the foreseeable fuiture, albeit from a low base. While purchasing power parity figures show higher financial measures-three times higher than traditional measures-the economic base is still small and informal. Basic financial indicators predict relatively limited improvement, particularly in terms of funding from and lending to the enterprise sector. AZERBAIJAN Current Focus: Azerbaijan is gradually moving forward with privatization and consolidation, raising minimum capital requirements, and improving the payments system. A new chart of accounts became effective in 2000, largely consistent with IAS. Prospects: Azerbaijan benefited from high GDP growth in 2000, but it still lags most countries in terms of actual per capita output. GDP is about US$4.5 billion equivalent, or US$560 per capita. While purchasing power parity figures show incomes at three to four times these measures, the economic base is still small and informal. Azerbaijan has a bifurcated economy, one geared to the oil sector, and the other that is more traditional. Basic financial indicators predict relatively limited development, particularly in terms of funding unrelated to the oil sector. KYRGYZ REPUBLIC Current Focus: The Kyrgyz Republic is focused on stabilizing the banking system after the 1998-99 crisis, strengthening supervision and resolution mechanisms, raising minimum capital, and improving the enforcement of prudential regulations to contain risks to financial sector stability. Prospects: The Kyrgyz Republic remains one of the poorest countries in the world, far behind m countries in terms of per capita output. GDP figures at year-end 2000 are estimated to be US$1.3 bill equivalent, or US$263 per capita. The country remains weak at the structural level, which has transla into a weak market for banking. Financial intermediation measures have declined since the mid-19S There is no expectation that this will change dramatically any time in the near future. MOLDOVA Current Focus: Moldova continues to follow its strategy of privatization, enforcement of prudential regulations to strengthen capital and liquidity management, and consolidation of the system for meaningful competition based on principles of sound governance and management. Prospects: Real GDP growth rates of 5 percent are projected for 2001, but Moldova still remains far behind most countries in terms of actual per capita output. GDP figures at year-end 2000 are estimated at about US$1.3 billion equivalent, or US$365 per capita. Trends have been negative since 1998-99, when GDP fell precipitously in real terms.10 Reflecting some of these adverse trends and developments, broad money is only US$222 million, a small base for economic development. The 2001 return of the Communist Party to power is not expected to improve prospects for banking system development or reform. 10 Real GDP growth was negative 6.5 and negative 4.4 percent, respectively, in 1998-99. Converting GDP to US$ at year end exchange rates, Moldova's GDP fell from US$2.2 billion in 1997 to US$1.1-1.2 billion in 1998-99. 22 None of the THE STRUCTURAL ENVIRONMENT FOR BANKING Nre ha the countries has the The four subject countries face a number of similar judicial capacity "architectural" and infrastructure-related challenges in needed for constructing a viable foundation for banking and financial enforcement of services, summarized in Box 2.2. While all four have adopted enforcement of two-tier banking legislation and strengthened their laws and creditor rights, even regulations (usually after a banking sector crisis), all four though legislation is have been slow in developing adequate bankruptcy sound mechanisms, introducing sound accounting and audit principles, modernizing paymnent systems, or developing formal non-bank financial services (e.g., capital markets, pension funds, insurance). As an indication of the insufficiency of laws and regulations, none of the countries has demonstrated the kind of judicial capacity needed for enforcement of a legal framework that supports creditor rights, even though the legislation is basically sound. BOX 2.2 CHALLENGES IN THE FINANCIAL SECTOR ENVIRONMENT Issues Challenges * Collateral enforcement weak L Bankruptcy and liquidation measures ineffective Legal * Judicial capacity insufficient and overburdened Judicial . Absence of formal and effective extra-judicial mechanisms for dispute resolution * Culture of non-payment R Banking supervision traditionally norms-based Regulatory * Risk-based techniques poorly understood Supervisory . Banking supervision marked by inconsistent treatment of banks S Skilled personnel in short supply Accounting * Limited development of accounting/audit profession and standards and * Tradition of tax avoidance and non-disclosure in enterprise sector Information * IAS charts of accounts only recently adopted Payment * Slow and costly processes make liquidity management inefficient and * Non-cash settlement based on culture of barter, arrears and netting has Settlement distorted valuation Legal and Judicial Framework Issues Problems in the legal framework regulating creditors' rights have slowed financial sector development. This has mainly involved collateral issues and bankruptcy procedures, both in terms of the law and implementation. While some countries have adequate bankruptcy legislation, such as in the Kyrgyz Republic (where creditors are the first to be compensated when a firm is liquidated), enterprises have been able to avert bankruptcy by running up fiscal 23 and utilities payment arrears. Under such circumstances, banks have had limited incentives to seek recourse through the courts. At least until recently, the bankruptcy law was considered weak in Albania. Enforcement in all four countries has been weak, except in the Kyrgyz Republic, where modest success has been reported." The banks themselves have often made the problem worse for themselves by rolling over loans they should have provisioned as losses. Thus, the challenge is more than just a legal and enterprise issue, and relates as well to the ability of banks to classify loans properly, provision and reserve for losses, and more broadly manage their risks. This is a challenge faced in all four countries, and has been a common problem throughout the transition economies at some point during the reform period. A strong culture A strong culture of non-payment permeates all four systems, of non-payment derived from their history, current economic conditions, and the permeates all practical inability of banks to foreclose on pledges and other four systems forms of collateral. With the imposition of hard budget constraints and the introduction of stricter prudential regulations (e.g., risk-weighted capital measures), it is simply safer for banks to not lend. This has contributed to low credit-to-GDP measures in all four countries, as well as very high levels of barter and arrears. The result has been to neutralize banks as a source of financial discipline in the intermediation process. In Moldova, banks have generally been segregated from this process since 1995 as a conscious part of monetary policy and banking sector stabilization. However, the high level of NPLs in Albania, Azerbaijan and the Kyrgyz Republic suggests that these countries are only starting to transform the culture of borrowers to one that is based on disciplined principles of commercial creditworthiness. This shows in the low levels of credit-to-GDP Banks have often ratios. been passive in Banks have often been passive in trying to recover unpaid loans. trying to recover This is partly due to banks' traditional internal weaknesses in all four countries, as well as the lack of success banks have had in the courts. unpaid loans The judicial process has traditionally been inefficient and biased against creditors, even when the law has provided them with nominal protections. The lack of capacity, support and information systems has translated into slow adjudication processes and substantial backlogs. In Albania, approximately three-quarters of civil judgments are reportedly not enforced. This is one of many reasons why term lending is uncommon, and why project finance and mortgage lending markets have not developed. Other judicial capacity issues include the need to train judges and supporting administrators, their low pay, and inadequate records. In Moldova, laws are considered adequate, but Moldova's judiciary lacks the experience to implement them effectively. The court system is reported to suffer from understaffing and poor pay, an invitation to backlogs and corruption. Most governments are aware of these weaknesses, and are attempting to improve incentives for banks to lend, although this is not universally true. In Albania, a new Collateral Law was adopted by Parliament in 2000, after which a new collateral registry started operations. The See EBRD Transition Report, 1999. 24 1996 Bankruptcy Law has been amended, and there are plans to introduce a credit information bureau. However, foreclosure on collateral and administration of bankruptcy cases have been slowed due to backlog in the courts. Development of the credit bureau will take time before it can be used effectively as a tool to mitigate credit risks. In Moldova, the government has made some progress in establishing a property registry, but the absence of market mechanisms for asset valuation undermines the use of such information. This has been a factor in banks not feeling sufficiently comfortable with the use of privatized land as adequate collateral for secured loans in the agricultural sector. In general, it has been safer for banks to secure loans with liquid assets to the extent possible. Until these problems are rectified, banks are not likely to lend or assume much risk. In the Kyrgyz Republic, some progress has been made with the use of 99- year leaseholds and the development of agricultural markets. However, liquidation and foreclosure rights and practices are needed for enterprises to be able to use these assets as collateral for loans. Otherwise, banks are unlikely to assume risk in this critical sector of the economy. In Azerbaijan, the government appears reluctant to make the bankruptcy process effective in the real sector due to the likelihood of personnel retrenchment and financial costs. Legislation is varied in terms of the kinds of non-bank activities in which banks are permitted to engage. In Albania, there has been little non-bank development apart from some movement in the insurance sector. For now, it is prudent for Albania to focus on a sound and stable foundation for banking to try to restore confidence after the pyramid scheme collapse in 1997. Recent and anticipated privatization should encourage this. In Azerbaijan, there are restrictions on banks being engaged in non-bank activities, as well as strict controls on the export of foreign exchange. In Moldova, the National Bank has exercised fairly tight and direct control over monetary instruments, while restricting the activities of banks until they can demonstrate their ability to manage risks. Given the anemic economy, there has been little activity from banks. Only the Kyrgyz Republic freely permits "universal" banking, which is premature given the weak state of banking and financial services, not to mention the supervisory processes. While restrictions on banks' risk-taking activities have often been prudent, they have also constrained earnings opportunities and their ability to strengthen capital. In all four countries, banks have had problems assessing credit risk. In many cases, political pressure has been applied to make imprudent loans, as reflected in high levels of NPLs at one point or another. In the absence of adequate credit management capacity and forbearance from pressuring bankers to make "political" loans, it would be unwise to permit deposit-taking institutions to engage in other kinds of trading and investment activities that they are unable to manage properly. However, without these kinds of income-generating opportunities and with the presence of a high proportion of non-earning and illiquid assets, banks are not able to generate earnings to build capital. Tight monetary policy has also imposed high reserve requirements, often wholly or partly unremunerated, further driving up the proportion of non-earning assets. In all four countries, capital will remain a constraint as long as these limitations exist. Foreign banks are generally welcome in all four subject countries, playing a role in providing financial services essential to modernize the banking system. In Albania, most banks are foreign-owned, and all others are private. In Azerbaijan, it is easy for foreign banks to obtain a license, although there is a 30 percent cap on foreign-owned capital as a percent of total banking 25 system capital. Foreign banks now reportedly account for about 20 percent of total share capital in Azerbaijan.12 In the Kyrgyz Republic, legislation is completely open to foreign investment in the banking sector. While recent toughening of licensing conditions may make it difficult for foreign banks unable to meet standards of fitness and propriety, sound financial management, etc. to enter, this is prudent, given the fragility of the Kyrgyz Republic economy and system. In Moldova, while the sector is open to foreign investment, Moldova has failed to attract prime- rated investment, although it does have several smaller banks operating with foreign capital. The four subject countries have approached bank insolvency and liquidity problems in differing ways, and have too often been reluctant to shut down major loss-making banks, particularly those owned by the state. The Kyrgyz Republic has been more forthright than many transition countries in its treatment of Subject countries larger banks with troubled portfolios, even though the judiciary have too often been is considered weak and inexperienced and the National Bank has not always succeeded in winning cases filed against bank reluctant to shut owners and managers for violations. Still, in the Kyrgyz down major loss- Republic, there do not appear to be any legal barriers making banks, preventing the authonities from closing down banks, even when they are large.13 In Albania, it took a crisis of the magnitude of particularly those the pyramid scheme collapse to trigger a change in approach to owned by the state banking sector stability. Since then, the government has moved forthrightly and with determination to restructure and recapitalize (to zero capital, eliminating negative net worth) the two remaining state banks (after having liquidated Rural Commercial Bank in 1997), and to privatize them with strategic investment. One (National Commercial Bank) was privatized this way in 2000, while the last state bank (Savings Bank) is currently in the process of selecting a privatization advisor. Before that, Albania had moved troubled loan portfolios from bank to bank while trying to rationalize branch networks, rather than dealing with the underlying fundamentals of bank insolvency. In Azerbaijan, 170 smaller banks have been closed, but the restructuring and privatization of large troubled state banks has taken more time. This may be due to the cushion Azerbaijan has in direct investment into the oil and gas sector, which has mitigated the sense of urgency regarding banking refonn. In Moldova, some banks have been closed, but most banks have been given time to recapitalize under a deliberate forbearance regime. This has been based on formal agreements and corrective action plans for banks, and has taken into account the underlying difficulties of the business environment. However, the National Bank of Moldova has complained about the amount of time and focus required for bank resolution issues in the absence of an efficiently functioning court system. 12 As of mid-2000, at least three banks were 100 percent foreign-owned, and another two banks were branches of foreign banks (with no domestic share capital). 13 After the 1998-99 banking sector crisis triggered by exposure to the The Russian Federation market, three of the Kyrgyz Republic's four major banks were liquidated, and another was placed under NBK administration. All together, these banks accounted for about half of system assets at the time. The public generally lost confidence in the currency and the banks as a result of the crisis. 26 As the four countries' banking systems have generally moved to the path of reform, they have done so in a manner that is linked to international standards. While many of the "architectural" prerequisites for stable financial systems need to be developed, the authorities in the four countries have cooperated with FATF and other institutions in putting in place mechanisms to deter or contain money laundering. While all four countries have their share of informal trade, the central banks and regulatory authorities have been stricter in recent years in clamping down on the detection of money laundering and the use of banks for these purposes. At the same time, because payment and settlement systems have been relatively underdeveloped and formal offshore financial flows relatively limited, the incidence of money laundering through the banking system is not considered to be all that great. The Prudential Regulatory Framework and Banking Supervision The prudential regulatory framework has been strengthened and modernized in recent years in all four countries, although there are still shortcomings in terms of information quality, banking supervision and risk management. All four countries introduced two-tier banking legislation Theprudential early in the 1990s, with accompanying prudential regulations. After facing some form of banking crisis, all four have since regulatory amended their two-tier legislation, and strengthened the prudential framework has framework. In the case of Albania, this occurred in 1998 after the been modernized... pyramid scheme collapse. Broad prudential requirements now adhere to Basle Committee recommendations and EU norms, and there are still Albania is receiving technical assistance to strengthen its banking shortcomings in supervision capacity. In Azerbaijan, prudential norms are generally informatin .consistent with international norms. However, implementation and information enforcement have not always been consistently applied, and quality, banking minimum capital for existing banks remains low. supervision, and Progress in strengthening banking supervision in the Kyrgyz risk management Republic was considered to be slow through 1999, particularly with regard to enforcement measures against banks for violating accounting standards and prudential requirements. Efforts since the 1998-99 crisis have been made to tighten up on monetary and bank reporting, monitor bank management and governance, develop on-site and off-site supervisory capacity, improve coordination between/among NBK units, and conduct regular inspections of banks. In Moldova, the prudential regulatory framework was modernized on a phased basis, with a detailed plan addressing minimum capital, capital adequacy, provisioning for loan losses, loan classification standards, credit exposure, foreign exchange exposure, licensing standards and practices, and resolution through purchase and assumption or closure on a least-cost basis. Moldova has pursued a narrow strategy focused 27 on straightening out fundamentals prior to moving on to universal banking. Efforts to build the system have been phased, systematic, and carried out without support from the state budget.14 Box 2.3 summarizes the key features of the prudential requirements in each of the four subject countries, and compares them to BIS Standards. Capital requirements still remain low in all four countries, although most have made efforts in recent years to tighten Capital requirements up on capital adequacy requirements. Albania now has remain low, although minimum capital of US$5 million equivalent. While this most have made is low, it is consistent with the EU minimum of Euro 5 million. As for capital adequacy, Albania increased tightened Up on minimum risk-based requirements to 12 percent, and capital adequacy imposed lending limits on banks that are unable to requirements comply as a corrective action. In Azerbaijan, minimum capital is only US$2 million equivalent for existing banks, but US$5 million equivalent for new and foreign banks. The authorities in Azerbaijan have held back on implementing earlier plans to apply the US$5 million standard to all banks. About 60 percent of Azerbaijan's banks were not in compliance with minimum capital requirements as of late 2000, and distorted and inaccurate financial information may also mean that many banks were not in compliance with minimum CARs of 10 percent. In the Kyrgyz Republic, minimum capital was raised to US$1 million equivalent, very low by global standards. Mid-2000 CARs indicate that banks have increased their capital adequacy ratios, although NPLs remain high. In Moldova, minimum capital is about US$2.5 million equivalent. Most banks complied in mid-2000, but barely. Efforts have been made recently to strengthen banking supervision capacity, mainly in off-site surveillance, on-site inspection and policy coordination. In Albania, banking supervision is being modernized, and is now based on more accurate and regular reports from the banks. The Bank of Albania has received technical assistance to strengthen banking supervision and the regulatory framework. Meanwhile, international accounting firms audit all of Albania's banks. In Azerbaijan, the prudential framework is broadly consistent with international norms. However, enforcement has not always been consistent with requirements. This has largely shown up in banks' under-provisioning for classified assets, and overvaluation of capital. In the Kyrgyz Republic, measures taken have included increasing staff levels, developing more of a risk-oriented focus, strengthening off-site surveillance, and establishing regular annual schedules for full scope examinations of banks. Technical assistance is being provided to enhance institutional capacity for more 14 While forbearance has presented a cost, this has been rational and reasonable under existing circumstances. Restructuring has proceeded within the context of limited budgetary resources, a weak economy, and no serious strategic investment in the banking sector. 28 Box 2.3 KEY FEATURES OF PRUDENTIAL REGULATORY REQUIREMENTS FOR BANKS BIS Albania Azerbaijan Kyrgyz Moldova standards Republic Minimum Capital EUR 5 million US$5 million US$2-5 million US$1 million US$2.5 million Minimum CAR 8% 12% 10% 12% 12% Insider lending < 10% total Fixed each year; < 10% total < 10% total < 20% total capital to one aggregate borrowings regulatory capital regulatory capital regulatory capital to insider; <20% of all insiders (direct to one to related parties affiliates; total total capital to or indirect) cannot shareholder; affiliated loans < all insiders exceed 100% of the <20% total related Tier 1 capital bank's regulatory party expense capital. Large borrowers: < 25% total < 20% total regulatory < 25% total < 25% total < 30% total single exposure regulatory capital regulatory capital regulatory capital regulatory capital capital Large borrowers: N/A < 140% total N/A N/A < 40% bank's total 10 largest regulatory capital credit portfolio Large borrowers: < 800% total < 700 % total < 800 % total < 800 % total < 500 % total all large credits regulatory regulatory capital regulatory capital regulatory capital regulatory capital capital Loan N/A Five classes, from Five classes, from Five classes, from Five classes, from classification and standard to loss; ratios standard to loss; standard to loss; standard to loss; provisioning are 1, 5, 20, 50, 100% ratios are 1, 5, 20, ratios are 0-2, 5, ratios are 0, 2, 5, 30, 50, 100% 25, 50, 100% 75-100% Liquidity N/A BoA determines based Asset/liquidity Liquid assets-to- Liquid assets-to- on bank's activities on ratios: current liabilities: > 30% deposits < than 30 "case-by-case" basis >40%; up to 1 days: > 30%+; long- month >80%; term loans (> 1 over 1 year < 80% year) < own funds + liabilities with maturities of > 1 year Government &/or N/A N/A N/A N/A Bank fumding from Central Bank GoM/NBM < 25% Financing of total funding Taxes N/A Profits tax at 30% Profits tax at 27% Profits tax at 30% Profits tax at 35%; provisions for loan losses fully tax- deductible since January 1, 1997 Reporting N/A Monthly financials N/A Monthly Monthly regulatory requirements and all reports on financials and all capital reports; regulations but reports on quarterly reports to provisioning reports regulations NBM which are quarterly Accounting N/A Annual external audits Annual external Annual external Annual external according to ISA/IAS audits according audits according audits according to to ISA/1AS to ISA/IAS ISA/IAS Notes: Minimum capital figures are approximated based on current exchange rates. Sources: Bank for International Settlements; laws/regulations of the four countries; Coats et al (Moldova) 29 effective supervision and enforcement of prudential norms. In Moldova, the National Bank has focused on the introduction of a CAMEL15 system for NBM internal ratings to prioritize supervisory concerns, development of a new chart of accounts consistent with IAS, general development of both on-site and off-site inspection capacity, and development of a bank resolution unit to attend to large bank problems and insolvent institutions. Staffing in banking supervision departments is relatively small, and it is sometimes hard to retain staff. In Albania, there are about 25 supervisors and related personnel (less than two per bank in the country's system) in the banking supervision department of the Bank of Albania. In Azerbaijan, staffing is only at about 45, or approximately 1.5 staff per bank. Moldova doubled its supervision staff to 38 (1.8 per bank), after the new banking law was introduced in 1995. In the Kyrgyz Republic, the departure of many supervisors in the mid-1990s was one of the reasons attributed to weak capacity in 1998-99. Private sector offers will be a continuous challenge to supervision departments, particularly as private firms provide more attractive compensation packages. This represents a potential challenge to supervision departments although, on the positive side, more trained supervisors employed by private financial services firms should also boost compliance with regulations and norms that contribute to financial sector stability. Despite advances in supervision, none of the four countries has None of the sufficient supervisory capacity in place for more complex risks that countries has might eventually be assumed under more developed conditions. In sufficient general, financial intermediation in the economy is low, and all four countries are still addressing basic issues relating to credit risk, interest supervisory rate risk and exchange rate risk. Likewise, the securities and contractual capacity for savings markets are undeveloped, so complex institutions and complex risks instruments have not formed. It will be important for all four countries to use the current period of stabilization and reform to strengthen supervisory capacity under sound prudential regulations. In the coming years, securities market development, pension reform, and insurance sector development are expected to provide greater breadth of financial services. This will require commensurate development and organization of supervisory capacity to ensure underlying financial sector stability. For now, this capacity does not exist. Development of supervisory capacity may be influenced by the different paths the four countries have followed with regard to current and capital account liberalization. Albania and the Kyrgyz Republic have open current and capital accounts and flexible foreign exchange regimes. While neither has attracted any meaningful direct investment or portfolio inflows, Albania has benefited from high levels of remittance inflows, contributing to its comparatively high broad money indicators. The Bank of Albania will have to maintain relatively strict oversight for underlying stability, while banks will need to properly manage their interest rate, exchange rate and maturity exposures if and when they assume a greater role in financial 15 CAMEL is a tool utilized by U.S.-based regulators to assess Capital adequacy, Asset quality, Management capacity, Earnings and Liquidity. 30 activity. Meanwhile, Azerbaijan and Moldova have exercised greater direct control of monetary and foreign exchange tools. Three of the four subject countries are holding off on formal deposit insurance schemes until the environment is more stable, given weak banks, limited institutional capacity, and recent financial sector dislocations. Albania is an exception in this regard, and privatization of the last state bank is expected to coincide with the introduction of a new deposit insurance scheme.16 For the three CIS countries, it is considered premature to introduce a comprehensive deposit insurance scheme, although the Kyrgyz Republic wants to utilize this as a tool to restore confidence in the banking system. Meanwhile, the government in Moldova is on record as rejecting the idea of a deposit insurance scheme until the banking system has consolidated. In fact, the deposit guarantee provided for deposits held at Savings Bank was rescinded in 1995 with introduction of the new Law on the National Bank. Accounting and Information Generating reliable, accurate and timely information has been a longstanding weakness in all four countries, although improvements are being made. Information from banks is improving with the introduction of the new chart of accounts, more regular reporting requirements for prudential compliance, and increasing preparation for the use of risk-based indicators for decision-making. Most authorities are also committed to improving the quality of statistics and information, as well as the disclosure of that information. This has been more evident in Albania, the Kyrgyz Republic and Moldova than in Azerbaijan. All four countries now have charts of accounts that are broadly consistent with IAS. The Kyrgyz Republic and Moldova adopted a new chart of accounts in 1997. In Moldova, annual external audits conducted by international firms according to IAS/ISA became mandatory for banks in 1998. In Albania, the chart of accounts became mandatory in 1999. All banks have been audited by international firms for years, including the state banks prior to privatization. In Azerbaijan, IAS accounts for the major banks (state and private) became mandatory in 2000. These banks are audited by Big 5 accounting firms. However, their annual statements are not currently available to the public. Business Most of the deficiencies in financial information are attributed to accountin is the real sector, and this is unlikely to change any time soon. g Business accounting practices are particularly weak, reducing weak, reducing bank capacity to assess creditworthiness and, consequently, their bank willingness willingness to lend. In Albania, micro-finance operations have to lend 16 BoA has prepared a draft deposit insurance law, with actual implementation expected by the time Savings Bank is privatized. The timing is considered propitious now that it appears confidence in the banking sector has been at least partially restored. This is evidenced in the steady increase in deposit placements with Albanian banks since 1999 (although growth has slowed in recent months), including the growth of foreign currency deposits (from foreign currency cash holdings) and the stable level of lek time and saving deposits (with maturities of three months to one year). 31 applied different approaches to obtain useful information. However, in the case of the comparatively larger enterprises that would be seeking loans from the banks, better information will eventually be needed to increase the willingness of banks to assume greater risk. The lack of professional capacity in the accounting and audit fields means full movement to IAS and ISA is still several years away. Strengthening internal audit functions at banks and enterprises, ensuring the independence of auditors, enforcing a code of conduct that is consistent with international standards, and observing more open standards of transparency and disclosure would contribute to more and better information for market purposes. Payment and Settlement Systems All four countries either lack a modern payment and settlement system, or have only recently made RTGS operational. In Albania, private banks have modern electronic systems for payment and settlement. However, the larger banks have traditionally conducted their transactions manually, causing delays of three to four days in Tirana and up to three weeks for transactions outside of Tirana. Privatization of NCB and Savings Bank are expected to reduce these delays. In addition to delays, the manual nature of settlement undermines confidentiality. In Azerbaijan, the payment system was outdated until 2000. In general, slow movement in this area in the four countries has made liquidity management less efficient for banks, and reduced the quality of services to clients. For capital markets, it has likewise translated into limited growth. To make the payment and settlement system more efficient, countries have been moving to real time gross settlement. Efforts are under way in Albania to achieve real time gross settlement (RTGS) in one to two years, and to link eventual securities market Countries are activity to this system. Azerbaijan has likewise had a modern generally moving electronic system since early 2000, and is currently in the process of to real time gross introducing a system for smaller payments. In the Kyrgyz Republic, the inter-bank payment system was largely automated by 1997. settlement However, given the predominance of non-cash transactions, much of the settlement system is based on debits-credits to offset claims. This is also true in Moldova, which will need to have the framework in place for a modern payment and settlement system if it is to have a modern banking system. This will also be needed as Moldova seeks to develop its securities markets over the next few years. Ownership Patterns The banking sectors in the four subject countries are now majority private. However, until recently, they have long been characterized by substantial state ownership, small and struggling private banks, and low levels of foreign participation (see Figure 2.1 below). The Albanian banking system is composed of 13 banks, of which one major bank is state-owned. Another major bank was state-owned until very recently, while a third state bank was liquidated in 1998. Albania expects that its banking system will be fully privatized by 2002 (the last state bank, Savings Bank, is selecting a privatization advisor in late 2001), 12 years after the transition 32 process began. In Azerbaij an, the two largest banks are state-owned, with privatization a remote possibility for one of them any time soon (United Universal) and unlikely for the other any time soon (IBA). In the Kyrgyz Republic and Moldova, the issue has been less state desire to control banks, and more the difficulty in attracting well-capitalized and well-managed private banks to enter the Figure 2.1 Structure of the Banking System by Type of Ownership Number of Banks Total Assets 70 10- 0% 60 r - - ; 90% - - , , -- 50 . .--.- 1 60% - -- - - -- 40 -5--- -0 % -- -- - - - 30 '-- ~~~~~~~~~~~~~~~~~~0 --- -- --- -- ----- - 20 - 60 - ...-... 30^ 3°%i > i~~~~~~~3 20 - 0 10 10% Albania Azerbaijan Kyrgyz Moldova Albania Azerbaijan Kyrgyz Moldova o1 Foreign Private Banks' Assets (% ) o No. of Foreign Private Banks (1999-2000) * No. of Dornestic Private Banks (2000) * Domestic Private Banks' Assets (%) (both domestic and foreign for Kyrgyz and Moldova) l No. of State Banks (2000) C State BankS' Assets (%) Sources: Central bank web sites; IMF; EU-TACIS; EBRD; World Bank estimates Notes: There is overlap between private and foreign banks; in Albania, the Savings Bank is the last bank to be privatized, but still commands major balance sheet figures; in Moldova, state bank assets are sometimes cited to be zero, but this excludes state ownership in the Economic Bank (savings bank). market. In the Kyrgyz Republic, the largest bank is currently owned by the National Bank, and will likely be privatized after the banking system goes through its current stabilization program following the banking crisis of 1998-99. In Moldova, there is one remaining bank to be privatized, the former Savings Bank. All other banks in Moldova became private in 1992. Even state ownership of the Savings Bank has not been viewed as a significant constraint on competition, as this bank lost its state guarantee on deposits in 1995 and has seen its share of the deposit market decline. State ownership has not been a constraint to opening up the banking system to private control. In Albania, the market opened up to private banking in 1994, and since then, 11 private banks have obtained licenses. In the CIS countries, the method of privatization for banks was based more on the CIS "pocket bank" mode of ownership transformation. This, in turn, left many banks running on low levels of capital, and stuck with poorly performning loan portfolios as a result of mismanaged and connected lending patterns. Private banks were permitted to operate as early as 1992 in the Kyrgyz Republic and Moldova. In Azerbaijan, there were as many as 240 banks by 1994, almost all of them "private." 33 Apart from Albania, foreign banks have a limited share of the market. In Albania, 12 of 13 banks have foreign participation, and it is anticipated that the Savings Bank (the last bank to be privatized) will also attract foreign investment. Thus, Albania will be wholly private and highly foreign in its banking composition later in 2002. In Azerbaijan, about 20-30 percent of banks are foreign in terms of equity investment. The Kyrgyz Republic is open to foreign investment in the banking sector and has several foreign participants, but has not succeeded in attracting major investment. In Moldova, there are several foreign banks that have small shares of the market, although none is a major bank by global, or even regional, standards. There is very little prime-rated investment in any of the banking sectors of the four subject countries. Only in Albania and Azerbaijan have internationally recognized banks invested. Even here, the investment has been limited. This is particularly striking in Azerbaijan, where only one major foreign bank (HSBC) has established operations despite the significant investment from abroad in the petroleum sector. No major banks have invested in the Kyrgyz Republic or Moldova. There is little Major foreign direct investment in the real sector is not enough prime-rated to develop a banking system. Azerbaijan is proof that banking investment in the and financial services can remain underdeveloped, even with banking sectors oJ major foreign investment in the economy. This is largely the four countries because the benefits of investment in the petroleum sector have not been broadly distributed through the economy, and the non-petroleum sector of the economy rernains troubled. While foreign direct investment may not be sufficient for banking sector development, its absence will only make it more difficult for the other countries to achieve banking sector viability. Market Concentration There is a high level of balance sheet concentration in the banking sector in three of the four subject countries. In Albania, the two largest banks accounted for 77 percent of assets, 82 percent of deposits, and 62 percent of credit as late as April of 2000. In Azerbaijan, there is the likelihood of continued concentration, most notably with the International Bank of Azerbaijan (still majority state-owned) and the recently established United Universal Bank (also state- owned, and the end result of the rehabilitation program of three other state-owned banks). These two banks accounted for about 70 percent of assets in 1999, and about half in 2000. In the Kyrgyz Republic, the four most active banks accounted for about 56 percent of credit, 48 percent of assets, and 51 percent of deposits in mid-2000. However, no one bank is predominant in terms of lending, deposit mobilization, or other banking fundamentals. In Moldova, there is still a high degree of concentration, with the five major banks (all formerly state-owned except one, which remains partly state-owned) accounting for about 80 percent of deposits, 52 percent of net capital, and 66 percent of total assets. There is no indication that these ratios will change any time in the near future. It should be noted that market shares are frequently overstated for the "large" banks, as these values are typically inflated by local accounting practices. In many cases, the volume of assets, loans and capital need to be adjusted for under-provisioning of loan losses. This is a special 34 problem for state-owned and former state-owned banks, which are affected much more than the new private banks by both the legacy of socialist accounting and banking practices, as well as pressures to make politically-based loans to loss-making enterprises. Once adjusted, the shares of larger troubled banks are much smaller, as is the general size of the banking system's balance sheet. This would also clearly have an impact on reported income flows of the troubled banks, although they also have competing incentives with regard to profits and losses-higher profits mean higher taxes, yet higher profits make it easier to comply with regulatory capital requirements if earnings are retained. Where concentration is not distortionary, this appears to be due to past closures of problem banks that were considered large by earlier standards. In the Kyrgyz Republic, there is concentration among the four largest banks. However, partly because five major banks have been shut down in recent years, no one institution has enough market share to dominate. The largest bank in the Kyrgyz Republic has only about 15 percent of total assets and deposits in the system. Because the absolute values of the banks are so low, no one bank appears to be strong enough to distort competition beyond case-by-case transactions. This is not the case in Azerbaijan, were the two state banks are comparatively large. However, in Albania and Moldova, there is decreasing concern about concentration and market distortion as banking systems become fully privatized, and new prudential frameworks are implemented. In the future, competition is expected to smooth out some of these shares over time, as has occurred recently with lending flows in Albania and increases in private market shares in Azerbaijan. For instance, in Azerbaijan, private banks accounted for about one-third of total assets in 2000 (net of any write-offs of NPLs), up from about 20 percent in 1997. This is partly the result of the transfer of the state banks' bad assets to a loan recovery agency. However, even as private banks' shares increase proportionally, absolute figures Even as private in terms of intermediation will remain small by global standards banks 'shares in all four countries. Likewise, the range of products and services is expected to be limited until the business environment increase, improves. intermediation will remain small by global standards 35 THE FINANCiAL CONDITION OF THE BANKING SECTOR Box 2.4 CHALLENGES IN THE BANKING SECTOR Areas Challenges Resource * Deposit levels are low, particularly in the CIS countries Mobilization * Access to capital markets is virtually nonexistent * Capital is low * Lending levels are low Intermediation * Incentives to lend are weak * Enterprise conditions discourage lending * NPLs are high, or have been carved out (Albania) but still reflect bad loans Portfolio from the past Quality * Banks are not aggressive in provisioning for loan losses in the CIS countries * Creditworthy enterprise borrowers are in short supply, which limits the market and weakens portfolio quality * Most banks and banking sectors are insolvent, or have been insolvent at some Solvency point in the last few years as a result of bad loans made by state banks and private "pocket" banks * Non-lending products/services are limited, which removes a potential source of earnings to rebuild capital * Costs of recapitalization have exceeded available fiscal resources The Business of Banking In all four of the subject countries, the business of banking is severely constrained by circumstances and conditions in the macroeconomic environment and/or in the enterprise sector. CIS banks have been constrained by a number of factors over which they have relatively li.tle control, including low levels of monetization resulting in high levels of non-cash transactions in their economies, lack of access to potential customers in the informal economy, widespread non-performing loans, low lending demand resulting from enterprise sector distress, and a lack of demand for fee-generating corporate services which enhance earnings without any call on bank capital. Other constraining factors, such as excessive fixed cost structures, the failure to resist political pressures to make imprudent loans, and a reluctance to move decisively to address problems in their loan portfolios are more within banks' control. In Albania, these problems have been a bit less severe, as the economy has shown far higher levels of monetization, banks have All of the countries generated considerably higher deposits per capita, and the ., . country has experienced a surprising degree of monetary and exhibit low levels oJ fiscal stability. Disciplined efforts to improve general loan monetization which portfolio quality have generated better asset quality figures, limits capacity to although lending is low. Nevertheless, Albania still has much farther to go to close the gap with more advanced banking undertakefinancial systems, and the depth and scope of problems faced by banks intermediation in all the subject countries means they face a very deep 36 Figure 2.2 Money is Scarce in the Three of the challenge for future development. Four Subject Countries All of the subject countries exhibit very low levels of monetization, which has, in turn, limited the capacity of their banks to undertake financial intermediation. Using broad money (M3) to GDP as an .... .... .... .. ................indicator, among the four subject ,V ~ ~ ~ ~ ~~. .--- ountries, Albania is the only one where mpo .~4broad money measures exceed 20 percent of GDP. In Albania, these measures | | . _ approached a comparatively high 60 Alaria Azb M*bm percent of GDP in 2000, as shown in a NUMP a R:Ra*KABd MYGP Figure 2.2. Monetization figures would be Sources: IMF; World Development Indicators (World lower in the four countries if GDP Bank); World Bank Calculations statistics were adjusted for purchasing Notes: M3 figures all 2000; PPP-adjusted GDP derived power. Contributing to Albania's M3 from World Development Indicators figures over the last several years are an open current and capital account, high levels of remittances flowing in from abroad, reasonably stable exchange rates, and relatively low levels of inflation. None of the three CIS countries enjoys this combination. The CIS countries show particularly low levels of monetization, by all measures, largely as a result of reliance on barter, arrears and netting dating back to central planning. All three CIS countries rely heavily on non-cash settlements. In part, this is rooted in traditional practices dating back to the Gosbank, and in part due to liquidity shortages emerging during and after the hyperinflationary period. In the Kyrgyz Republic and Moldova, reliance on these measures has been more necessary because of the curtailment of directed credit that imposed hard budget constraints on state enterprises, forcing them to seek non-bank financing, usually in the form of arrears on utilities, tax, social fund, and enter-enterprise accounts. Even in Azerbaijan, broad money measures were only 12-13 percent of GDP from 1995-99, rising to 19 percent in 2000 due to inflows related to Oil Fund savings. Given its fairly high direct investment totals of US$3.1 billion during the 1995-99 period, this indicates that Azerbaijan is channeling its investment resources through other institutions rather than the banks, namely in the form of the Oil Fund and state enterprise arrears. All four subject countries exhibit a high level of commercial activity that is in the informal sector. This activity is difficult for banks to finance, based on modern standards of credit risk management, as well as the tax-driven reasons enterprises prefer to stay in the informal economy (which is culturally based on non-disclosure). Informal activity is generally in the services sector, which was underdeveloped during the socialist period. In Albania, about half of total economic activity is considered informal, and this accounts for much of the country's remittance-driven investment. In Azerbaijan, there are estimates that the official statistics only capture a quarter to a third of economic activity, largely because of the high level of cross- border trade with ethnic Azen's in Iran and other countries. In the Kyrgyz Republic, there is reported to be substantial economic activity that is informal, also composed largely of cross- border trade. While Moldova's figLures for informal sector activity are relatively low by 37 transition standards, at 13 percent of total GDP, comparable figures are estimated to be far higher in the other three subject countries. The predominance of such informality contributes to low levels of credit from the formal banking sector, as enterprises outside the formal sector seek to avoid drawing attention to themselves from tax and other regulatory authorities. This high level of informality also translates into a high proportion of cash sales, or sales on credit through informal institutions based on personal connections and a desire to avoid taxes. By obscuring fundamental cash flow information, all of this also works against traditional fundamentals of credit risk management. Impaired assets have also been a major problem in all four countries at one point or another. The drain on earnings represented by NPLs adds to pressure on banks to maintain high intermediation spreads, which depresses demand for loans and adds to Impaired assets the risk of default as a result of high debt servicing burdens. As pressure banks intermediation spreads increase, the possibility of adverse selection to maintain high increases that, in turn, increases portfolio problems further. Such a intermediation cycle also makes it hard to develop non-lending services, which then intermedatin leads back to problems in mobilizing funds. In Albania, NPLs were as spreads high as 61 percent in 1997. These have since been transferred to a loan collection agency and do not represent an ongoing portfolio problem for the banks, although the riskiness of lending does remain a challenge to banks in general. In Azerbaijan at end 1999, fully 62 percent of total loans were classified as non- performing. Netting out banks' impaired assets, total assets in 1999 would have been little more than half of total assets reported. In the Kyrgyz Republic, problem assets peaked at 36 percent in 1999. While lower than many other peak NPL figures recorded by CIS and other transition countries, these bad assets accounted for a major share of banking system capital, triggering a new round of banking sector reforms. In Moldova, less than 20 percent of loans are non- performing, but provisions for loan losses still consume about two-thirds of net interest income. Few banks in the subject countries provide any meaningful corporate banking services. While private banks have provided corporate services in Albania, largely payment, transfer, custodial and trade finance services, these banks are largely focused on a niche client base in the capital or nearby port. Most enterprises and households have not benefited from such banking services, although privatization Few banks provide of the last two state banks is expected to improve product any meaningful and service offerings. In Azerbaijan, the International Bank corporate banking of Azerbaijan is the only domestic bank that provides services meaningful corporate banking services, while HSBC has focused on its foreign corporate clientele. Apart from these banks, service levels and capacity are considered poor. In the Kyrgyz Republic and Moldova, only basic services are available. Part of this is due to the absence of major foreign investment, as major foreign banks often follow multinationals when they make strategic investments in countries. However, in the Kyrgyz Republic, apart from Canadian investment in the gold mining sector, which is projected to decline in the next few years, there has been no major investment. Likewise, in Moldova, there has been scant direct investment from any source until very recently. 38 The distressed condition of the enterprise sector presents limited incentives for most banks to lend in the four subject countries. The preferred customers of banks - large enterprises which borrow in substantial volumes - are rarely creditworthy because neither the state-owned enterprises nor the recently privatized enterprises have yet undergone the deep restructuring which would permit them to overcome the politically-inspired inefficiencies which were originally built into them by central planners. Most de novo businesses in all four countries are at the micro level, or very small-scale at best. Small borrowers are costly for banks to service on a per client, transaction or loan basis, and they present limited income-generating opportunities for banks at best. Most of these small enterprises (and virtually all households) have few assets. Moreover, their liquidity is constrained by systemic demonetization and the prevalence of barter and arrears in the system. Thus, there is little likelihood that these potential borrowers would place sizeable deposits and strengthen the banks' funding bases. If domestic in orientation (as is almost always the case), there is little opportunity for trade finance income. In summary, banks in the four subject countries are trapped by a series of circumstances, much of which is not of their own making. The macroeconomic framework has presented difficult environmental circumstances in which to operate. Structural weaknesses in the real sector have seriously impaired the banks' ability to accumulate a greater proportion of earning assets with which to rebuild their capital levels. The resulting negative impact on the banks has constrained their ability to mobilize external funding of all kinds, which effectively precludes them from intermediating at levels that would support real sector growth. It is little wonder, in such circumstances, that bank owners and managers feel little urgency to invest in refining banking skills that have little use in the environments in which they operate. Mobilization of Funding Banks in the three CIS countries are faced with severe limitations on their sources of funding, although this is not a problem for Albania. Deposits in the CIS countries are constrained by a significant lack of confidence in the banks and local currency values, and the limited cash liquidity of enterprises. However, there are some positive trends emerging, mainly linked to high, positive real interest rates paid on deposits that may bring household and enterprise funds increasingly into the banking Banks face severe system over time. However, access to capital and syndicated limitations on borrowings, particularly foreign capital and borrowings, remains their sources oj problematic in light of the banks' poor loan portfolios and earnings. In the CIS countries, greater reliance is placed on funding borrowings from central banks, which increases government dominance of the banks' operations and is generally an unsustainable source of borrowings. 39 Deposits Deposit levels are low in the four countries studied, as Figure 2.3 Per Capita Bank Deposits are Extremely Low shown in Figure 2.3, although by Global and Regional Standards Albania's per capita deposit levels are far higher than in Bank Deposits Per Capita, US$ the three CIS countries. Among the four countries 500 studied, Albania's per capita 400 deposit base was equivalent to 350 - US$481, whereas in the other 300.- ... .................... three countries, the range was 250 --..-... : US$14-US$70. In Azerbaijan, 200 where the per capita deposit 150 figure is US$70, this is 150 primarily due to a major o A I , , I influx of foreign currency cIs Albania Azerbaijan Kyrgyz Moldova deposits placed by the Oil u Domestic Curency * Foreign Currency Fund with the International Sources: IMF; central banks; World Development Indicators (World Bank of Azerbaijan. Overall, Bank); World Bank calculations. Albania's per capita deposit Notes: Albania: 2000 are projections; Azerbaijan: 2000 from ratio is about 12 times the September; Moldova: 2000 from June; figures for Tajikistan, average of the other three Turkmenistan and Uzbekistan not available (CIS measures); foreign countries. By contrast, the five currency deposits only for The Russian Federation and Ukraine (CIS measures). leading candidates among __ transition countries to join the European Union have about US$1,889 in per capita deposits, nearly four times the level in Albania, more than 40 times the averages in the three CIS countries studied, and 13 times the average for CIS countries in general. There appears to be a positive correlation between confidence in There is positive the local currency and comparatively high per capita deposit correlation levels. Albanians have about three quarters of their deposits in local currency, while the Kyrgyz Republic and Moldova have between confidence about half their deposits in foreign currency and Azerbaijan has in the local fully 86 percent in foreign currency (see Figure 2.4). In currency and high Azerbaijan's case, this is due to the proportion of oil sector transactions that are denominated in US dollars on a global basis. per capita deposit However, in general, CIS countries show less confidence in their levels local currency instruments than non-CIS countries. 40 The CIS countries are Figure 2.4 The Currency Composition of Bank Deposits finding it particularly Reflects Depositor Confidence. difficult to restore confidence and build a more vable eposi baseBank Deposits by Currency more viable deposit base for banking. The absence 100%-/o _ of both confidence and 900/0- - - _ - deposit insurance has kept 80%- / - -------- - -------- ----- ---- people from entrusting 70/o - their funds to banks, as 60% - . . . ... have the banks' weak 500/o - - ...; - service delivery and the 400/o '- _ - absence of an active retail 300/o network. Levels of tax 20r/6 avoidance by citizens and 100/o corporate entities are high, 0- and concerns about as Abania Azerbaijan Kygyz Mcbsda account confidentiality * Domestc Cenwcy (%/6) * Foreign urency (%/6) also serve as disincentives also une s dsincankes. Sources: IMF; central banks; World Development Indicators (World to keep funds in banks. Bank); World Bank calculations. The limited access that Notes: Albania: 2000 are projections; Azerbaijan: 2000 from September; most households and small Moldova: 2000 from June; figures for Tajikistan, Turkmenistan and businesses have to loans Uzbekistan not available (CIS measures); foreign currency deposits only also provides them with for The Russian Federation and Ukraine (CIS measures). little incentive to place funds with banks. In Azerbaijan, the failure of 170 banks and earlier pyramid schemes (in 1994) have compounded the problem. In the Kyrgyz Republic, there are signs that some confidence is returning after the Russian crisis. However, overall deposit levels remain low, and most broad money remains outside the banking system. In Moldova, as was the case earlier in the Kyrgyz Republic, deposit levels were increasing in 1996-97 after the introduction of banking reforms. However, enterprise problems associated with exposure to the Russian market led to a withdrawal of funds in 1998. Household liquidity requirements followed this pattern. Since then, Moldova has begun to show some initial signs of rebounding confidence, although its economy remains very weak. Under such circumstances, it has been virtually impossible to make progress with the deposit base of the banking system. Poverty has also made it difficult to develop a viable and broad funding base for the banks. Some of the problem associated with developing a more viable funding base has to do with the predominance of people living in rural areas, where poverty is fairly widespread and trust in banks and other formal financial institutions is limited. This is true in Albania, the Kyrgyz Republic and Moldova. For instance, in Moldova, about half of GDP is accounted for by agricultural production and processing. However, much of production is subsistence, large agricultural operations have been dismantled, and the prevalence of barter, arrears and netting have left much of the rural sector illiquid. This has reduced deposit mobilization opportunities, and added substantially to the risk of loan exposure. While the population of Azerbaijan is more urbanized, its wealth has not been well distributed and banks have not branched out much beyond Baku. As elsewhere in underdeveloped transition economies, the enterprise sector still 41 keeps its funds outside the banking system. In Azerbaijan, this also appears to apply to most of the household sector, slowing deposit mobilization. Albania and Azerbaijan have been hurt by earlier pyramid scheme debacles in 1997 and 1994, respectively. In particular, Albania's economy was nearly devastated by losses associated with pyramid scheme collapses in 1997, with losses estimated to be US$1.2 billion. At least one source has noted that the nominal value of the pyramid schemes' obligations-US$1.7 billion at their peak- approached half of Albania's GDP 7. Losses of US$1.2 billion approximated 45 percent of average GDP from 1997-98 (based on nominal GDP figures from the IMF). Meanwhile, Azerbaijan also experienced losses from pyramid schemes in 1994. This coincided with some of the highest levels of hyperinflation recorded anywhere, and triggered the introduction of a new monetary policy, new currency, and new banking laws. When measured in US dollar terms, deposit mobilization trends since 1995 have been encouraging for Albania, and highly respectable by comparison with EU accession candidate norms. However, trends have been discouraging for the CIS countries. Total deposit values in Albania increased nearly US$1 billion from end 1995 to end 2000. This is nearly a 140 percent increase in just five years, and would likely have been far higher had the pyramid schemes not existed. On a per capita basis, this means total per capita deposits have increased US$277 in the last five years, almost exclusively in the household sector. In Azerbaijan, there has been a net increase in deposits of US$407 million since 1995, essentially based on the 2000 deposit from the Oil Fund with IBA. However, as in the other CIS countries, there was a net decline in domestic currency deposits. Overall, on the strength of the large foreign currency deposit in 2000, per capita deposits have increased 2.5 times since 1995. However, because the deposit base is not well distributed, this cannot be viewed as a sustainable trend. In both the Kyrgyz Republic and Moldova, per capita deposits were flat or slightly negative in 2000 when compared with 1995 levels. This reflects both enduring poverty and stagnant resource mobilization. In the Kyrgyz Republic, per capita deposits averaged only US$14 at end 2000, putting the country among the absolute lowest in terns of per capita deposits. This reflects the low level of monetization and points to the general underdevelopment of the economy. In this sense, physical assets and subsistence agriculture are much more likely to account for peoples' savings than money or other financial instruments. This is also true in Moldova and rural Albania and Azerbaijan. By contrast, the five EU accession candidates showed an aggregate increase in deposits of US$22.6 billion over the five years, an increase of US$353 per capita, and equivalent to nearly half the collective annual GDP of the four subject countries. Table 2.2 highlights trends in domestic and foreign currency bank deposits from 1995 to 2000 for the four subject countries, the five EU accession candidates, and the CIS in general. 17 See Jarvis, "The Rise and Fall of the Pyramid Schemes in Albania," IMF, July 1999. 42 Table 2.2 Net Bank Deposit Growth - 2000 vs. 1995 EUS 5 cis Albania Azerbaia yrgyz eublic Moldova (millions, US$) Domestic Currency N/A -7,183 727 -6 -25 -37 Foreign Currency N/A 1,869 217 413 19 32 Total 22,566 -5,314 943 407 -6 -4 (per capita, US$) Domestic Currency N/A -25 214 -1 -5 -10 Foreign Currency N/A 44 64 51 4 9 Total 353 -19 277 50 -1 -I Notes: Albania: 2000 are projections; Azerbaijan: 2000 from September; Moldova: 2000 from June; figures for Tajikistan, Turkmenistan and Uzbekistan not available (CIS measures); foreign currency deposits only for the Russian Federation and Ukraine (CIS measures) Sources: IMF; central banks; World Bank calculations Deposit growth trends have begun to differentiate in each of the four subject countries, as shown in Figure 2.5. Albania is the most remarkable in its accomplishments, given the magnitude of impact from the collapse of the pyramid schemes in 1997. While total deposits declined in 1997, this was mainly represented by the deposits of private sector enterprises and, to a lesser extent, state enterprises and farms that withdrew funds to speculate in the pyramid schemes. Households also participated in the pyramid schemes, although not, apparently, with funds withdrawn from banks.18 In fact, households showed an overall increase in deposit placements with banks, responding in particular to the comparatively high interest rates paid on three-month time deposits.19 In general, aside from 1997, Albania has shown deposit growth each year from 1995-2000 in both local and foreign currency. In Azerbaijan, performance has been inconsistent, with growth in 1997, declines in 1998, sluggishness in 1999, and then a major boost in foreign currency petroleum deposits in 2000. With 86 percent of total deposits now in foreign currency, this reflects significant dollarization of the deposit base. The willingness to place funds with banks in the absence of a deposit insurance scheme would suggest that confidence may be returning, and that individual and corporate depositors are becoming less concerned about other issues noted above (e.g., account confidentiality). However, the placement with banks also reflects the lack of altemative investment opportunities. 18 This is not to say their participation did not equal or exceed those speculative investments made by the enterprise sector. There were two million participants in the pyramid schemes before their collapse. However, much of the investment appears to have come from sources that were not captured by formal monetary or banking measures. Thus, deposit withdrawals from the banks were mainly from the enterprise sector rather than from the household sector. 19 Banks paid between 34 and 37 percent on three-month lek time deposits in late 1997-early 1998. These were estimated to be positive in real terms when measured against the average CPI (32 percent in 1997), although less than year end CPI (42 percent at end 1997). 43 Figure 2.5 Deposit Growth Rrends Have Begun to Vary in the Four Subject Countries Year-to-Year Deposit Growth, Kyrgyz Republic Year-to-Year Deposit Growth, Albania _ Dorrestic Currency - Foreign Currency - Total - Donmestic Currency ---e Foreign Currency - Total 200% - 200% - = 1 60%160. 120% - ...1.... .. . ... ... ... 120% . . ......... ... ...... ..... ........... ... 80% _0% Ar . 40% 40% -40% -400% -80% -80% 1996 1997 1998 1999 2000 1996 1997 1998 1999 2000 Year-to-Year Deposit Growth, Moldova Year-to-Year Deposit Growth, Azerbaijan Dorrestic Currency -.- Foreign Currency - Total - DorresticCurrency -a- ForeignCurrency -o Tota 160% 160%. . 120% - ....... ...... ... 120% 80% 800 / .-- 40% 40% 0% 0% -40% - -40% -80% -80% - 1996 1997 1998 1999 2000 1996 1997 1998 1999 2000 Sources: IMF; central banks; World Bank calculations Notes: Albania: 2000 are projections; Azerbaijan: 2000 from September; Moldova: 2000 from June In the Kyrgyz Republic, there have been some positive trends in depositors shifting holdings from foreign currency deposits to local currency savings/time deposits. However, with only 5 percent of total deposits held in demand deposit accounts, this suggests that many households and small depositors have withdrawn their funds from the banks, and that they are not likely to return soon in any major way. This point is reinforced by the proportion of funds held outside the banks - 56 percent of total money at end 2000, up slightly from 54 percent at end 1999, and roughly the average of 1998-99. In Moldova, the economy has been weak for years. Notwithstanding major difficulties, banks showed increases in both domestic and foreign currency deposits in 2000, partly resulting from the remittance inflows from Moldovan workers employed in the Russian Federation. Since 1997, Moldovan banks have shown increases in foreign currency deposits, while local currency deposits contracted by more than 60 percent. By this measure, deposit levels nearly halved in real terms in 1998. This was felt across the board, reflecting obvious cash needs on the part of the public. 44 Depositors respond to real interest rates in differing Figure 2.6: Real Rates Paid on Local Currency Deposits ways, based on their own 400/o- Kr Replic personal/enterprise liquidity, 35 - their confidence in the 30%/o- - -_ currency, and their comfort 25% with the safekeeping capacity 2 - Ml"a of the banks (see Figure 2.6). AP In Albania, deposits have 5% increased significantly in the last several years. It is noteworthy that local currency -10 (, deposits have continued to m 1997 1998 1999 2000 increase as real rates paid Sources: IMF; central banks; World Bank calculations have declined. In Azerbaijan, N foreign currency deposits have Notes: all rates for comparable maturities; 2000 rates mid-year increased significantly and rea; rates paid on local currency deposits have declined, reflecting declining demand. In the Kyrgyz Republic, real rates on deposits have been highly positive, yet resource mobilization has been negative. Nonetheless, there are some promising signs of a return to confidence in 2000. Deposit mobilization trends earlier pointed to a dual trend in which demand deposit levels diminished to very low levels, while the net shift in funds gravitated towards foreign currency accounts, signaling a decline in confidence in the som. However, recent increases in som time deposit accounts suggest that confidence in the currency is being restored. In Moldova, real rates have been negative, as have deposit mobilization trends. In both countries, this points to limited liquidity on the part of most depositors, widespread poverty, and the vulnerability of these weak economies to trends in the Russian Federation. Moldova is now benefiting from remittance flows from and exports to the Russian Federation, but the economy overall is weak. 45 Borrowings and Capital Non-deposit funding for the banks is generally limited to Banks have low capital, borrowings from the central bank, as shown in Figure 2.7. This wea reflects low levels of capital, weak domestic securities markets, k domestic and no real link to international capital markets. In Albania, securities markets, and deposits account for 85 percent of total liabilities and equity. no link to international This is partly a reflection of underdeveloped securities markets, and portends fairly liquid debt and equities markets in the capital markets future should trends Figure 2.7 Structure of Bank Funding Sources continue. In the CIS countries studied, the ratios Structure of Banks' Liabilities and Equity vary, based on capital 581 figures. In Azerbaijan and 2 68 Moldova, deposits 175 i- accounted for about half of 150 --------------------------- total funding, whereas in the .2 125 08------------------ .... Kyrgyz Republic, deposits were about 72 percent of 100 . _ . . ............ ......................... .................... . ...................... *7 * _ ~ ,i _total liabilities and equity of ~~ 75 ~~~~~... ~~~the banks. Compared with 50 the five EU accession 12 5 E candidates, funding 0 structures are not all EU 5 CIS Albania Azerbaijan Kyrgyz Moldova together different because of * Total Deposits n Other Borrow ings 3 Net Capital the CIS banks' borrowings from their central banks. Banking System Liabilities and Equity However, EU accession 100%-- candidate banks are able to 90%- * - - access the syndicated market 80%- - - - for loans more readily than 70% ---- _ - CIS countries. More striking 60% . . f v ...... are the differences in per- _0% . bank values, as the EU accession candidates' 40% average liabilities and 30% -_ - _ capital were about six times 20% - those of Albanian banks, 10%- more than 60 times the 0% average among the CIS EU 5 CIS Albania Azerbaijan Kyrgyz Moldova countries studied, and more ONet Capital 0 Other Borrowings * Total Deposits than 20 times the average for all CIS banks. Sources: IMF; central banks; EBRD; World Bank calculations Notes: EU candidate figures based on 208 banks; CIS figures based on 2,063 banks 46 Although most countries have increased minimum capital requirements, actual net capital remains very low by global standards, both for systems and on a per bank basis. In Albania, capital for the banking system was still only about US$242 million in mid-2000, or US$18.5 million on average per bank. In Azerbaijan, total system net capital was US$269 million, or US$4 million per bank. In Net capital remains fact, most banks are miniscule considering that IBA, United low by global Universal and a handful of others account for most banking standards, both for system capital. In the Kyrgyz Republic, banks are very small and limited in all resource categories. Minimum capital was systems and on a recently raised to about US$1 million, yet average net capital per bank basis was less than half of that. Total net capital of the whole banking system in the Kyrgyz Republic was less than US$8 million. In Moldova, the average bank has US$ 4 million in net capital. By contrast, the five EU accession candidates show net capital of about US$108 million per bank, more than 13 times the average for all CIS banks, and more than 26 times that of the three CIS countries studied. Sources of financing other than deposits, capital and central bank borrowings are virtually non- existent for all four banking systems. In Albania, other sources accounted for only US$56 million, or 3 percent of total bank funding in 2000. Flows of Financing other private long-term (syndicated) loans have been zero since 1998, than deposits, and were only US$1-2 million in 1996-97 after Albania paid out a capital and central net US$29 in 1995. Some of the foreign banks operating in Albania have used their own resources for limited lending. In the bank borrowings CIS countries where other borrowings are about 20 percent of are virtually non- total, these are primarily borrowings from the central bank. In existent for allfour Azerbaijan, central bank borrowings constitute two thirds of other exnkinsvstentmfor al/borrowings, and in the Kyrgyz Republic, they are 87 percent. banking systems Moldova is a bit different, with borrowings from NBM only 15 percent of other borrowings. By contrast, the five EU accession candidates all have access to international capital markets, and this accounts for a major share of their US$35 billion in non-deposit liabilities. Levels of Lending Tables 2.3 and 2.4 provide a snapshot of credit "stocks" and "flows" for the four countries, and more broadly for the EU candidate countries and the CIS. Aggregate lending levels have been low in all four subject countries for several years. In Albania, bank lending at end 2000 was US$331 million, or 9 percent of GDP, and about US$25 million per bank (see Figure 2.8). 47 Table 2.3 Stock Credit Figures for Banking Systems for year end 2000 1 ~~~~~EU5 j I lai zerbaijan Iygy I I I l | | Republic | Moldova Total Bank Credit (mn, US$) 98,058 53,664 $331 $521 $58 $192 Private Sector Credit (mn, US$) 79,661 >30,665 $160 $122 $32 $149 State Sector Credit (mn, US$) 18,397 <22,999 $171 $399 $26 $43 Notes: State sector credit includes government plus state enterprises/farms; no data available for Tajikistan, Turkmenistan and Uzbekistan (CIS figures) Sources: IMF; central banks; World Bank calculations Table 2.4 Net Credit Flows: 2000 vs. 1995 (US$ million) EU 5 CIS Albania Azerbaijan Kyrgyz EU 5 C l .t Republic Moldova Net Total Bank Credit -$3,610 -$8,330 $58 $174 -$138 -$61 Net Private Sector Credit $13,329 -$1,729 $51 $94 -$149 $53 Net State Sector Credit -$16,939 -$6,601 $7 $80 $11 -$114 Notes: State sector credit includes government plus state enterprises/farms; data not available for Tajikistan, Turkmnenistan and Uzbekistan (CIS figures) Sources: IMF; EBRD; central banks; World Bank calculations Figure 2.8 Levels of lending per bank are About half of lending was to the private extremely low. sector and half to the state sector (mostly government). This represents a slight trend Bank Credit per Bank upward since 1998, when credit controls (US$ millions) were removed on private banks,20 yet remains low as a percentage of GDP. In 20 4 Azerbaijan, total bank credit was US$521 175 - -million, or 12 percent of GDP, and about 150 ---------- --- --------------------------US$7.5 million per bank. In the Kyrgyz 125 -------------- -----------------------Republic, bank credit was only US$58 1C0 ......... . ... million in total, equivalent to 4.5 percent ..i ... .i . . . .i of GDP, and only US$3 million per bank. D0 91 X a G | l I I_ I Low levels of credit reflect a combination 25 ! _ l l l _ 1 of limited resources available for lending, o -- Aa and tightening standards in the aftermath WJ5 as Abia an KWz M\bw of the 1998-99 banking crisis. Total bank, Sources: IMF; central banks; World Bank calculations 20 Following the pyramid scheme crisis of 1997, the Bank of Albania applied tight credit controls over all individual banks. These ceilings were removed in 1998 for most private banks. While the ceilings were non- binding in most cases, they were perceived as a deterrent to new investment by banks. 48 Figure 2.9 Deposits Exceed Lending to the Non- credit exposure in 2000 was only Gvrnmer . . about one-third of 1995 levels. In Government Sector In all Four Countries. odv,lnig wsU$9 . ~~Moldova, lending was US$ 192 million in total, or 15 percent of GDP, Credit Ratios and about US$9 million per bank. 500% Lending to non-government (the 400% enterprise sector and households) in 300% - all four countries is less than deposit 200% levels. In Albania, the amount of 100%- A outstanding credit to the non- % -- -- government sector stood at about /0% -US$175 million at end 2000 against EU 5 CIS Albania Azerbaijan Kyrgyz Moldova total deposits of about US$1.6 billion. Thus, deposit placements were nearly * Enterprise Sector Credit/ Deposits 10 times total credit outstanding to * Enterprise and Household Sector Credit/ Net Capital the real sector. In the enterprise sector (state and private), deposits were . ~double the credit received from the Sources: IMF; central banks; World Bank calculations banks a t ed 9.ine Afrbaija banks at end 1999. In Azerbaijan, banks' lending to enterprises (state and private) amounted to about 90 percent of total deposits in late 2000. The amount of outstanding credit to the Azeri non- government sector was about US$508 million in early 2000, compared with total deposits of US$567 million. Netting out bad assets shows that in 1999, deposits were 15-20 percent greater than loans. In the Kyrgyz Lending to Republic, which is among the very poorest of CIS and transition enterprises and economies, outstanding credit to the non-government sector was households is less only US$55 million as of June 30, 2000. These credit levels are now less than total deposits, partly as a result of asset quality than deposits problems that are being resolved. As of mid-2000, loan-to- deposit ratios approximated 81 percent, down from as high as 225-258 percent in 1995-96. In Moldova, outstanding credit to the non-government sector stood at about US$162 million as of September 30, 2000. Low as these credit levels are, they exceed total deposits held with banks by about US$10 million. The gap is presumably financed by foreign liabilities, central government deposits, and credit from NBM. Lending to the private sector is low in the four countries studied (see Figure 2.10), reflecting a wide range of real sector weaknesses. These include the lack of legal recourse banks have in recovering secured assets, poor information regarding enterprise Lending to the creditworthiness, weak management and governance in the private sector is enterprise sector, inadequate capital structures by firms (which translates into excess reliance on bank debt), and a tradition of loan low .. and has default and arrears. Among the four countries, only Moldova has a declined to the majority of its loan portfolio outstanding to the private sector. state sector However, this only amounts to US$149 million in total, or US$7 49 Figure 2.10 Lending to the State Sector is Declining, but million per bank. Remains Significant in all Four Countries. Lending to the state sector remains high in some countries, Credit to Private vs. State sector although it has declined as (% of GDP) governments have curtailed 45% directed lending and/or moved 40% - ---------------------------- forward with privatization. In 35%~ --------------------------- Albania, lending to state 30%. . . . ... ... . enterprises and farms was less 25% - than US$12 million in 2000, or 20% i n- about 3.5 percent of total loans. In 15%~ _ _ Azerbaijan, as privatization has 10%- * i been slower, the banks' exposure 5% i i _ to the state sector is higher, at 0% - ' about three quarters, and almost D O . <,, t e all of it is to the state enterprise < se 0 E sector. While proportionally high, this translates into only US$386 * Private Sector Credit/GDP O State Sector Credit/GDP million in loans to the state enterprise sector, or about US$5.5 Sources: IMF; central banks; World Bank calculations million per bank. In the Kyrgyz Notes: State sector credit includes govenmment plus state enterprises Republic, the government and farms; no data available for Tajikistan, Turkmenistan and curtailed directed lending in the Uzbekistan (CIS figures). early 1990s, as did Moldova. However, large-scale privatization slowed in the Kyrgyz Republic. Consequently, bank exposure to the state enterprise sector is 38 percent of total loans. In US$ terms, this is only $22 million, or about US$1 million per bank. While low in the aggregate, such exposures approximate three times net capital of the banks. In Moldova, lending has been low over the years, including to state enterprises and farms (since 1998). Banks' credit exposure to the state enterprise sector was only US$13 million in total in 2000. (Figures differ slightly from those presented in Table 2.3 because the table includes lending to the government as well. The text figures are strictly to the enterprise and farm sector, not government as a whole.) While banks traditionally have been used to finance a major proportion of deficits, lending to the state sector has declined in recent years. However, arrears are likely serving as the alternative to bank financing. In Azerbaijan, where the share of lending to the state sector has been comparatively high, the government has reduced fiscal deficits in recent years. Neither the central bank nor the commercial banks are being utilized to finance deep deficits. However, non-bank arrears are high, reflecting quasi-fiscal deficits. In Moldova, fiscal deficits are fairly low in the 3 percent range. Deficits are higher in Albania and the Kyrgyz Republic. Banks in Albania (mainly Savings Bank) hold a higher proportion of government securities. In the CIS countries, deficits are financed more through arrears. In the future, expansion of the treasury bill market may lead to an increase in bank investment in government securities. 50 Limited lending by banks has been the result of efforts to stabilize monetary policy, to ensure that problems of NPLs do not recur on the order of magnitude that has decimated banking systems in the past. This should create capacity for an increase in future lending flows to the enterprise sector in Albania, but there are questions about lending capacity in the CIS countries. In US dollar terms, total commercial bank credit in Albania to enterprises is estimated to be about US$172 million, or US$13 million per bank. This is low, yet from a safety standpoint, appears to be prudent. As a percentage of total deposits, such levels of credit to enterprises are small, at 10.5 percent of total. As a percentage of capital, these loans were only 1.5 times. Thus, banking system ratios are conservative at this point, and suggest that problem loans are now less of a risk to banking system safety than they were a few years ago on the condition that banks' exposure to government is safe. In the CIS countries, it is not at all clear if the same conditions apply. In Azerbaijan, 62 percent of loans were considered non-performing in 1999. Combined with problems most banks had in complying with minimum capital requirements, credit capacity is limited. In the Kyrgyz Republic, the banking system is small to begin with, and currently undergoing what is likely to be a long-term stabilization phase. In 2000, credit exposure was nearly eight times net capital, suggesting that there is little additional financial capacity to significantly increase lending flows. In Moldova, loans already exceed deposits, and are more than two times capital. Given the general weakness and vulnerability of the economy, it is unclear how much additional capacity is prudently available to the banking system. The Quality of Bank Assets and System Solvency The quality of the banks' loan portfolios is low in the CIS subject countries, and would be even lower if all loans were subject to the scrutiny implied by international standards. In Albania, loan figures are currently good, but this is partly due to the carve-out of bad loans in the state banks. New lending flows since have been limited in defense against a weak legal environment and the riskiness of lending in Albania. On a system-wide basis, the CIS countries studied have insolvent or near-insolvent banking sectors. Investments in government securities, at times the result of pressures to finance continuing budget deficits, are high, but do not represent either nsk-free returns or generate adequate income to permit recapitalization through internally generated profits. These conditions are unlikely to abate in the foreseeable future, without addressing the fundamental problems presented by connected lending, the culture of non- payment, state fiscal imbalances and the core problems of the enterprise sector. The Loan Portfolio Asset quality has been a major problem for all four banking systems at one point or another, usually the consequence of connected lending on non- commercial terms combined with a culture of non-payment. In Asset quality has Albania, NPLs at one point reached 61 percent of total for the been a problem state banks (1997), although this was only equivalent at the time for state and to 7.2 percent of their total deposits. They were as high as 54 percent in 1999, but have since been cleaned up. In Azerbaijan, private banks 51 NPLs reached 62 percent of total in 1999, or US$332 million. In the Kyrgyz Republic, NPLs were as high as 36 percent of total in 1998. In mid-2000, this had declined to 23 peicent, but this was still equivalent to US$13 million in a banking system with only US$8 million in net capital. In Moldova, bad loans were about US$25 million in 1999, about a third of net capital. Asset quality has been a problem for both state and private banks. However, due to the high level of balance sheet concentration of the state banks, banking systems have been weakened by poor state bank asset quality. In Albania, the state banks were the major problem, accounting for most problem assets. This is no longer the case, with a lending freeze imposed on the last state bank (which is currently in the process of selecting a privatization advisor). In Azerbaijan, problems were particularly acute at Prominvestbank and Agroprombank prior to their merger with Savings Bank into United Universal Bank in mid-2000. These two banks at one point had 90 percent of their portfolios classified as non-performing. As occurred in Albania with the state banks, credit controls were put in place, restructuring plans were implemented, and balance sheets were recapitalized. In the Kyrgyz Republic, several state banks have been closed down due to insolvency. In Azerbaijan and Moldova, private banks have been liquidated. These have been pocket banks operating on behalf of shareholders, managers, and related parties. However, the costs associated with these banks have generally not been as severe as the costs associated with the state banks. The magnitude of NPLs would have left all banking systems insolvent at peak, except possibly in Moldova. In Albania, ROA and ROE were negative through 1998 due to the poor performance of the state banks, although the private banks appear to Bad loans would have made profits. The liquidation of Rural Commercial Bank and have left most subsequent restructuring and privatization of National Commercial banking systems Bank and Savings Bank have fixed the problem of banking system insolvency. In Azerbaijan, bad loans approximated 1.4 times deposits insolvent atpeak held with the banks. If fully written down, NPLs would have eliminated the net capital of the banking system, leaving it US$65 million insolvent. Meanwhile, a straight 10 percent of total (adjusted) assets would have translated into an additional US$40 million needed to achieve a very basic capital adequacy measure. In the Kyrgyz Republic, bad loans were US$13 million. If written down, this would have brought capital to negative US$6 million. To comply with prudential requirements of a basic 12 percent capital adequacy ratio, it appears the banking system was at least US$15 million undercapitalized in 2000. In 1999, losses accounted for a decline of about 40 percent of capital. About half the banks were insolvent and/or experiencing serious losses. Moldova's NPLs reportedly peaked at US$55 million in 1996, or about two-thirds of net capital. Thus, while the system would not have been insolvent, it would have been severely undercapitalized. However, asset quality measures in 2000 show that the balance sheets of the major banks are still encumbered by a high proportion of illiquid assets, and that bad loans as a percentage of total credit are still a problem. The five major banks that account for 66 percent of total assets had half to two-thirds of their own assets in illiquid form in mid-2000. While this appears to be less of a problem for the other smaller banks, the smaller banks also show high levels of illiquid assets. Based on very rough estimates, the banking system had about US$175 million in illiquid 52 assets in 2000, more than two times net capital. If half of these assets were loss, than this would bring net capital to zero. The culture of default permeates all four systems, and needs to be addressed for long-term financial stability. The key problem for Albanian banks has been recovering loans from private sector borrowers. While The culture of state farms and enterprises were broadly considered to be in default permeates bad shape in the early-mid 1990s, their non-performing levels allfour systems have been contained, while the quality of loan exposure to the private sector deteriorated. The lack of recourse through the courts has proved difficult for all four systems. Meanwhile, in the Kyrgyz Republic (and presumably elsewhere), banks have been passive in trying to recover troubled loans until recently. Poor asset quality demonstrates that it is not enough to transform ownership classification through privatization. Proper governance, management, systems, discipline, disclosure and skills are also needed to manage credit and market risks. Without these prerequisites for success, there is a strong possibility that lending flows to private borrowers will eventually lead to problems of asset quality, as they have in Albania and most of the CIS countries today. However, here, there is often pressure on banks to lend even when borrowers and projects fail on tests of commercial viability, capital structure, cash management, and other fundamentals of prudent lending. A tightening of classification standards to bring them in line with international norms has brought to light the poor underlying condition of loan portfolios, banks, and banking systems. Inadequate asset classification and under-provisioning have overstated profits and capital on bank balance sheets. When finally recognized, the financial condition of banks and the consolidated banking system proves to be far weaker than earlier diagnosed. This, in turn, helps to explain problems associated with intermediation (e.g., low levels, high costs), and ultimately leads to different approaches to help restore capital and strengthen liquidity. This also reflects the shortage of personnel trained in risk management, both in the banks and among the regulatory community of these countries. This has been true in all four subject countries, as it has throughout the transition world. Investments in Government Securities Banks in the four subject countries have varying degrees of participation in the securities markets, principally government securities, although in no case are these holdings either extensive or remunerative enough to offset losses in the loan portfolio. In Albania, financing of government became the major activity of the state-owned banks after the pyramid scheme collapse, and prior to their privatization. This has weakened earnings. In Albania, banks' marketable securities were US$162 million, mostly held by the Savings Bank. With prudential requirements for capital adequacy at a minimum of 12 percent, the assignment of a zero risk weight for investment in T-bills, and the absence of alternatives, commercial bank investments in Treasury bills and government securities were about equal to bank credit exposure to the enterprise sector in 2000. Overall, these investments have helped to defray a sizeable portion of the government deficit, particularly in 1996-97. Since 1999, commercial bank investment in government securities has covered about half of the Government of Albania's fiscal cash deficit. 53 In contrast, banks' government securities holdings in Azerbaijan, the Kyrgyz Republic and Moldova, are limited. Rather than using banks to finance fiscal deficits, CIS countries have implemented fairly strict monetary policy, while running up arrears and settling accounts through barter and netting instead of through the banking system. In Azerbaijan, T-bill activity has only amounted to about US$55 million each year over the last two years. In the Kyrgyz Republic, banks' investment in government securities was little more than US$6 million. Discounting loan rates by 23 percent (comparable to NPLs), real net spreads on lending would be more profitable for banks by about 22 percent, given that T-bill rates were barely positive on a real basis. In Moldova, banks' investment in T-bills was about US$53 million. In 2000, nominal spreads on T-bills were negative against deposit rates. On a nominal basis, banks make more money on loans, but the ratio of provisions to net interest earnings equates to a discount of about two-thirds of the net differential. Thus, while banks may show a positive nominal spread on loan rates, the pre-tax differential has generally not exceeded 3 percent. At that rate in a market such as Moldova's, a key issue for management is whether it is worth the time, cost and risk of assuming credit risk when earnings from such risk are so limited. This challenge is made more problematic due to the limited alternatives for income generation. Trends in Bank Earnings Due to the poor quality of assets and limited range of financial services provided by the banks, earnings have been weak for years, although more recent figures show positive earnings in Albania. Returns on both assets and equity for banks in Albania throughout the decade were negative, and this applied to some private banks as well as the state banks. Losses resulted mainly from high loan arrears. With the imposition of greater discipline since the pyramid scheme collapse, Albania's banks posted profits in 2000 of about US$24 million. However, performance in the CIS banks has been weak. Azerbaijan posted only US$7 million in after-tax profits in 1999, and this could well be an overstatement of earnings with the high NPLs portfolio of the banks. Meanwhile, Kyrgyz Republic banks posted a loss of US$4 million, largely due to higher provisioning for losses. Table 2.5 summarizes the profit and loss indicators in the most recent year for which data are available for the four subject countries. 54 Table 2.5 Profit and Loss Indicators: 1999-2000 (US$) l Albania ^rAzerbaian Kr ublic Moldova Provisions for Losses $6 million $59 million $4 million $25 million After-tax Profits/Losses $24 million $7 million -$4 million N/A ROA 1.4% 1.0% -4.0% N/A ROE 28.7% 6.7% -0.4% N/A Notes: Figures for Albania from November 2000; Azerbaijan and Moldova from 1999; Kyrgyz Republic from March 2000; Sources: Central banks; World Bank calculations In the CIS countries, if loss loans were fully provisioned, earnings would weaken considerably, as shown in Table 2.6. Such losses would then fully deplete capital. In Azerbaijan, with 62 percent NPLs in 1999, adjusted results would have shown losses of US$266 million, a major difference from reported profits of US$7 million. Cumulative provisions from 1995-99 only amount to about US$122 million, or 37 percent of the outstanding stock of bad loans in 1999. Thus, at a minimum, the system should have provisioned another US$2 10 million. This, in turn, would have clearly brought ROA and ROE figures into deeply negative ranges. Adjusting provisions for the full amount of NPLs, ROA would have been negative 37 percent, and ROE negative 246 percent. Some of these have since been removed with the transfer of bad assets from the three state-owned banks that are now United Universal. However, there is a strong possibility that the banking system still ran up significant losses, and that these were pervasive throughout the system. Table 2.6 Adjusted Profit and Loss After Provisioning: 1999-2000 .-._._.-_-_ __. _ Azerbaijan _ Kyrgyz Rpublic i Provisions for Losses 332 million $13 million After-tax Profits/Losses -$266 million -$13 million ROA -36.7% -14.2% ROE -245.7% -1.5% Notes: based on figures in Azerbaijan from 1999 and Kyrgyz Republic from March 2000 Sources: Central banks; World Bank calculations In the Kyrgyz Republic, comparable adjustments would have produced losses of US$13 million instead of US$4 million. This suggests that the Kyrgyz Republic is actually making progress on its loan portfolio problems, as the difference is not all that great. In the last two years, Krygyz banks have had to provision for losses more aggressively than they did prior to the banking crisis of 1998-99. Provisions increased from 33.5 percent of total loans at end 1998 to 81 percent by March 31, 1999, and have since approximated 35-50 percent. Thus, one third to one half of net interest earnings2l have generally been consumed by whatever earnings Kyrgyz Republic banks have been able to squeeze from their loan portfolios. This has made it difficult for banks to generate the kind of retained earnings needed to strengthen their capital and reserves. 21 Net interest margin on the profit and loss statement. This is prior to other sources of earnings (e.g., commissions, fees, foreign exchange trading), which are very limited in the Kyrgyz Republic. 55 Earnings have also been constrained by the narrow range of fee-generating services provided. Most banks have not provided a wide range of lucrative fee-income services. In Albania, most private banks offer payment and transfer services, trade finance, custodial, payroll, and other common services. However, the absence of major trade and investment in most countries limits higher commissions from foreign exchange trading, mergers and acquisitions, corporate advisory services, funds management, and other activities that boost earnings in more developed markets. The savings banks of the four subject countries have not developed modem retail services, despite their extensive branch networks, a choice which has also limited their income levels. None of the four subject countries' traditional savings banks has introduced a broad array of electronic and plastic card services. In Azerbaijan, some retail innovation has come from IBA and Most Bank. However, these generally cater to Baku, and are not broadly distributed throughout the country. Likewise in Albania, electronic services and cards are available, but mainly in Tirana. Privatization with strategic investment of the Banca de Economii (the savings bank) in Moldova and the Savings Bank in Albania could change this. Likewise, the privatization of National Commercial Bank in Albania may also lead to improvement and modernization of services out of its branches. Real net spreads have been high for borrowers, but reasonable in most cases for banks considering their cost structures. In Albania, real net spreads have been 8-9 percent since 1999. In Azerbaijan and the Kyrgyz Republic, net spreads have been much higher than in Albania and Moldova, likely reflecting a higher proportion of non-earning assets (including higher reserve requirements). In Azerbaijan, real net spreads were 14 percent in mid-2000. In the Kyrgyz Republic, real net spreads were 37 percent, about double the spreads in 1999. This may reflect the increasing pressure on banks to meet minimum capital requirements. By contrast, in Moldova, real net spreads have been less than 1 percent since 1998. This likely reflects a general lack of demand, rooted in a weak economy and high arrears to non-bank creditors. Banks currently are in the distorted position of net spreads being positive based on negative real rates on both loans and deposits. This situation has prevailed at least since 1998, and possibly earlier. High levels of NPLs would justify higher costs of intermediation, although this would jeopardize enterprise cash flow (and credit quality) in a low inflation Earnings have environment. In Azerbaijan, net loan spreads in real terms against been constrained deposits have been attractive, at 10-20 percent since 1995. However, by the narrow the volume of performing loans at these rates has not been sufficient {, to generate strong earnings. Thus, if NPLs are factored in, banks' range offee- real net intermediation costs would approximate deposit rates paid. generating For instance, 1999 NPLs were 62 percent of loan stocks. If real loan services rates were discounted by 62 percent, the effective real loan rate would approximate 10.7 percent, barely above real rates paid on deposits. Following the same methodology, real loan rates were less than deposit rates paid in 1998. Thus, while the costs of intermediation are considered high by the real sector, banks find that existing spreads barely fund their costs of deposits. In the Kyrgyz 56 Republic, with NPLs at 23 percent of total credit and 45 percent of total reserve requirements unremunerated, this translates into non-earning assets of at least 29-30 percent22 in 2000 (net of fixed assets, etc.). This would require that banks charge at least 30 percent above the average cost of funding to break even on deposit mobilization. With nominal deposit rates averaging 29.6 percent, this would imply matched loans equivalent to at least 38.5 percent, nearly 2.5 times rates received on T-bills. Finally, reserve requirements have hurt bank earnings in all four subject countries. Reserve requirements were 13 percent in Albania, 12 percent in Azerbaijan, 20 percent in the Kyrgyz Republic, and 13 percent for banks in Moldova. In the Kyrgyz Republic, these are particularly high, since these apply to all deposits and remuneration for required reserves has been scaled back. BANKING SECTOR RESTRUCTURING AND RECAPITALIZATION Most banks that have failed or been recapitalized deteriorated as a result of insider/connected lending, mismanagement, poor systems, and bad governance. More often than not, the major problem banks in all four countries were state-owned. However, private pocket banks in CIS countries have also been closed down for the same reasons. In Albania, only one bank has been liquidated, a former state bank. The other two state banks have been (NCB) or will (Savings Bank) be recapitalized up to zero net worth prior to privatization. In Azerbaijan, more than 170 small banks have been permitted to fail, although the four largest have been recapitalized. In the Kyrgyz Republic, large and small banks have been permitted to fail, including comparatively large state banks as far back as 1995. In this sense, and combined with efforts to curtail directed lending in 1993-94, the Kyrgyz Republic has shown a low level of tolerance for many insolvent banks. However, in other cases, banks have been recapitalized, thus resolution experience has been more case-by-case than consistent and systematic. In Moldova, eight banks have failed and/or been liquidated, and others are expected to merge, grow or be shut down through consolidation in the coming years. In Albania, the government restructuring plan for the two state banks has explicitly focused on rehabilitation and privatization via strategic investment. Total cost will not be finally determined until the Savings Bank is privatized, likely in 2002. However, to date, costs have been about US$150 million. This implicitly accounts for some of the cost associated with the earlier liquidation of Rural Commercial Bank, which involved (i) the transfer liquid assets and liabilities from this troubled bank to Savings Bank, with gap later filled by T-bills, and (ii) the transfer of non-performing assets to a loan collection agency. In the cases of NCB and the Savings Bank, the government decided to float bonds equivalent to the value of negative capital figures to tenninate the banks' technical insolvency, and to enable their privatizations to move forward. 22 Deposits were 71 percent of total liabilities and equity. With 20 percent reserve requirements, this is equivalent to 14.2 percent of assets. Of this, 45 percent is unremunerated, equivalent to 6.4 percent of assets. On top of 23 percent NPLs, this equals 29.4 percent. 57 In Azerbaijan, the cost of gradual restructuring of the banking system has been felt more in the banks' sluggish earnings performance than in the fiscal cost of recapitalization. However, to date, the explicit cost of recapitalization has been US$175-200 million. This theme is found throughout CIS banking systems, where governments have generally not bailed out the banks, yet have failed to put in place adequate measures to protect against mismanagement, poor governance, low capitalization, and imprudent insider lending. In Azerbaijan, some of these characteristics have been in evidence, particularly weak governance, insider lending, and low capital. The Govermment of Azerbaijan has recapitalized some state banks, while permitting others to disappear. Bad assets of Agroprombank, Prominvestbank and Savings Bank (roughly estimated to be about US$125 million) were consolidated into Agroprom's portfolio, which is being assumed by a loan recovery agency under the Ministry of Finance. In the Kyrgyz Republic, recapitalization of the banking system has cost about US$85 million over the years. The current restructuring program is estimated to be less costly than earlier efforts. By 1996, recapitalization efforts had been estimated at about 5 percent of 1996 GDP, or about US$70 million at the time. Since then, the government has taken efforts to restructure and recapitalize potentially viable banks, liquidate others, and provide liquidity support in the interim. This has included about US$6 million in liquidity support for banks, US$9 million in assets removed from troubled bank balance sheets, and another US$4 million in bonds to Kairat Bank during the current restructuring program. General estimates of the overall cost of the rehabilitation and restructuring program for the banks were 1.0-1.5 percent of GDP, although it could turn out to be less with private recapitalization of banks, higher than expected loan recoveries, and a general reduction in NPLs. If estimates of 1.0-1.5 percent of GDP hold, the overall cost would approximate US$10-15 million. This roughly approximates estimates of the net insolvency of the banks, based on year-end 1999 NPLs at 30 percent. However, with NPLs down to 23 percent as of April 1, 2000, net insolvency would be closer to US$5-6 million. This might then mean the costs of recapitalization will be closer to half original estimates. In Moldova, the cost of recapitalization has not been fiscal as there has been no recapitalization of the banks. Discipline imposed in the banking sector has separated the overall problem of arrears of the real sector from the banks, although such structural and payments weaknesses have removed the potential market for banks to lend. This is reflected in weak funding of banks, and a general lack of commercial creditworthiness of the enterprise sector. Banks have faced the challenge of working through their old stock of NPLs, while limiting their new lending flows to enterprises that are able to service and repay. NPLs were in Most banks that the 10-15 percent range through 1999, although actual problems of performance may have been underestimated due failed did so due to to the practice of rollovers that made non-performing credits insider/connected appear to be performing. At 10-15 percent of total credit, non-performing loans to the banks represented about US$19- lending, 29 million in mid-2000, or about 1.8 percent of GDP. mismanagement, poor systems, and bad governance 58 THE NON-BANK FiNANcIAL SECTOR There is little In all four of the subject countries, there is very little capital markets activity. Banks constitute by far the most developed part capital markets of formal financial markets, notwithstanding the limited activity... banks intermediation and financial services capacity of almost all banks constitute the most in the subject countries. The Government T-bill market tends to dominate the securities markets, yet there is virtually no developed part of secondary trading in government securities, and there is no financial markets portfolio investment in the form of short-term capital inflows. despite limited Corporate debt securities are uniformly absent from the markets. The privatization programs have created a substantial float of intermediation and equity securities (including defacto derivative equity instruments, services in the form of the share options incorporated in privatization vouchers), as well as stockholders in great numbers. However, there is no transparent, liquid mechanism either to value or to trade these equity instruments fairly. Where formal stock exchanges have been created, they tend to focus on foreign exchange trading, with very limited corporate listings and even less activity. Contractual savings institutions are in the early stages of development beyond the limited roles they played under the socialist system. Substantial progress remains to be achieved in creating a legal, financial and business foundation for such activities before they can participate more broadly in the financial sector (as investors) and the economy (as service providers). Box 2.5 CHALLENGES IN THE NON-BANK FINANCIAL SECTOR Area Challenges Capital Markets * Legislation is uneven and untested Infrastructure * Regulators are inexperienced and not well-coordinated with banking regulators * Enterprises cannot satisfy listing disclosure requirements Volume of * Listing and trading figures are small in value Activity * Off-market trading is comparatively significant, yet not well-regulated * Government securities investment that fund budget deficits crowd out corporate and other issuers Lack of * Political, commercial and market risk seen as too high for domestic and Investors international portfolio investors * Contractual savings systems not yet sufficiently developed to produce local institutional investor demand Contractual * The insurance sector is underdeveloped and under-regulated Savings * Foreign participation is sometimes discouraged in the insurance sector * Pension reform is hampered by the fiscal limitations on governments, as well as political issues 59 Securities Market Development Market Infrastructure In several of the subject countries, the legal and regulatory infrastructure is weak for securities market development. In Azerbaijan, the corporate sector's financing needs have traditionally been met by bank loans or from other channels. Thus, there is no active tradition of listing on securities exchanges, nor is there a tradition of meeting meaningful disclosure requirements to support corporate debt There is no active or equity activity. The Kyrgyz Republic has introduced a tradition of listing reasonable framework for securities listings, disclosure requirements, share depository, and clearing and settlement on securities mechanism. The weakness in this country has more to due exchanges or with macroeconomic vulnerability, real sector structural meeting disclosure problems, and general financial weaknesses. From an infrastructure standpoint, the major weakness is lack of requirements experience and testing due to the limited activity of its markets. In most countries, adequate listing requirements are in place for stock exchanges, yet companies are unable to meet the standards. This reflects weaknesses and lack of competitiveness by global standards in the real sector, as well as an absence of useful accounting standards. In Azerbaijan, the number of potential issuers that are financially sound is limited, and none so far has been able to comply with the strict listing requirements in place for the BICEX, apart from two commercial banks. Even here, the banks are only in partial compliance with listing requirements. In the Kyrgyz Republic, only a few companies can meet the KSE's main exchange listing requirements. There were 69 firms listed as of March 2001. In Moldova, 58 of Moldova's 840 registered joint stock companies were listed on the main exchange at end 1999, up from 21 as of end 1998. Under securities legislation, non-listed companies with more than 300 shareholders must be traded on the exchange. Market Size The securities listed and the market capitalization of stock exchanges in all four of the subject countries is very low. In Albania, there are no listed companies, therefore no real capitalization. In Azerbaijan, there are two listed companies, at about US$2 million each. In Moldova, there are 207 companies listed, of which 58 are on the major exchange. The average per company capitalization is less than US$1 million. The Kyrgyz Republic had the highest relative value, at about 24 percent of 1998 GDP. However, this collapsed in 1999. Market capitalization was only US$232 million at end 2000, or US$3.4 million per listed firm. These values compare with US$62 billion in total market capitalization for the five EU accession candidates, about $1,000 per capita, and 27 percent of GDP on average. On the strength of market capitalization figures in the Russian Federation and a few other CIS countries (e.g., Kazakhstan recently), aggregate CIS market capitalization is estimated to be US$76 billion. This suggests that more highly 60 populated and resource-based CIS economies have greater potential for the development of functioning securities markets, an important lesson for Azerbaijan. Trading volume on the securities exchanges is predictably low in all four subject countries, and usually concentrated in a few firms. For example, in the Kyrgyz Republic, trading volume in 2000 approximated US$23 million. In Moldova, the volume of trade was only US$16 million from January-July 2000. Table 2.7 summarizes securities market indicators for three of the four subject countries, the EU accession candidates and OECD countries. Table 2.7 Summary of Securities Market Indicators OECD EU 5 CIS Azerbaijan Kyrgyz _______ ------ Republic iMoldova Stock Mkt Capitalization (US$, mn) $32,291,799 $61,659 $75,772 $4 $232 $38 Stock Mkt Capitalization/ GDP 141% 27% 19% <1% 24% 3% Stocks Traded (US$, mn) NA $26,318 $3,060 N/A N/A >$16 Stocks Traded/GDP 95% 12% <1% N/A N/A >1% No. Firms Listed on Major Exchange 22,325 504 508 2 69 58 Notes: Albania has no stock exchange Sources: World Development Indicators; EBRD; Kyrgyz Republic Stock Exchange; OECD Off-market trading is Off-market trading is more active, but subject to significant more active, but pricing distortions and a lack of transparency. In the Kyrgyz subject to significant Republic, off-market trading was estimated to be about 30 times the trading volume on the KSE in 1998. This ratio may pricing distortions have decreased since then, but on-market trade remains at low and a lack of levels, as it does in the other CIS countries studied. To enhance transparency transparency and increase liquidity over time, the Kyrgyz Republic is planning to develop an electronic OTC market. Market development has been slowed by the ability of enterprises to obtain financing from other, informal sources, mainly in the form of arrears. Traditionally, enterprises were able to obtain funding from the banks, while the banks would rely on funding from the central bank or the government. With the tightening of monetary policy and banking standards, patterns shifted, more often than not leading to the build-up of arrears. These and other factors have led to behavior that is not creditworthy, undermining the development of market-based standards in all four subject countries. Almost all of the market capitalization of "stock" exchanges is Almost all market from governnent securities. As of November 2000, T-bill capitalization oj market values in Albania approximated US$770 million, of "stock" exchanges which about 20 percent was held by banks (mostly the Savings Bank). In Azerbaijan, the government securities market had is from government only about US$55 million in turnover in 1999, and about the securities 61 same in 2000. Only about three quarters of supply was purchased in 2000,23 and only two-thirds in 1999.24 By extension, there is also no significant secondary market for the small T-bill market.25 In the Kyrgyz Republic, the total value of T-bills and medium/long-term government securities was estimated to be about US$78 million in mid-2000, about one-third the market capitalization of the Kyrgyz Republic Stock Exchange. Of this, the banks held about 8.5 percent of total, equivalent to about US$6.6 million, or roughly 7 percent of total assets of the banking system. As of June 2000, the stock of T-bills outstanding was US$53 million, about one third of total internal debt. Equity values and turnover are minimal on exchanges. In Azerbaijan, Equity values privatization vouchers have been virtually the only securities traded. and turnover Market capitalization for these was only about US$80 million in 2000, are minimal and there is no organized market. Net of privatization vouchers, about US$4 million in equities were listed on the BICEX, although there was no on exchanges trading in 1999 or 2000. In the Kyrgyz Republic, most trading takes place off the main exchange. Trading usually involves privatization vouchers that were issued earlier in the 1990s as part of the 1,200 companies in the mass privatization program. The Moldova Stock Exchange has had to share stock trading with the unregulated over-the-counter market, which accounted for 48 percent of transactions in 1999. Direct monetary control in recent years has been part of the reason why markets have not developed, particularly in the CIS countries. In Azerbaijan, the T-bill market is generally illiquid, and yields are intentionally kept low as a function of deliberate interest rate controls. In the Kyrgyz Republic, securities market development has been and will be constrained by fairly tight monetary and fiscal policy and the absence of overall market development. According to the Poverty Reduction and Growth Facility (PRGF) framework, the government is expected to bring down the inflation rate from 14 percent in 2000 to 7 percent in 2002, while fiscal deficits (cash basis) are expected to decline from 7 percent of GDP in 2000 to less than 6 percent by 2002. This tightening will reduce the scope for securities development in the absence of major direct or portfolio investment, neither of which is anticipated. Liquidity Liquidity is in short supply for capital markets development, due to economy-wide demonetization and the absence of institutional investors. In Albania, the major insurance company is the only firm at the moment that represents a potential institutional investor. This may change in the coming years as Albania proceeds with financial market development. In Azerbaijan, the Oil Fund is the closest equivalent to a potential institutional investor. However, the large placement of deposits with IBA signals the likelihood that the government will direct resources from this fund, rather than encouraging a more market-based investment approach. 23 T-bills sold were 72 percent of the supply offered in the first eight months of 2000. 24 T-bills sold in 1999 were 67 percent of the supply offered. 25 Secondary trading in the T-bill rnarket in 1999 was less than US$ 100,000. 62 Restrictions on the entry of foreign insurance companies also constrain development of institutional investors. In the Kyrgyz Republic, for example, insurance companies and private pension funds are virtually non-existent. In Moldova, this is also true, despite the privatization of its main insurance firm with strategic investment. Several of the subject countries are planning to diversify activity in the government securities market, driven by a shift in monetary management based on the use of indirect instruments through open participation in the market. In Albania, the govermment plans to increase the participation of private banks in the lek deposit market and T-bill market, strengthen the functioning of primary markets by increasing transparency, improve the marketability of T-bills through retail outlets (e.g., post offices) for greater household participation in the market,26 and foster development of secondary markets for government securities. In Moldova, the implementation of monetary policy is expected to rely increasingly on open market operations of T-bills, which should stimulate the development of the secondary market. Non-residents are permitted to participate in the T-bill market in all four subject countries, reflecting desirable openness. This will improve prospects for development in the future, but will require proper regulatory oversight to ensure the risks of large, unanticipated capital inflows and outflows are contained. In Azerbaijan, there has been virtually no international investment in the T-bill market. Portfolio investment has been zero since 1999, and was never greater than US$1.1 million (net) in 1997. In Albania, the Kyrgyz Republic and Moldova, nonresident participation is encouraged. With large-scale privatizations planned for these three countries, this could establish some momentum for nonresident investment into the markets. Prospects are varied for development, with Albania and Azerbaijan showing the best prospects among the four subject countries from a financial resource perspective. In Albania in particular, there is substantial potential to develop a liquid market due to high levels of remittances from abroad. Private transfers averaged about US$375 million from 1998-2000, which was about 1.4 times merchandise export earnings, about 3.5 times official transfers,27 and nearly six times direct investment during the samne period. In Azerbaijan, these prospects are tied to the oil sector and the potential to introduce well capitalized firms on to the equity markets. There is also potential to harness transfers from the substantial cross-border trade that occurs with ethnic Azeris and others in nearby countries, but that is currently unrecorded in the balance of payments. In the Kyrgyz Republic and Moldova, prospects are less favorable due to more vulnerable income and investment streams. In the Kyrgyz Republic, privatization of several large-scale enterprises may stimulate investment, but it will be a while before an active securities market materializes for corporate debt and equity. Meanwhile, most transfers are official rather than private, and private transfers have generally been outflows rather than inflows. In Moldova, remittances have increased in recent years from workers transferring funds from the Russian Federation. However, overall, balance of payments inflows have been weak in Moldova. 26 Households held less than I percent of T-bills in early 2000. 27 Official transfer figures only apply to the capital account. Current account figures for 1999 official transfers related to the Kosovo crisis are not included. 63 Contractual Savings: Insurance and Pension Funds The insurance sector is severely underdeveloped in all four of the subject countries, particularly in the CIS countries. In Albania, motor vehicle insurance accounted for about 85-90 percent of premium revenues. Investment assets approximated about US$35 million in 1999. The government has plans to open up the insurance market with the privatization of INSIG, initially targeted for the middle of 2001. Efforts are expected to include a review of the legal framework and investment promotion. In Azerbaijan, the insurance sector is underdeveloped. Barriers to entry have been in place for foreign insurers, which severely stunts development of the sector. In the Kyrgyz Republic and Moldova, the insurance sector is likewise underdeveloped. In Moldova's case, the attraction of foreign investment into the sector can potentially be helpful. However, such investment will have limited effect in the absence of adequate regulation, information (statistics and data) for effective actuarial tables, and purchasing power and functioning capital markets for investment. Some of the CIS countries have initiated reform of their pay-as-you-go (PAYG) pension systems. In general from 1995-99, pension expenditure in Azerbaijan has risen steadily as a share of GDP, from 1.8 percent in 1995 to 4.2 percent in 1999. It is possible that 2000 represents the first year in which the growth rate relative to GDP has been reduced since 1995, although GDP has increased and the 2000 rate remains twice what the relative pension expenditure was in 1995. Major problems include low collection (and high evasion), utilization of the state pension fund for social protection as well as social insurance, and the absence of a central registry that is automated and up to date with information on contributions. The government has managed to contain expenditure by gradually increasing retirement ages, imposing ceilings on reference wage rates, and introducing selective indexation rules. In the Kyrgyz Republic, pension expenditure exceeded revenues by 32 percent in 1997. Thus, increasing budgetary transfers and/or accumulation of arrears was the only way to fund pension obligations. Recognizing this as unsustainable, the government reformed the PAYG system. Reforms have included increases in the retirement age, and establishment of individual accounts as a vehicle for contributions to supplement the base pension. In Moldova, efforts over the last few years have involved cost containment, and paying off arrears to pensioners. All four subject countries need to reform their pension systems for long-term sustainability. In Albania, there is substantial need from a fiscal standpoint to move on with pension reform. The 2000 budget shows social insurance outlays (including health) to be 5.7 percent of GDP, and 18 percent of total expenditure. However, in 1999, pension costs were 26 percent of total fiscal expenditure and 6.8 percent of GDP. In Azerbaijan there is only one social insurance fund, and this accounted for 20 percent of state budget expenditure and 3.4 percent of GDP through 3Q 2000. This was 97.5 percent of the existing official social safety net in Azerbaijan. In the first half of 2000, more than 40 percent of these expenditures were in the form of transfers from the state budget. In the Kyrgyz Republic, deficits and arrears have Alli four countries come down but not been eliminated. The replacement rate was need to reform brought down from 47 percent in 1998 to 40 percent in 1999, but their pension still remains high. Meanwhile, total state budget transfers and systems for long- arrears were projected to approximate 0.5 percent of GDP in 2000 term sustainability 64 and be zero in 2001, as compared with an average 1.2 percent per year from 1994-99. In Moldova, there is substantial fiscal need to move on with pension reform. The government ran up significant pension arrears in the past, leaving much of the population poorer and illiquid. While there was some initial pension reform in 1998, there has been limited progress to date in building a new and sustainable system. The 2000 budget still showed social insurance outlays to be 13 percent of total expenditure. While these figures are not excessive, Moldova lacks the resources to operate a PAYG system sustainably. There is potential to develop a contractual savings market in all four subject countries. In Albania, deposits have increased in recent years, partly because banks appear to be the only formal place for safekeeping peoples' savings. By offering alternative instruments, some of these savings would be expected to move to longer-term instruments. Additional funds not placed with banks would also likely wind up in comparable instruments. In all the subject countries, banks may be reluctant to provide long-term savings instruments for the foreseeable future. This niche can be filled by life insurance companies and private pension funds. As with the development of viable securities markets, all four subject countries will need to establish a framework for market confidence to take hold for the contractual savings market to develop. This will include a separation of life and non-life activities, clear investment policy guidelines, and strong consumer protection provisions. Introduction and development of regulatory and supervisory capacity will also be needed, learning from the lessons and mistakes of the past in the banking sector. Thus, solvency and liquidity thresholds, governance and management standards, and licensing requirements will need to be clear and prudent. Meanwhile, in all the subject countries, restoration of confidence in the banking sector is tied to progress in the non-bank markets, particularly as these relate to long-term savings. The benefit of progress in the banking Allfour countries sector is that it should spill over with time to the insurance and , private pension fund market as additional vehicles for the need to establish a safekeeping of savings. framework for market confidence for the contractual savings market to develop 65 ALBANIA Current Status: The non-bank financial sector is less developed in Albania than the country's banking system. The money and capital markets consist only of primary issues and a narrow secondary market for Treasury bills. As a department of the Bank of Albania, the Stock Exchange currently operates only as the servicing entity for Treasury bill issues. The insurance industry continues to be dominated by INSIG, the original state monopoly, although there are now competitors that are strong enough to challenge INSIG's dominance in the vehicle insurance market (which is most of the market in Albania). There are two investment funds that have purchased privatization vouchers, but both are presently inactive and waiting for the expected privatization of state-owned companies. Prospects: Albania is moving to privatize INSIG, which may help with the establishment of a more prudently and privately managed insurance sector. However, it will take time to develop the framework and marketplace for insurance to play a major role outside the vehicle insurance sector. Meanwhile, reinsurance will continue to play a key role, which means foregone revenues for the domestic insurance industry. Albania is several years away from having a developed securities market, which will also serve as a disincentive to development of insurance and private pension funds. AZERBAIJAN Current Status: Azerbaijan has virtually no securities or capital markets, no private pension funds, and a limited insurance sector. There is slow movement to activate the T-bill market, although fiscal prudence and monetary control have limited the volume of trade on this market. Slow privatization in the medium- and large-scale SOE sector has constrained the growth of equities. Restrictions on foreign insurers constrain modernization of the insurance sector. From a regulatory and institutional standpoint, structures and a framework are not yet in place. Meanwhile, where sound procedures and standards are in place, such as for the issuance of corporate securities, there are no firms that can fully comply with requirements. Prospects: There are plans to develop a multi-pillar pension scheme in the coming years, but not imminently. KYRGYZ REPUBLIC Current Status: The non-bank financial sector is embryonic, apart from the limited T-bill market. While the Kyrgyz Republic introduced mass privatization early in the 1990s, there has been limited securities market development to date. Likewise, the insurance sector is inactive. To the Kyrgyz Republic's credit, it has put in place reasonable legislation and regulations, and sought to shore up institutional capacity for enforcement in the aftermath of the Russian crisis in 1998-99. Prospects: Recent efforts to rationalize the pension system bode well for future sustainability. However, all parts of the financial sector suffer from severe liquidity constraints, and this makes it difficult for banking and non-bank financial services to emerge. MOLDOVA Current Status: The non-bank financial sector is virtually non-existent in Moldova, apart from the thin T-bill market. The securities market functions on a very limited basis, primarily in the government securities sector. There is very little information on the insurance sector, and most of this is focused on motor insurance. Predictably, there are no major institutional investors providing contractual savings, although Moldova has attracted foreign investment from Australia into the sector. Prospects: The Government has introduced basic pension reform in 1998 to make the existing system more financially sustainable, but movement towards a multi-pillar pension scheme has stalled. Accordingly, the prospects for a significant expansion of the non-bank financial sector are dim. 66 SUMMARY OF FINANCIAL SECTOR CHALLENGES, ISSUES AND RISKS All four subject countries have emerged from major shocks and challenges in recent years. All four have shown progress, although there are still several weaknesses and challenges that could undermine stability. Coming off of the pyramid schemes just four years ago, Albania's reform program has shown progress in introducing confidence-building measures that are intended to put the banking sector on a viable and sustainable track. Azerbaijan has stabilized its macroeconomic framework over the last five years, and been in the fortuitous position of attracting major investment into its petroleum sector. The Kyrgyz Republic has strengthened the financial position of the banks, and moved to introduce strengthened institutional capacity. In Moldova, progress has been made as part of a deliberate and well-planned reform program in the banking sector. The general environment for risk-taking will remain weak in all The environment four subject countries for the foreseeable future. In Albania, new for risk-taking will laws have been adopted, including in the area of secured remain weakfor the transactions, but it will take time for institutional capacity to develop and demonstrate that the environment provides adequate protection for creditors. Until then, banks will be less likely to assume traditional risk, which will limit the amount of lending that banks will be willing to provide. In Azerbaijan, lending is undermined by a poor business environment, a culture of loan default, weak collateral enforcement, and poor financial information for risk assessment. In the Kyrgyz Republic and Moldova, it will take time to build judicial capacity to provide risk-takers with the confidence they need to assume risk. Inadequate and incomplete accounting information from potential clients remains a problem for banks. A culture of default and arrears predominates. Even where companies seek to comply with credit agreements, there is substantial market risk associated with the vulnerability and fragility of the Kyrgyz Republic and Moldovan economies. There is a general risk that banks all want to be universal. This is understandable, in light of the limited investment and fee-based earnings opportunities most of them face. However, because they also lack the experience and management systems, there is a risk that banks might seek to move aggressively into markets (e.g., leasing, factoring, smart cards) without being able to manage the risks involved. Capital remains extremely low. While most countries have tightened minimum capital and CARs, it does not appear that aggregate capital will grow substantially for quite a while. This is because all four markets remain relatively small, earnings sources are limited, and there is no active securities market or serious institutional investor interest in the banking sector. In Azerbaijan and the Kyrgyz Republic, many banks have been unable to meet minimum capit2 requirements, which are low by international standards. In Moldova, most banks barely meet minimum capital requirements. Asset quality could be undermined in the future by the continuing problems in the real sector, the weakness of information provided by companies, weak credit risk management skills of the 67 banks, and in some cases, political pressure on banks to lend. Information is universally poor from the enterprise sector in all four subject countries, and credit risk management capacity has been passive and reliant on inaccurate collateral values. Meanwhile, privatization of major banks in Albania and Moldova and low lending levels in all four subject countries may lead to pressure on banks to lend to stimulate economic growth. NPLs have deeply hurt overall asset quality and eamings, as shown in banks' poor ROA/ROE measures despite high intermediation margins. As banks more accurately provision for loan losses, profits will continue to be undermined. In Albania, banks are beginning to show improved performance. However, the eamings performance of the CIS countries has been poor for years. Banks in all four subject countries will have to address the Banks will have challenge of a small market, a culture of default, and the risk of to address the adverse selection to compensate for a small volume of eaming assets. Meanwhile, borrowing costs are high for companies. challenge of a Even if they are well managed, the high costs of borrowing small market, a alone present financial risk to credit quality. Macroeconomic vulnerability adds to this risk in the CIS countries, although culture of default, movement towards hard currency has offset some of the risk. and the risk oj Movement towards privatization in the financial sector has led adverse selection to the introduction of new incentives for improved bank to compensatefor govemance and management. However, it remains to be seen a small volume oj how management will fare as banks assume greater balance earning assets sheet exposure over time. Even with reforms, risk management systems are considered fairly modest in all four subject countries. Banking supervision will need to strengthen and play a key oversight role as long as the market and institutional capacity remain underdeveloped. In the future, it is expected that banks will be professionally managed, diversified, competitive, and stable. However, it will be years before modern capacity is comprehensively in place. Until then, it will be important for banking supervision departments to strengthen early warning and off-site surveillance capacity, and to build on efforts to link on-site examination strategies and timing with a focus on key risk areas as ascertained from regular off-site reviews and assessments. Governance and management remain weak in the real sector, which means lending to the private sector is no panacea for the banking system's ills. All four subject countries still have large-scale privatization transactions to initiate. Meanwhile, many private companies are, in fact, privatized companies that have not yet sufficiently restructured to be credit worthy. Other private companies are often subject to poor standards of governance and management, and therefore represent serious credit risks. This points to the need for comprehensive reforms in all subject countries, and the insufficiency of financial sector reform in the absence of a more disciplined incentive structure for enterprises. 68 In countries that rely on strategic investment, there is the potential that strategic investors will either misread the market or be excessively risk-averse, causing disappointment. This would appear to be a potential risk in Albania with the privatization of its two state banks. Performance in the insurance sector of Moldova shows strategic investment is not sufficient for development. As found in other markets, not all strategic investors succeed in transition economies. Nor do they all remain in these economies when competition intensifies, or if the market does not develop as anticipated. Banks will need to Most of the subject countries are planning to move to a more open market basis for the implementation of monetary policy. Under strengthen risk such circumstances, banks will need to strengthen risk management management systems, and demonstrate their ability to manage ystems, and interest rate, exchange rate, pricing and maturity risk. This should not be hard to achieve as they represent core preconditions for demonstrate their successful banking. However, along with pricing risk, this has ability to manage often caused major losses in advanced institutions. Given the high level of sensitivity to commodity output in the CIS countries (e.g., oil in Azerbaijan, gold in the Kyrgyz Republic, agriculture in exchange rate, Moldova), there will be constant risk that sharp movements could pricing and materially impact the relatively small portfolios of these countries' risk banks. This, in turn, could severely affect liquidity, and lead to insolvency in the absence of adequate capital and reserves. In small and comparatively open markets like Albania and the Kyrgyz Republic, this can spread more quickly due to higher levels of sensitivity to market risk. Foreign exchange risk appears to be a particularly important risk in Azerbaijan, with nearly 90 percent of banking system deposits in the form of hard currency. This also applies to the other CIS countries, as the Kyrgyz Republic and Moldova have about half their balance sheets in foreign currency. In Azerbaijan, banks' open foreign exchange positions were 35 percent of total capital at end 1999, despite prudential regulations limiting net exposure to 30 percent. Even more importantly from a risk standpoint, any erosion in the exchange rates of net assets held in dollars and other foreign currencies would subject the banks (and, potentially, the system) to foreign exchange risk. Foreign exchange-related risks are also linked to fundamental credit risk. Again, Azerbaijan may be the most sensitive to these risks. Traditionally, banks have matched foreign currency loans and deposits. However, these loans have generally gone to state enterprises, which is also the sector most responsible for NPLs. Should SOEs default for foreign exchange reasons, this would present a major cost to the banks, the central bank, and/or the government. Liquidity management has been fundamentally weakened and exposed to unnecessary risks in all four subject countries due to weak payment and settlement systems. In Albania, this has been due to reliance by state banks on manual settlement and physical transport of cash and documentation. Lack of integrated branch network systems has also raised risks about internal bank data and information. Banks in Azerbaijan will need to modernize their liquidity management systems and practices as plans to modernize the payments system materialize. In Moldova, liquidity management has been weak due to the non-cash approach to settlement. 69 There are funding concentration issues in some countries that could interfere with inter-bank markets once developed. In Albania, Savings Bank's deposit market share undermines other institutions' funding and liquidity. In the future, such predominance in the T-bill and inter-bank market could subject the market to liquidity problems if Savings Bank's asset management practices are unsafe and Concentration excessively risky2 . In Azerbaijan, IBA's position in the foreign currency deposit market could potentially undermine issues could other institutions' funding and liquidity. While the inter-bank interfere with inter- market is small now, IBA could eventually assume a major bank markets once position. If so, many smaller banks' liquidity positions could be sensitive to the financial condition of IBA. In Moldova, a developed 30 percent limitation on asset and deposit concentration has been imposed. Small-scale intermediaries have generally been unregulated, and could potentially become a risk to the financial system. In Albania, savings and credit associations have not been required to report to the banking supervision department of the Bank of Albania. This has been identified by the IMF as a potential risk to the quality and timeliness of monetary data. In light of Small-scale Albania's past history with pyramid schemes, it will be Small-scale important to ensure that these institutions are not used to siphon intermediaries have off funds from the public, engage in pyramid schemes, or been unregulated, engage in any other activities that could undermine financial sector stability and public confidence. In Azerbaijan, a law and could become a permitting credit unions to operate was passed in 2000. In the risk to the financial future as these emerge, the National Bank of Azerbaijan will system need to ensure these and other smaller institutions are not used as unregulated channels for deposit-taking that could lead to a revival of pyramid schemes. Savings and credit associations have not become a risk issue in the Kyrgyz Republic or Moldova. Should they or other micro-credit institutions emerge in any meaningful way, they will also need to be subject to proper oversight to protect against abuses. These are partly offset in Moldova by village ownership, which has also emerged in rural Albania as an effective alternative at the micro level to bank financing. The reputations of all four subject countries for perceived corruption are weak by global standards. As one example, Transparency International ranked 99 countries by perceptions of corruption. Of these, Albania ranked 84th (along with Georgia and Kazakhstan), Azerbaijan ranked 96th (along with Indonesia), the Kyrgyz Republic ranked 87th (along with Uzbekistan and Federal Republic of Yugoslavia), and Moldova ranked 75th (along with Ukraine and Vietnam). By contrast, Denmark and Finland were first and second, while Nigeria and Cameroon were 98th and 99th, respectively. All four governments have tried to improve their country's image abroad. However, notwithstanding recent cooperation between the central bank and international authorities, Albania has a reputation for money laundering, smuggling and 28 This risk could be two-way. If Savings Bank's financial position deteriorated, other banks' financing would decline, perhaps causing a liquidity panic or shock. Likewise, if Savings Bank were overly exposed to banks whose financial condition deteriorated, this would raise the risk of Savings Bank experiencing a liquidity problem, which could then lead to panic in the deposit market. 70 fraud that undermines the willingness of strategic investors to commit resources. Over time, as Albania seeks to join European and global institutions, there will be pressure to align systems with standards recommended by BIS and others to protect against financial crimes. In Azerbaijan, the government has been effective at stabilizing the macroeconomic framework and introducing some legislation conducive to foreign investors. However, Azerbaijan has a poor reputation for corruption. The Kyrgyz Republic and Moldova have created favorable incentives for foreign investment and the central banks have positive reputations, yet their countries are also encumbered by a reputation for corruption. In the case of Moldova, the political dispute over Transnistria has permitted a channel for money laundering, smuggling and fraud. Finally, regardless of the effectiveness with which the four subject countries implement reforms that improve the infrastructure and policy environment for banking, it is inescapable that the fate of banks is tied inextricably to the condition of the banks' customers - the enterprises. Unless the health and creditworthiness of enterprises improves, banks will have only two choices. They can continue to serve as a reservoir for the problems of their borrowers, with large volumes of NPLs, limited funding opportunities, and cost structures which effectively prevent them from growing out of their problems. Alternatively, they can withdraw from funding the real sector, on the grounds that the risks are too high, and focus instead on funding the government and seeking the safest, highest return investments for what funds they manage to attract. This is largely what has happened in Moldova. However, in such a case, these financial institutions will no longer be "banks," and they certainly will be in no position to provide the financial resources for recovery and sustainable growth in the economy at large. 71 Box 2.7 A COMPOSITE OF MARKET RISK IN THE FOUR SUBJECT COUNTRIES ALBANIA: Efforts have been made to contain risk-taking behavior, and to hold management accountable for prudent standards. There are no major exposures to external markets, although interest rates in Greece, Turkey and Italy have an effect on local interest rates. Now that Kent Bank (of Turkey) is entering the market as a strategic investor, it is anticipated that this will help with more sophisticated risk management systems. Market risk in Albania is generally internal, as shown in the pyramid scheme collapse in 1997. However, increasing openness of the system will eventually increase vulnerability to movements and developments offshore. Meanwhile, in the future, if remittances/transfers are eventually channeled into more formal capital markets opportunities on the local securities market, Albania will need to closely monitor the potential for rapid outflows. AZERBAIJAN: Market risk has been contained in recent years by direct foreign exchange controls. This was evident with the GKO crisis in the Russian Federation, with Azerbaijan faring better than most countries due to limits on exposures. Moreover, the central bank has been effective in stabilizing the macroeconomic framework, and preventing a recurrence of pyramid schemes. However, the economy will remain sensitive to oil/commodity price movements. Seventy percent of export revenue derives from oil, and about two thirds of foreign investment has been linked to the petroleum sector. KYRGYZ REPUBLIC: Macroeconomic vulnerability is a perennial problem. Notwithstanding efforts to implement relatively sound monetary policy to bring down inflation rates, these efforts have been undermined by comparatively loose fiscal policy, slow restructuring in the large-scale enterprise sector (still mainly state-owned), a weak banking system, and erosion of confidence in the local currency. Combined with general structural weaknesses, this has resulted in a steady increase in the Kyrgyz Republic's debt profile. Debt service obligations doubled in 2000 to 10.5 percent of exports, and future requirements will be more difficult to meet if there is deterioration in the balance of payments. Initial targets are for the current account deficit to decrease to about 7 percent of GDP by 2002, as compared with 16 percent in 2000. There is also a lack of strategic commodities to attract high levels of FDI, net of what has been attracted in the gold sector. The Kyrgyz Republic is also vulnerable to regional political risk. Major trading partners include Kazakhstan and Uzbekistan, with which KR has had disputes in the past over trade, payment, and netting arrangements. There is also the risk of civil disorder, as occurred in 1999. MOLDOVA: Since 1995, efforts have been made to stabilize the banking sector. Most efforts have involved reasonable forbearance to get potentially viable banks to a financial position where they could operate in the market without undermining stability. Major exposures to external markets are in the form of debt, most of which is held by the government rather than the banks. (GoM also has large internal exposure.) Moldovan banks fared reasonably well during the Russian crisis in 1998 because exposures were limited by prudential regulations. However, Moldova remains subject to market and political risk. Political risk abounds over the lack of definitive resolution of the Transnistria region. Any major conflict with the Russian Federation would harm export markets, undermine remittance flows, and complicate negotiations on payments for critical energy supplies. Meanwhile, increasing openness of the system will eventually increase vulnerability to movements and developments offshore. 72 CHAPTER THREE: PRIVATE SECTOR DEVELOPMENT AND PRIVATIZATION Since the start of the transition, all four countries have made impressive progress with some aspects of privatization. In particular, the privatization of small enterprises is largely complete in all four countries. Two of the four countries, Moldova and the Kyrgyz Republic, have made modest progress with the privatization of large enterprises. However, despite these positive steps, some problems remain. The privatization of 'strategic' Allfour countries enterprises, including the privatization of infrastructure, has been have made slow. In several cases, privatizations have been delayed for political reasons and/or by overly optimistic valuations out of line impressive with market standards. Further, the privatization programs progress with the employed in the four countries have often resulted in dispersed privatization of and insider ownership, with a negative impact on enterprise performance and corporate govemance. Finally, weak legal small enterprises protection for minority owners makes it easier for controlling owners to divert revenues from minority owners, and discourages outside investors from taking minority stakes. Box 3.1 CHALLENGES IN PRIVATIZATION Areas Challenges Large-scale . Progress has slowed for the largest loss-makers Privatization . Governments place excessive emphasis on cash receipts * Enterprises are given unrealistic valuations Infrastructure . Progress is slow, as values deteriorate Privatization . Regulation needs greater market orientation * The power sector is excessively integrated Corporate . Disbursed and insider ownership has impaired the quality of governance Governance . Legal protection for minority owners is not well defined or enforced 73 PROGRESS WITH PRIVATIZATION Small-scale Privatization Small-scale privatization is largely complete in Albania and the Kyrgyz Republic, and significant progress has been made in the other two subject countries studied. In Albania and the Kyrgyz Republic, small-scale privatization was largely complete by 1995. In fact, at this time, both subject countries scored slightly higher on the EBRD's index for small-scale privatization than the average for the EU accession candidates. Although both subject countries lagged behind the EU accession candidates by 1999, their progress with small-scale privatization is still impressive. Azerbaijan and Moldova made little progress with small-scale privatization in the early years of the transition. However, since 1993, they have accelerated and, by 1998, the EBRD (2000) rated their programs as nearly complete. Table 3.1 summarizes the EBRD index ratings for the four subject countries, the CIS and the EU accession candidates in both small-scale and large-scale privatization, for the beginning, middle and end of the 1990s. Table 3.1 EBRD Indexes for Small and Large Scale Privatization 1991 1995 1999 Small-scale Privatization Albania 2.0 4.0 4.0 Azerbaijan 1.0 1.0 3.3 Kyrgyz Republic 1.0 4.0 4.0 Moldova 1.0 3.0 3.3 Early Applicants to the EU 2.2 3.9 4.3 Cormmonwealth of Independent States 1.0 2.6 3.3 Large-Scale Privatization Albania 1.0 2.0 2.0 Azerbaijan 1.0 1.0 1.7 Kyrgyz Republic 1.0 3.0 3.0 Moldova 1.0 3.0 3.0 EU accession Candidates 1.4 3.5 3.7 Commonwealth of Independent States 1.0 2.1 2.5 Source: EBRD (2000) Notes: Higher values mean greater progress. For large-scale privatization, 1 means little private ownership; 2 means comprehensive scheme almost ready for implementation; 3 means 25 percent of assets privatized or in process; and 4 means more that 50 percent of large enterprise assets in private ownership. For small-scale privatization, 1 means little progress; 2 means substantial share privatized; 3 means nearly comprehensive program implemented; and 4 means complete privatization with tradable ownership rights. The highest possible score is 4.3, which means that conditions are similar to conditions in advanced industrialized economies. For the complete scoring system see EBRD (1997, p. 15). 74 Large-scale Privatization Although both Moldova and the Kyrgyz Republic have In contrast to the made some progress with large-scale privatization, progress is more limited in Albania and Azerbaijan, even progress made with compared to countries in the CIS. In contrast to the small-scale progress made with small-scale privatization, progress privatization, with large-scale privatization has been slow. Although the Kyrgyz Republic and Moldova had made more progress progress with large- than the other two subject countries by 1999, they still scale privatization lagged behind the leading economies in Central Europe has been slow and the Baltic states. Privatization of Infrastructure Other than in the telecommunications sector in Albania and the power sector in Moldova, the four subject countries have made only limited progress with reform of and private participation in, infrastructure. By the beginning of 2001, none of the four subject countries had fully privatized their fixed-line operators, and there was little competition in cellular telephony. This is very much at odds with developments in the EU accession candidates, where at a minimum, private competition in the mobile phone market has existed for several years and at least partial privatization of fixed line operators has proceeded or been transacted. Although all four subject countries had taken steps to start privatizing their fixed line operators, these attempts have not yet been successful, and there have been numerous cancellations and delays. Other than in Moldova, where electricity distribution has been partially privatized and distribution and generation have been separated, the power sectors of the four subject countries were generally dominated by vertically integrated state-owned enterprises at the end of 1999 (see Table 3.2). This is also at odds with some of the EU accession candidates, which have shown a willingness to move forward with privatization in electricity distribution. Progress in other areas of infrastructure (i.e., railways, roads, and water) has also been slow29, once again, deviating from the practices of the EU accession candidates that have introduced private toll roads, portions of railway operations, and municipal water services. 29 EBRD, 2000. 75 Table 3.2 Private Sector Participation in Infrastructure in 2000 Albania Azerbaijan Kyrgyz I I I Republic Moldova EBRD Index for Telecom Reform 3.3 1.3 2.3 2.3 EBRD Index for Power Reform 2.3 2 2.3 3.3 Independent Telecoms Regulator Yes No Yes No Independent Electricity Regulator Yes No Yes Yes Privatization of Fixed Line Operator* No** No No** No** Number of Cellular Operators 1 *** 2 2 2 # of private cellular operators 1 2 2 2 Notes: Source for EBRD indices and independence of regulators in EBRD (2000). Higher values for the indices imply greater reform In the power sector, a 1 means sector operates as government department. 2 means some distance from the government and improvements in corporate governance but no separation of businesses. 3 means a law has been passed for complete restructuring including vertical unbundling and the setting up of a regular. 4 means the law has been implemented. In the telecommunications sector, a 1 means little progress with commercialization, private sector participation and regulation. 2 means modest progress with commercialization of main operator. 3 means substantial commercialization, separation of postal and telecom services and liberalization in the mobile market. 4 means privatization of the dominant operator, comprehensive regulator and institutional reform and liberalization of entry. Maximum score is 4.3. For complete descriptions see EBRD (2000, p. 43). * Privatization means private operator has controlling stake in enterprise. **As of the beginning of 2001, Albania, Moldova and the Kyrgyz Republic were in the process of privatizing their fixed line operators, although the process has been delayed several times in all these countries ***A second cellular license was auctioned in January 2001 in Albania. APPROACHES TO PRIVATIZATION Although all four subject countries have used multiple methods to privatize both small and large enterprises, the primary methods used for small-scale privatization favored insiders (i.e., workers and managers). Most of the subject countries relied on cash sales or auctions, although Moldova included 560 small enterprises in its voucher privatization program and privatized a number of small-scale enterprises through lease buyouts, while Albania used management and employee buyouts. All four of the subject countries used voucher privatization schemes as their main strategy for the privatization of medium- and large-scale enterprises, in many cases with significant preferences for insiders. Only in Albania were the preferences for insiders minimized, with all citizens having equal access to vouchers and shares. Lease buy-outs and, for the most attractive companies, case-by-case sales and international tenders were also used, albeit not very often. Although investment funds were permitted in all of the subject countries, they generally did not attract a significant share of vouchers from the population. Therefore, they were able to acquire no more than minority equity interests, which gave them little impact on corporate governance and highly uncertain dividend earnings. Table 3.3 summarizes the main features of the schemes used in the four subject countries for small-scale and medium- and large-scale privatization. 76 Box 3.2 Do PRIVATIZATION AND OTHER REFORMS CAUSE POVERTY? One concern often expressed about the structural reform process is that, although reform may improve overall economic performance, the gains might be spread unequally throughout the population. In this view, market reforms might increase inequality and, perhaps, poverty. For example, privatized enterprises, facing hard budget constraints, might be forced to restructure, reduce the size of their workforces, or close down altogether. In the absence of well-functioning labor and property markets and effective social safety nets, this could plunge many families into poverty. However, a recent study, World Bank (2000, p. 113) finds that the opposite is true: "Countries that implemented inconsistent or partial reforms experienced larger drops in output than those that chose to move quickly. This relationship is reflected in the association between policies and poverty: good policies that allowed for a faster recovery of output appear to have led to lower poverty levels. Poverty is highest where reforms have stalled or have been only partially implemented... This evidence suggests that countries that moved quickly and comprehensively to implement market reforms experienced a smaller increase in poverty than countries that delayed the start of reform." Lower poverty among faster reformers does not appear to be only due to faster economic growth. World Bank (2000, p 156) shows that, in general, income inequality is lower in those countries that have privatized more. The chart below shows the correlation between enterprise reform and changes in inequality. 0.35 0.3 ... ...............-------------------------------.....--------- AD 0.25 . - - . . . - - - - - - - - - - -... ........................ 8 0.2 .- ----- ---- ---- 2 --- - - - - -- - - - - -- -- - - - - -- - - ------ - 0.15 -- .. ' 0.1 -- a,~~~~~~~~~ -0.05 ._ ...I . 1.5 2 2.5 3 3.5 4 (EBRD enterprise reform) Data: World Bank (2000) and EBRD (2000). Note: EBRD enterprise reform is of average of EBRD rankings for large-scale privatization, small-scale privatization, enterprise reform and competition policy. Dethier et al (1999) shows a similar correlation for the EBRD index of privatization programs and levels of inequality. In general, the high correlation between different measures of reform makes it difficult to distinguish which reforms are most important. 77 The privatization schemes used in the four subject countries Privatization has resulted in relatively dispersed ownership, often with employees resulted in holding significant stakes in their companies. Although some relatively dispersed analysts have suggested that employee ownership can improve enterprise performance by aligning the interests of employees ownership, often and shareholders, employee ownership has not been effective in with employees many transition economies. A recent survey of the literature holding significant suggests that employee-owned enterprises have performed worse than private enterprises with concentrated ownership.30 The stakes ... employee perforrnance of employee-owned enterprises was especially poor ownership has not in the CIS, possibly because ownership incentives do not been effective function effectively in weak institutional environments, such as in the CIS and the four subject countries. CHARACTERISTICS OF STATE-OWNED AND PRIVATE ENTERPRISES In addition to macro-level data, this section of the report uses data from the World Business Environment Survey (WBES), a major survey of over 10,000 enterprises in 80 countries conducted by the World Bank and several other agencies. The survey of the transition economies, which was conducted in collaboration with the European Bank for Reconstruction and Development, included over 3,000 enterprises from 21 transition economies. Box 3.3 contains a more complete description of the WBES. In the four subject countries, privatized enterprises are smaller than state-owned enterprises but larger than de novo enterprises. This is consistent with previous evidence that more progress has been made with small-scale than large-scale privatization. A similar pattern can be observed in the EU accession candidates and other CIS economies. In three of the subject countries, privatized enterprises were more likely to be found in industry than either de novo or state-owned companies. In the fourth, Azerbaijan, state-owned enterprises were most likely to be in industry, followed by privatized companies, then de novo enterprises. This result is consistent with the EBRD rating of Azerbaijan's large-scale privatization effort, which trails the other three countries by a substantial margin. The size and sectoral distribution of enterprises differ between countries, as can be seen from Table 3.4. Enterprises in Albania and Azerbaijan tend to have fewer employees than enterprises in Moldova and the Kyrgyz Republic. De novo enterprises (i.e., enterprises that were formed as private enterprises rather than being privatized) were smaller and more likely to be in service sectors than state-owned or privatized enterprises. Even relative to privatized enterprises, de novo firms tend to be small. In all the 30 Djankov and Murrell found that, in general, ownership by foreign enterprises and individuals, ownership by investment funds, and concentrated individual ownership was more effective than employee- ownership at improving sector performance. 78 subject countries, in the EU accession candidates and in the CIS as a whole, more than 50 percent of de novo firms had 50 employees or fewer. De novo firms were also more likely to be in service sectors in all of the subject countries and in the CIS as a whole, although the difference between de novo firms and other enterprises was small in Azerbaijan. 79 Table 3.3 Approaches to Privatization Albania Azerbaijan Kyreuz Moldova Republic Small-scale privatization Year Initiated 1991 19962 1992 1993b Management and employee Cash auctions Sales to individuals Voucher privatization Main Method byuswith concessions and workerprga for insiders. collectives Conversion to joint- Other Methods Cash auctions, voucher stock companies Direct cash sales, Used privatization program. with some shares leases. reserved for insiders Medium- and Large-scale privatization Year Initiated 1995 1997 1992 1993b Main Method Voucher privatization pivattion Voucher Voucher pnvatization program ~~~programn programprga Auctions, leases, transfers Cash auctions Conversion to joint- Direct sales, leases, Other Methods to local governments, and asbyas stock comnpanies auctions, lendes Used direct sales for 'strategic' and case-by-case with some shares auctions, int'l tenders enterprises. competitive sales reserved for insiders for best enterprises. Voucher Privatization Year Initiated 1995 1997 1994c 1993 In most cases, E some shares Five percent of mployees allowed to purchase up to 20% of Preferences for Equal access voucher were reserved for stes were shares for nominal price; Insiders privatization program enployees on workers at no 50% of shares of agro- concessional cost processors given to terms collective farms. Permiitted, but few set up; banks not allowed to set up Permsitted, but Permitted; about 40 Voucher funds; investment funds few vouchers closed-end funds Privatization prohibited from owning Permitted placed in emerged, but these have Funds >40% of companies (or investment had little imnpact on investing >1O% of their funds. corporate governance assets in single company Sources: EBRD (1996, 2000), World Bank (2000) and McHale and Pankov (2000). Information from these sources was supplemented with information from other sources. Notes: All subject countries used multiple privatization methods, with many countries using multiple methods for individual enterprises. a Although the law on small-scale privatization was adopted in January 1993, the first privatization took place in 1996 and there was little progress before 1997. b Privatization Law was passed in July 1991, but little progress was made outside of the agricultural sector before 1993. 'An earlier voucher privatization program was initiated 1992, but little progress was made. 80 Box 3.3: WORLD BUSINESS ENVIRONMENT SURVEY This section of the study draws extensively upon data from the World Business Environment Survey (WBES). Since the survey was completed in 1999, all parts of this study that refer to these data refer to data for 1999, unless otherwise noted. Consequently, the data from the survey will not take into account change in policy since 1999 that might have improved the business environment in the case study countries. Similar sampling and survey techniques were used in all countries to ensure cross-comparability of data and to avoid systemic bias for individual countries. As discussed in Hellman et al. (2000), which describes the survey and the sampling methods in detail, the WBES was designed to gather data from a representative sample of enterprises. Based on information from government statistical offices in each of the countries, quotas for a breakdown based on sector, number of employees and location were constructed. Hellman et al. (2000) note that some judgment was necessary since the quality and quantity of information varied considerably across countries. Several additional quotas were applied to ensure that different groups of enterprises were represented in the sample. These additional restrictions were: (i) manufacturing and service enterprises were represented according to their relative contributions to GDP, with a 15 percent minimum imposed on each sector (ii) at least 15 percent of the sample was made up of small (less than 50 employees) and large (over 200 employees) enterprises; (iii) at least 15 percent of the sample came from small towns or rural areas; (iv) at least 15 percent of firms had some foreign ownership and (v) at least 15 percent of enterprises were required to export at least 20 percent of their output. In addition to these quotas, the sample was selected so that 20 percent of enterprises were state- owned, with the remainder being privately owned. However, within these ownership classes, no attempt was made to ensure that the samples were representative. Hellman et al (2000) note that they did not attempt to construct weights and that the information needed to do this was not available in most countries. Since the samples for the subject countries had very few majority foreign-owned firms (see below) and only 20 percent of firms were state-owned in accordance with the sample selection rule, most of the enterprises were domestically owned private enterprises. Several additional steps were taken to ensure that the data would be comparable across countries. First, the survey instrument, which was prepared in English, was translated into local languages and then independently translated back into English to ensure accuracy of translation. Second, the survey was extensively piloted in all countries to ensure that enterprise managers correctly understood questions. Third, the interviews were conducted face-to-face to avoid problems related to low response rates from postal surveys and the difficulty of dealing with sensitive issues in telephone surveys. Fourth, the same international survey firm conducted the survey using local staff trained according to a common format. Despite efforts to ensure that the data would be comparable across countries, it is still possible that some bias remains. For example, in part of the survey, managers were asked to assess how serious several obstacles were to the operation and growth of their businesses on a scale of one to four. These questions could be affected by perception bias if respondents rate problems more or less severely than an objective outsider would. Further, if cultural differences mean than individuals' biases are correlated within countries, this could result in some bias at the country level. A recent study, however, found that there is no strong evidence of persistent bias in the WBES. Hellman et al. (2000) compared several of the answers on the WBES with objective measures of country performance and found no evidence of systematic bias for any of the countries in the W3BES samples for Eastern Europe and Central Asia except for Armenia. * Although some of the authors of this study are employed by the World Bank or the EBRD, the study notes that the views expressed in findings, interpretations, and conclusions in that paper reflect only the views of the authors and do not necessarily reflect the views of the World Bank (or EBRD), their member countries, or the Executive Directors of those organizations. 81 Table 3.4 Sector and size distribution of state-owned, privatized and de novo enterprises in 1999 SmaHl Medium-sized Large Industry Services Other Total 63% 28% 9% 38% 50% 13% AlbaniaI Privatized 38% 52% 10% 69% 27% 4%: Total 61% 32% 7% 52% 48% 1% Azerbaijan Privatized 53% 41% 6% 53% 47% 0% Total 29% 45% 26% 74% 25% 2% Kyrgyz Republic Privatized 17% 45% 38% 87% 13% 0% Total 32% 30% 38% 65% 34% 1% Moldova I I Prnvatied 16% 32% 52% 82% 18% 0% Total 49% 27% 24% 43% 52% 4% EU accession I candidates Privatized 21% 40% 39% 66% 31% 3% Commonwealth Total 41% 35% 24% 59% 40% 1% ofn-pedet of Independent Privatized 20% 39% 40% 73% 27% 0% Data Source: WBES Notes: Columns might not add to 100 percent due to rounding. Small enterprises are enterprises with less than 50 employees, medium-sized enterprises have between 50 and 200 employees and large enterprises have over 200 employees. ENTERPRISE REFORM Box 3.4 CHALLENGES IN ENTERPRISE REFORM Areas Challenges Competition * Both competition and competition policy remain weak Poicy Enterprise * Incentives to restructure privatized enterprises remain weak Restructuring 82 Competition and Competition Policy Competition policy remains weak in the subject countries, with all four significantly lagging behind the EU accession Competition candidates. Although the four subject countries made policy remains significant progress with small-scale privatization by the end of weak the 1990s, less progress was made to increase competition and improve competition policy (see Table 3.5). EBRD (2000) gave all four countries low scores in this area. Further, and in contrast to the EU accession candidates, only Albania has made any progress in this regard since the middle of the 1990s. Table 3.5 EBRD Index of Competition Policy. 1991 1995 1999 Albania 1.0 1.0 1.7 Azerbaijan 1.0 2.0 2.0 Kyrgyz Republic 1.0 2.0 2.0 Moldova 1.0 2.0 2.0 EU accession candidates (Average) 1.6 2.5 2.8 CIS (Average) 1.1 1.7 1.9 Source: EBRD (2000). Higher values for the indices imply greater reform 1 implies no competition legislation or institutions. 2 implies competition policy legislation and institutions set up; some reduction of entry restrictions of enforcement action on dominant firms. 3 implies some enforcement actions to reduce abuse of market power and to promote a competitive environment. 4 implies significant enforcement actions to reduce market power and promote a competitive environment. Maximum score is 4+. In all the subject countries other than Azerbaijan, fewer enterprises said that their customers would switch suppliers if they raised prices by 10 percent than in the EU accession candidates (see Figure 3.1). This provides further evidence that competition is weak in the four subject countries. 83 Enterprise Restructuring In contrast to their significant progress Figure 3.1 Enterprises in the Four Subject Countries with privatization, none of the four were Less Likely to Fear Losing Customers to Their subject countries has made Competitors if They Raised Prices by 10 Percent comparable progress with enterprise reform since the start of the transition. 50% EBRD (2000) gave Albania, the e 45% Kyrgyz Republic and Moldova a score *N 40% - of 2.0 for enterprise reform and a) 35% - Azerbaijan a 1.7, with scores not X 30% . I I . | improving in any of the four subject < 25% countries since 1995 (Table 3.6). This E 20% - _ _ * l 1 indicates that incentives for E) 15% S l - lll l l l restructuring privatized enterprises are 6 10% particularly weak in the four subject 8 5% countries, that creditors have a a. 0% L difficult time enforcing their claims MGc an N n > D c > against enterprises, and that n 2) 3C bankruptcy legislation is only weakly w < enforced. Although the four subject countries score below the EU Data Source: WBES.aceso cadat, th erl Note: Results for medium and large enterprises with over 50 employees n accession candidates, the early 1999 applicants have also made little progress in this area since the mid- 1990s. Table 3.6 EBRD Index of Enterprise Reform 1991 1995 1999 Albania 1.0 2.0 2.0 Azerbaijan 1.0 1.7 1.7 Kyrgyz Republic 1.0 2.0 2.0 Moldova 1.0 2.0 2.0 EU accession candidates (Average) 1.6 2.9 3.0 CIS (Average) 1.0 1.7 1.8 Source: EBRD (2000). Higher values for the indices imply greater reform. A score of I means soft budget constraints and few reforms to promote corporate governance; 2 means moderately tight credit and subsidy policy, but weak enforcement of bankruptcy legislation and little action to strengthen competition and corporate governance; 3 means significant progress with hardening budget constraints and to promote effective corporate governance; and 4 means substantial improvement in corporate governance and significant new investment. Maximum score is 4+. Most enterprises Although enterprise reformn has been disappointing, most enterprises have taken steps to had taken at least some steps towards restructuring their operations their between 1996 and 1999 in all the subject countries. The WBES restructure tnelr asked about 15 areas of restructuring, including opening and closing operations plants, increasing and decreasing employment by at least 10 84 percent, developing new products, and exporting to new countries. In all countries, more than half the enterprises undertook at least one of the fifteen actions between 1996 and 1999 (see Table 3.7). Restructuring was especially common in Moldova, where almost all enterprises took at least one action towards restructuring. The low level of restructuring among de novo enterprises is probably because most did not need the same level of restructuring as state-owned and privatized enterprises. Table 3.7 Enterprise restructuring in the four subject countries between 1996 and 1999 Increase labor force Decrease labor force by No Restructuring by at least 10% at least 10% State-owned 21.9% 12.5% 25.8% Albania De novo 45.3% 22.7% 6.7% Privatized 19.2% 15.4% 30.8% State-owned 33.3% 4.2% 25.0% Azerbaia De novo 28.8% 7.6% 11.5% Azerbaijan Privatized 17.6% 5.8% 23.5% State-owned 25.0% 10.7% 14.3% Kyrgyz De novo 7.7% 11.5% 11.5% Republic Privatized 12.0% 4.0% 44.0% State-owned 0.0% 15.4% 53.8% Moldova De novo 0.0% 23.8% 28.6% Privatized 2.9% 14.3% 60.0% EU accsState-owned 10.1% 12.7% 52.5% EU accession De novo 14.0% 38.6% 19.2% candidates Privatized 6.0% 17.1% 50.4% Commonwealth State-owned 7.9% 15.8% 44.8% of Independent De novo 5.2% 24.9% 33.6% States Privatized 6.7% 48.8% 15.8% Data Source: WBES Notes: De novo enterprises exclude enterprises that did not exist before 1996. De novo and Privatized refer to outsider-owned private enterprises. No restructuring implies that the enterprise did not restructure in any of the 15 areas that the WBES asked about. The most dynamic enterprises in terms of employment growth are de novo enterprises. De novo enterprises are more likely to increase employment and less likely to decrease employment than either state-owned or privatized enterprises in all the subject countries. Although the low number of layoffs among de novo enterprises is not surprising, since they are probably less likely to be overstaffed, it is The most dynamic encouraging that they are also more likely to expand enterprises in terms employment than either state-owned or private enterprises. of employment This emphasizes that de novo private enterprises are likely to be one of the main engines for economic and employment growth are de novo growth in these countries. enterprises 85 THE BusINEss ENVIRONMENT AND CONSTRAINTS ON ENTERPRISE OPERATIONS Box 3.5 CHALLENGES IN THE BUSINESS ENVIRONMENT Areas Challenges Macroeconomic * Even moderate levels of inflation (between 20 and 40 percent) Stability are a constraint on enterprise performance and growth Taxation and * High tax rates are the largest constraint on enterprise Regulation perforrnance and growth in the area of taxation and regulation * Tax evasion and arrears are major problems and result in high tax burdens for enterprises that pay taxes * Excessive regulation encourages corruption Crime and * Crime is a significant problem for enterprises, especially in Corruption Albania * Corruption is high, harming domestic enterprises and .deterring foreign investnent * Corruption often affects small enterprises more than medium- sized enterprises * Successful enterprises are especially harmed by corruption Access to Finance * High interest rates and poor access to long term loans are the most significant problems related to access to finance * Access to finance is a greater constraint for small enterprises than for medium or large enterprises The Legal System * Existing commercial laws are poorly enforced and the legal infrastructure is weak Macroeconomic Stability Although inflation was very high in the early 1990s in all the subject countries, it had fallen below 25 percent by 1996. At the start of the transition period, the collapse in economic output was accompanied by a rapid acceleration in inflation. In all three subject countries in the CIS, inflation reached hyper-inflationary levels in the early 1990s (see Table 3.8). However, inflation was reduced quickly, as the countries replaced the ruble with their own currencies and tightened fiscal and monetary policies. By 1996, inflation had been brought down to 21 percent in Moldova, 32 percent in the Kyrgyz Republic, and 20 percent in Azerbaijan. 86 Table 3.8 Inflation Rates (1991-1999) (in percent) 1991 1992 1993 1994 1995 1996 1997 1998 1999 Average 96-99 Albania --- 226 85 23 8 12 33 20 0.39 16 Azerbaijan --- 912 1,129 1,665 412 20 3.6 (0.7) (8.6) 3.5 Kyrgyz Republic --- 855 772 229 41 32 23 10 36 25 Moldova --- 1,276 789 330 12 21 8.0 6.6 46 20 EU accession 89 262 40 26 21 17 12 11 5.9 7.7 candidates CIS --- 1,055 2,008 1,337 348 151 33 19 49 63 OECD 5.1 3.8 3.4 2.7 2.9 2.2 1.8 1.6 1.5 1.8 Source: IMF International Financial Statistics By 1999, tight monetary policies and successful stabilization programs resulted in inflation close to zero percent in both Albania and Azerbaijan. Although the dispute over the Nagorny- Karabakh region delayed stabilization and reform in Azerbaijan, the government enacted a stabilization program after the 1994 cease-fire, which succeeded in reducing inflation below 20 percent by 1996. Since 1996, the government has continued to enforce tight monetary policies, and inflation became negative in 1999. In Albania, inflation rates rose to about 226 percent in 1992, before being reduced to less than 10 percent by 1995. Although inflation accelerated following the collapse of the pyramid schemes, the government quickly re-stabilized the economy, and inflation was only 0.4 percent in 1999. In Albania and Azerbaijan, as in the EU accession candidates, enterprises saw inflation as a minor to moderate obstacle to enterprise operations and growth in the WBES. Since inflation was close to zero or negative in both of these countries, and close to or below 10 percent in all the EU accession candidates, it is not surprising that inflation was seen as a relatively minor problem. However, this does not imply that macroeconomic stability should be taken for granted - the importance attached to macroeconomic stability in the Kyrgyz Republic and Moldova stresses the importance of continued vigilance to achieve stability. In contrast, although inflation quickly fell from the hyper-inflationary levels of the early 1990s in the Kyrgyz Republic and Moldova, it remained at moderate levels through the end of the decade. Inflation averaged over 20 percent per year between 1996 and 1999 in both countries, and reached about 40 percent by 1999, with predictable destabilizing impact on the exchange rates for local currencies. Although these moderate levels of inflation do not indicate severe macroeconomic imbalances, enterprise managers saw even this modest level of inflation as the largest constraint on enterprise operations and growth in these countries (see Table 3.9). The exchange rate, which was ranked as a moderate to major constraint in both countries, was seen as the third largest obstacle. The importance attached to inflation and exchange rates in these countries with only moderate levels of inflation - in the face of high levels of corruption, poorly developed domestic financial sectors, continued political instability, and weak legal institutions -attests to the high value placed on macroeconomic stability by the real sector. 87 Table 3.9 Macroeconomic Obstacles to Enterprise Performance in 1999 Inflation Rate Albania Mean Rating 2.7 2.5 Albania (Position) (9) (12) Mean Rating 2.6 2.3 Azerbaijan (Position) (5) (9) Kyrgyz Republic Mean Rating 3.7 3.5 Kyrgyz ~~~~~~~~(Position) (1) (3) Moldova Mean Rating 3.8 3.5 (Position) (1) (3) EU accession candidates (PeanRating (245) 2 Commonwealth of Independent Mean Rating 3.4 3.0 States (Position) (1) (5) Data Source: WBES Notes: High scores mean greater constraint. Table 3.10 Inflation as an Obstacle to Enterprise Performance in 1999 No. of Minor Moderate Major Countries Constraint Constraint Constraint Very Low Inflation (< 5%) 1 1 100% 0% 0% Low Inflation (between 5 % and 15 67% 33% 0% 10%) Moderate Inflation (between 10% 24 14% 86% 0% and 40%) High Inflation (between 40% and 28 0% 42% 58% 100%) Very High Inflation (> 100%) 6 0% 50% 50% Data Source: WBES Notes: Moldova and the Kyrgyz Republic have been excluded from this table. Countries are classified by inflation based on the highest year-end inflation between 1995 and 1999. Recent studies suggest that inflation does not deter economic growth significantly until it becomes as high as 40 percent per year. It thus might seem puzzling that enterprises rated moderate levels of inflation in these countries as a major constraint on enterprise operations and growth.31 However, enterprises in other countries also have similar reservations about even moderate levels of inflation. Enterprises rated inflation as a major constraint on enterprise operations and growth in over 50 percent of the countries where year-end inflation exceeded 40 percent in at least one year between 1995 and 1999 (see Table 3.10). Even in countries where the highest year-end inflation was between 10 and 40 percent between 1995 and 1999, 31 See World Bank (1997) and Bruno and Easterly (1998). Bruno and Easterly (1998) find that there is little evidence to support the assertion that moderate inflation (between 20 and 40 percent) affects growth. However, they also note that, since there were very few cases of inflation between 40 and 80 percent in their sample, it is difficult to tell whether the threshold is even higher than 40 percent - perhaps as high as 80 percent. 88 enterprises in 86 percent of the countries rated inflation as a moderate constraint on enterprise operations and growth.3 Regulation and Taxation Taxation and regulation were seen as significant obstacles to Taxation and enterprise performance in the four subject countries in 1999. regulation are Enterprises ranked this grouped constraint between first and fifth seen as significant relative to the eleven main constraints (see Table 3.11). One notable difference between regulation and taxation and the other obstacles to constraints is that enterprises in the four subject countries do not see enterprise taxation and regulation as a greater constraint than enterprises in the EU accession candidates.33 However, given the high level of performance corruption in the subject countries, and the possibility that corruption might be caused by excessive or over-zealous regulation, regulation and taxation might be indirectly more burdensome in the four subject countries than it is in the EU accession candidates.34 Table 3.11 Taxation and Regulation as Obstacles to Enterprise Performance in 1999 Overall High Tax Tax Administration Business Rates and Regulations Licensing Albania Mean Rating 3.0 3.7 3.1 2.1 (Position) (5) (1) (2) (5) Azerbaijan Mean Rating 3.0 3.4 3.0 1.9 (Position) (1) (1) (2) (3) Kyrgy Repblic Mean Rating 3.5 3.7 3.2 2.1 Kyrgyz Republic (Position) (2) (1) (2) (3) Mean Rating 3.5 3.5 3.1 2.2 Moldova (Position) (4) (1) (2) (3) EU accession Mean Rating 3.0 3.5 3.0 2.0 candidates (Position) (1) (1) (2) (3) Commonwealth of Mean Rating 3.4 3.5 3.0 1.9 Independent States (Position) (2) (1) (2) (4) Data Source: WBES Note: High scores mean greater constraint 32 It is not clear whether enterprise mnanagers object to moderate levels of inflation or whether they fear a return to very high levels of inflation. Since the transition economnies often went through periods of hyperinflation in the early years of the transition, it is possible that enterprise managers in countries where inflation remains moderate fear the possible return of hyperinflation. 33 Out of the other ten constraints, the average rating for enterprises in the subject countries exceeded the average rating in the EU accession candidates in 39 of 40 comparisons. 34 See, for example, Bardhan (1997, p. 1335) for a discussion of the link between corruption and regulation. 89 Within the broad areas of regulation and taxation, enterprises saw high tax rates as the Figure 3.2 Government Revenues are Low largest obstacle in all four subject countries, as in the Subject Countries in 1999. well as in the CIS and in the EU accession candidates. Although this might suggest that a- 50 - there would be benefits to across-the-board o 40 - reductions in tax rates, it is important to note the S 35 - relatively weak fiscal situation in the four X 2- subject countries largely precludes this. The 25 , 20 - * r* four countries all have relatively high fiscal ? 15 - deficits, and general gover-nent revenues were 2-1> 5- _ ON U - - lower there in 1999 than they are in the EU 0 o - * * to a, to 'O * * accession candidates (see Figure 3.2). Although 'C . D t° D &low tax revenues might seem inconsistent with D we co ° Q t C st O high tax rates, the high levels of arrears and low a' Olevels of revenue reported to tax authorities (see Chapter 4) partially explains this discrepancy. As a result, improving the tax system by Data Source: EBRD (2000) reducing tax evasion and bringing the informal sector into the tax system, while relying less on corporate taxes, could reduce the tax burden on those enterprises that do pay taxes. Azerbaijan recently adopted a several provisions to improve tax collection from the non-oil sector, including adopting a new tax code. In response, tax collections from the non-oil sector improved in 2000, with revenues from the profit tax increasing from 207 billion manats in 1999 to 375 billion in 2000 and revenues from the VAT increasing from 621 billion manats to 692 billion. Other than tax administration and regulations, enterprises were not very concerned about other areas of regulation, viewing them as only minor obstacles to enterprise performance. In addition to being asked about tax rates and tax administration, the WBES also asked about business licensing, fire and safety regulations, environmental regulations, exchange rate and foreign currency regulations, labor regulations, and foreign trade regulations. In all four subject countries, and for all forms of regulation, enterprises saw these as only minor obstacles. Although this might suggest that regulation in general is not very important, it should be noted that excessive regulation is often linked to corruption, which was a more serious problem in several of the four subject countries (e.g., in Albania and Azerbaijan). Moreover, even if managers of domestically owned enterprises are used to dealing with regulation in their home countries, excessive regulation might deter foreign direct investment in the transition economies (see Sherif et al, 2000). Political Instability, Crime and Corruption Crime, political instability and corruption were the most important obstacles to enterprise performance in Albania. The collapse of the pyramid schemes in 1997, which resulted in 90 significant civil unrest, led to high levels of political instability in Albania in the late 1990s.35 This high level of political instability has made government policy less predictable, which has had a negative impact on the business environment. The collapse of the pyramid schemes also worsened crime. During the rioting that followed the collapse, thousands of weapons were looted from army barracks. Although many weapons found their way into Kosovo, many others ended up in the hands of local criminals. While order has been slowly restored since 1999, crime and civil unrest remain significant problems. In contrast to Albania, in none of the three subject countries from the CIS, was street crime or organized crime seen as a significant constraint on enterprise performance. Enterprises in Moldova ranked political instability as the second largest obstacle to enterprise operations and growth. In part, this may reflect the failure of any party to win a parliamentary majority in the 1998 elections, and ongoing conflict between the President and Parliament.36 Table 3.12 Crime, Corruption and Political Instability as Obstacles to Enterprise Performance in 1999 Political Street crime, theft Organized Corruption Instability and disorder crime Albania Mean Rating 3.4 3.4 3.2 3.3 (Position) (2) (1) (4) (3) Azerbaija Mean Rating 2.4 2.2 2.4 2.8 Azerbaijan (Position) (7) (8) (11) (3) Rebli Mean Rating 3.4 3.4 3.1 3.3 Kyrgyz Repuillc (Position) (5) (6) (8) (7) Mean Rating 3.6 3.1 3.1 2.9 Moldova (Position) (2) (7) (6) (9) EU accession Mean Rating 2.6 2.1 1.7 2.0 candidates (Position) (2) (7) (11) (9) Commonwealth Mean Rating 3.4 2.5 2.4 2.5 of Independent (Position) (2) (8) (9) (6) States Data Source: WBES Note: High scores mean greater constraint. 35 Following the collapse, the government resigned and parliamentary elections were held. This led to defeat for the incumbent Democratic Party and a new coalition led by the Socialist Party of Albania. However, the Democratic Party disputed the election results and, subsequently, decided to boycott Parliament, not returning until mid-1 999. In addition to the dispute with the Democratic Party, the government coalition has proved to be unstable, with three different Prime Ministers in place between 1997 and 1999. 36 Although the Communist Party of Moldova won the largest number of seats in the 1998 elections, it failed to win a majority, and the three center-right parties that forned the government have failed to work together, resulting in several changes in government since this time. 91 Enterprises in the four subject countries saw Enterprises saw corruption in 1999 as a larger problem than corruption as a larger enterprises in the EU accession candidates. This is consistent with other contemporary evidence on problem in 1999 than corruption. For example, based on surveys of enterprises did in the EU corruption throughout the world, Transparency accession candidates International rated corruption in all four of the subject countries as worse than in any of the EU countries accession candidates in its 2000 survey. Similarly, the two countries included in the 2001 survey also ranked below the EU accession candidates. Moreover, enterprises in the four subject countries reported paying a higher share of their revenues in bribes in 1999 than enterprises in the EU accession candidates, further emphasizing that corruption is a more significant problem in the four subject countries (see Table 3.13).37 Since the WBES was conducted in 1999, and even Transparency International's most recent ratings reflect data collected between 1999 and 2001, perceptions about corruption in the survey countries might not reflect recent efforts to reduce corruption. Further, since changes in policy probably affect perceptions only in the medium-term, recent actions might be immediately reflected in changes in perceptions. For example, the Government of Azerbaijan has taken several steps to reduce corruption and improve the function of civil society since 1999. The steps that it has recently taken to reduce corruption and increase accountability include: (i) adopting a new law on public service; (ii) enacting new procedures to increase the public accountability of civil servants; (iii) reviewing the legal and institutional framework for licensing and the adequacy of provisions for transparency in administrative and grievance procedures; (iv) putting in place new administrative arrangements and working on revising regulations for the Oil fund; (v) working to finalize a government anti-corruption program; and (vi) adopting a new law on public procurement. Further, the Govemment has also taken steps to increasing the role of non-governmental organizations, including: (i) passing a new Law on NGOs and working to adopt a new Law on the Registration of NGOs; (ii) ensuring that NGOs are better represented in government decisions by establishing state commissions/working groups in several areas and (iii) adopting new regulations to ensure that newly established ministries (including the recently established Ministry of Economic Development take decisions in consultation with NGOs. 37 To encourage honest responses to questions about bribery, and to allow enterprise managers to avoid implicating themselves when answering questions about frequency and level of bribe payments, the WBES asked about bribes paid by 'firms like yours' rather than about the manager's own fimn. Consequently, enterprise manager's responses did no imnply that their enterprises actually paid any bribes at all. Although it is important to keep this in mind when looking at the following results, the study refers to these questions as if the managers were talking about their own enterprise. 92 Table 3.13 Bribes as Share of Enterprise Revenues in 1999 Kyrgyz EU Albania Azerbaijan Republic Moldova accession CIS candidates Average 2.72 3.25 3.24 2.60 1.98 2.56 Less than 10 3.3 3.3 3.3 4.2 2.3 3.1 employees 10 to 49 employees 3.0 3.4 3.4 3.5 2.2 2.9 50 to 99 employees 2.2 2.7 3.0 1.8 2.0 2.5 100 to 199 employees 2.0 3.4 3.3 1.9 1.8 2.2 Over 200 employees 1.3 4.0 3.4 2.2 1.5 2.1 Data Source: WBES Note: Response was with reference to a hypothetical firm similar to the manager's own enterprise. Question was "On average, what percent of revenues do firms like yours typically pay per annum in unofficial payments to public officials. Categories are 1 - 0%; 2 - less than 1 percent' 3 - between 1 and 1.99%; 4 - between 2 and 9.99%; 5 - between 10 and 12%; 6 - between 13 and 25%; 7 - over 25%. Although enterprises in Azerbaijan saw corruption as a lesser constraint than enterprises in the other subject countries, they reported paying a greater share of revenues in bribes than enterprises in either Albania or Moldova. Although this might seem contradictory, factors other than the level of bribes might affect an enterprise manager's view of corruption. One such factor might be whether bribes are primarily used to avoid taxes and regulations, or whether they are needed to get services that should be delivered by government agents. For example, an enterprise that is able to bribe tax collectors to avoid large tax bills increases its profits and, thus, might not see corruption as a major obstacle to enterprise performance. This point, of course, should not diminish the negative impact of corruption on the economy as a whole. While one manager might prefer to pay a bribe than to pay a higher tax bill or comply with fire safety regulations, the loss to public finances or public health could far exceed the benefit to the manager. Second, corruption might encourage needless regulation or overly zealous enforcement, as government officials try to maximize their income from bribes. As a share of revenues, small enterprises in Albania, Moldova, the EU accession candidates and the CIS paid higher bribes than medium or large enterprises did. In contrast, in the two countries where bribes accounted for the largest share of revenues (Azerbaijan and the Kyrgyz Republic), the relationship was u-shaped - large enterprises and small enterprises generally paid greater shares of their revenues in bribes than medium-sized Bribe-takers target enterprises. Since small enterprises are an important engine for profitablefirms, employment growth, corruption might be especially destructive turning corruption in countries where small enterprises pay higher bribes. into another tax on Firms with higher margins (i.e., high sales prices relative to profits and operating costs) paid a greater share of their revenues as bribes profitscouragthan firms with low margins did in all countries except for discouraging efficiency 93 Azerbaijan.38 This suggests that bribe-takers target profitable firms, turning corruption into another tax on profits and discouraging efficiency. Corruption was more 'disorganized' in Azerbaijan and Albania, with enterprises reporting that additional bribes for the same service were more common after paying a first bribe. Since disorganized corruption is generally seen as more destructive than 'organized' corruption, this might partly explain the high rating that managers in Albania gave corruption as an obstacle to enterprise performance (see Table 3.12) relative to bribes as a share of revenues (see Table 3.13). However, it does not explain the low rating received by Azerbaijan relative to bribes as a share of enterprise revenues. Access to Financing Enterprises in the four subject countries saw access to finance as a greater obstacle to enterprise performance in Enterprises see access 1999 than enterprises in the EU accession candidates, to finance as a greater with the most common complaints being high interest obstacle than in the rates and access to long-term loans. Enterprises in the Kyrgyz Republic and Azerbaijan rated access as a EU accession moderate to major constraint, while enterprises in countries, with the Albania and Moldova rated it as only a moderate most common constraint. These ratings are broadly consistent with c . macro-level financial data - private sector credit complaints being high accounted for a lower share of GDP in Azerbaijan and interest rates and the Kyrgyz Republic (2.5 percent and 2.8 percent, access to long-term respectively) than in Albania and Moldova (3.3 percent loans and 4.6 percent, respectively), although levels in general are low in all four subject countries. Remittance flows may serve as a partial substitute, particularly in Albania. 38 The positive correlation between bribe payments and margins was statistically significant in Albania and the Kyrgyz Republic. The coefficient was positive but statistically insignificant in Moldova and negative and statistically insignificant in Azerbaijan. 94 Table 3.14 Financing as an Obstacle to Enterprise Performance in 1999 High Access to Banks lack Overall Interest long-term money to Rates loans lend Mean Rating 2.8 3.1 3.0 2.5 Albania (Position) (7) (1) (2) (4) Azerbaijan Mean Rating 2.8 3.1 3.4 3.2 (Position) (2) (3) (1) (2) Mean Rating 3.4 3.8 3.5 2.9 Kyrgyz Republic (Position) (4) (1) (2) (4) MeanRating 3.5 3.7 3.0 2.7 Moldova (Position) (5) (1) (2) (4) EU accession Mean Rating 2.6 3.1 2.1 1.6 candidates (Position) (3) (1) (4) (10) Commonwealth of Mean Rating 3.2 3.3 3.0 2.4 Independent States (Position) (3) (1) (2) (6) Data Source: WBES Note: High scores mean greater constraint. Positions for high interest rates, access to long-term loans and banks lack money to lend are relative to the 12 sub-constraints related to access to financing. The constraints are the three obstacles above, collateral requirements, bank paperwork and bureaucracy, access to lease finance, need connections, access to export finance, inadequate credit information, access to foreign banks, and corruption of bank officials. In all four subject countries, small enterprises saw access to financing as a greater constraint to operations and growth than medium-sized and large enterprises Small enterprises did. This was also true in the EU accession candidates, but was see access to not true in the CIS as a whole. Given that small de novo enterprises are the engine for employment growth in the financing as a transition economies, this could have important implications for greater constraint long-term growth. Although small enterprises consistently rated to operations and access to finance as a more significant constraint than medium and large enterprises (see Table 3.15), there was less consistency growth than in the responses of medium-sized and large enterprises. In medium-sized and Azerbaijan and the Kyrgyz Republic, large enterprises rated large enterprises do access to financing as a greater obstacle than medium-sized enterprises, while the reverse was true in the other two subject countries. 95 Table 3.15 Effect of access to financing on small, medium and large enterprises in 1999. Kyrgyz ~EU Albania Azerbaijan Kyrgyz Moldova accession cis Republic candidates Very Small 2.9 3.1 3.6 3.7 2.7 3.1 (< 10 employees) Medium-sized 2 8 2.6 3.4 3.5 2.6 3.1 (between 10 and 100) Large 2.6 3.0 3.4 3.4 2.5 3.3 (> 100 employees) Data Source: WBES Note: High scores mean greater constraint. Legal System, Functioning of the Judiciary and Other Constraints Current legislation in the areas of commercial law and the protection of shareholders' rights are an area for improvement in the four subject countries. Based on a survey of lawyers and academics, EBRD Commercial law and (2000) assesses the "extensiveness" (i.e., quality of protection oj written laws) and "effectiveness" (i.e., enforcement of shareholders'rights are those laws) of commercial law in the transition economies. In general, commercial law in the four areasfor improvement subject countries was farther from international standards than commercial law in the EU accession candidates (see Table 3.16). Similarly, existing laws appear to protect shareholders' rights slightly less well in the four subject countries than they do in the EU accession candidates. However, legal protections for creditors appear, at least on paper, as strong in the four subject countries as they do in the EU accession candidates. It is important to note that, in general, the transition economies score relatively well on many measures of shareholder and creditor rights, even compared to developed economies.39 For example, all the four subject countries score better than Germany with respect to shareholder rights and, as well as or better than Germany with respect to creditor rights (see Table 3.16). This largely reflects the fact that many transition countries have taken the opportunity to write new laws that reflect best current practices. Industrialized countries, in contrast, are often burdened with imperfect laws that have a long history, and enterprises as well as creditors have learned to accommodate their behavior and expectations to their specific shortcomings. It is also important to note that measures other than "effectiveness" of commercial laws are based on written laws, and do not reflect quality of enforcement. Here, scores would decline in the subject countries, and probably improve in the better performing advanced economies. 39 See Pistor et al, 2000. 96 Table 3.16 Quality of Commercial Law in 1998 and 2000 Commercial Law in 2000 (includes pledge, bankruptcy and Legal Protections in 1998 company) Extensiveness Effectiveness Shareholder Rights Creditor Rights (out of 4.3) (out of 4.3) (out of 6.0) (out of 4.0) Four Subject Countries Albania 3.3 1.7 3.0 3.0 Azerbaijan 3.0 2.0 2.0 4.0 Kyrgyz Republic 3.3 3.0 2.25 3.0 Moldova 3.0 2.0 3.5 4.0 EU accession candidates Czech Republic 3.0 3.3 3.0 3.0 Estonia 3.7 3.3 3.75 4.0 Hungary 4.0 3.7 3.0 3.75 Poland 3.7 4.0 3.0 2.25 Slovenia 4.0 3.7 2.5 4.0 Commonwealth of Independent States Armenia 3.7 2.0 5.5 3.0 Belarus 1.0 2.3 1.5 2.0 Georgia 3.0 2.0 3.0 2.75 Kazakhstan 4.0 3.7 5.25 2.75 The Russian 3.7 3.0 5.5 2.5 Federation Tajikistan 2.0 1.7 --- --- Ukraine 3.3 2.0 3.5 4.0 Uzbekistan 3.0 2.3 3.5 2.5 Developed Economies Germany --- --- 1.0 3.0 United Kingdom --- --- 5.0 4.0 United States --- --- 5.0 1.0 Data Sources: EBRD (2000) for company laws Pistor et al. (2000) for legal protection for transition econonies; and La Porta et. al (1998) for legal protection for shareholders and creditors. In all cases, higher scores indicate better protections. Although the four subject countries score only slightly lower with respect to the quality of written commercial laws than the EU accession candidates, they do far worse with respect to enforcement of such laws. All four subject countries except the Kyrgyz Republic received a score of 2.0 or lower on legal effectiveness, indicating that legal rules are unclear and that few meaningful procedures are in place to enforce them. This is troublesome, since recent research suggests that measures of enforcement are far better predictors of capital markets development than measures based on written laws.40 This suggests that it is more important for the four subject countries to build the legal and judicial infrastructure to enforce existing law better than it is to work on re-writing the laws themselves. Some recent actions might improve legal effectiveness in the subject countries in the medium term. For example, the establishment of regional land registration offices in Azerbaij an might improve the process of registering land titles and make transactions and land market development easier. 40 Pistor et al, 2000. 97 Interestingly, although all four subject countries scored poorly with The slow pace respect to legal effectiveness, enterprises in these countries did not Of reform is see the functioning of the judiciary as a major constraint on enterprise perfornance (see Table 3.17). However, consistent with reflected in the the previous evidence concerning legal effectiveness, enterprises in poor quality oj these countries did rate the functioning of the judiciary as a greater obstacle to enterprise perfonnance than enterprises in the EU Ifrastructure accession candidates did. Table 3.17 Quality of the Judiciary, Infrastructure and Anti-competitive Behavior as Obstacles to Enterprise Performance in 1999 Functioning of Anti-competitive the judiciary Infrastructure behavior Azerbaijan Mean Rating 2.4 2.3 2.7 (Position ~(6 ~ 10) (4) Moldova Mean Rating 2.5 2.6 2.9 (Position) (11) (10) (8) Commonwealth of Mean Rating 2.0 2.1 Independent States (Position) (I 1) (10) (7) Data Source: WBES Note: High scores mean greater constraint. In general, infrastructure reform in the four subject countries has lagged behind reform in the EU accession candidates (see Figure 3.3), Figure 3.3 Infrastructure Reform in the Four although progress has recently been made in Subject Countries is Lagging. some countries. The slow pace of reform is reflected in the poor quality of infrastructure in 3.5 - __ _. these countries. For example, the number of 3.5 j main telephone lines is lower in the four 3 -J subject countries than for the EU accession 5 candidates or for the CIS as a whole (see 2.5 Figure 3.4). Some recent reforms might * 2 improve infrastructure in the medium term. : 1.5- s * X * U X _For example, in Albania, 85 percent of the U | | * * * state-owned mobile telecommunications " 1- - - - " " - ~ company, Albanian Mobile Communications, - 0.5- _ * m m * * _ was sold to a consortium of Norwegian and _ * * * * Greek companies for US$96 million in July 0 2000. Further, in February 2000, three of five co 2=* Cn O O : power distributors in Moldova were sold to a cr @ P a XO = _Spanish company for US$25 million, and a Data Source: EBRD (2000) 98 second cellular license was sold to a Moldovan-Turkish joint venture. Except in Albania, where infrastructure was ranked as a moderate obstacle to enterprise performance, infrastructure was not seen as a major problem in the four subject countries. In the three CIS countries, infrastructure rated as only a minor or minor to moderate obstacle. In fact, on average, it was rated as the least or second least significant constraint in all three countries. This suggests that rehabilitation of infrastructure is important, but should not necessarily be as high a priority in strategies to improve the business climate as other concems articulated by the enterprise sector (i.e., corruption, taxation). In contrast, enterprises in Albania saw infrastructure as a moderate constraint on Figure 3.4 Te Pon e Co verae was a enteprie pe6°ranc andgroth. he oorProblem in the Four Subject Countries in enterprise performnance and growth. The poor 1999. rating given to infrastructure in Albania reflects several problems. First, the 40 telecommunications network in Albania is 35 less developed than in the other three four subject countries (see Figure 3.4). Second, 3 the quality of the road network, which was 25 already poor, suffered further damage during R20 the Kosovo conflict due to heavy usage by e r,_l NATO forces.4' In addition, the 1 underdeveloped power distribution network, E 10 below average rainfall that reduced 5 hydroelectric capacity, and inadequate construction of generating capacity has led to , D D r 42~~~~~~~~~~ power rationing and supply cuts. Finally, M £ the rail network deteriorated in the early Gb years of the transition as track and signal lines were sold as scrap.43 The problems with Source: International Telecommunications Union. infrastructure in Albania are especially noticeable given the comparisons to the other three subject countries, where infrastructure only ranked as a minor or minor to moderate constraint. This underlines the importance of privatization and regulatory refonn in Albania. Except in the area of telecommunications (where the mobile network was privatized, privatization plans have been made for the fixed line operator, and a second mobile license has been auctioned), significant reform is still needed. 41 Economnist Intelligence Unit, 2000, p. 18. 42 Economist Intelligence Unit Country Report (2000) states that hydroelectric stations generate 98 percent of power, and that no new generating capacity in Albania has been built for 20 years. Economist Intelligence Unit, 2000, p. 19. 99 The poor development of the four subject Figure 3.5. Internet Access was countries' telecommunications sectors means that Inadequate in the Four Subject Countries internet access is scarce, even by regional in 1999 standards. Further, enterprises in the three CIS subject countries were even less likely to have 70% access to the internet than enterprises in other 60% - CIS countries. In large part, the low level of 50% - internet connectivity reflects the poor 50% telecommunications infrastructure in the four 40%0/ subject countries." There were fewer main lines 30* per 100 inhabitants in those countries (see Figure 3.5). Internet connectivity is especially low in 20% the Kyrgyz Republic and in Azerbaijan, where 10%- _ only 3.2 percent and 7.8 percent of enterprises, respectively, reported having connections to the 0%- _ 2 C, internet (see Figure 3.5). The low level of c 8 , _ internet connectivity, and the poor development .0 C ' R n 2 ,, U U Unless vouchers are tradable and foreign :g o 2 investors can freely purchase vouchers and participate in voucher privatization Data Source: EBRD (1997, 1999) schemes without additional restrictions, voucher privatization schemes - even if 62 Further, Blanchard and Aghion (1996) point out that in environments where unemployment is high, there is a wedge between the value of an enterprise to insiders and the value to outsiders, which prevents the efficient sale of enterprises to outsiders. 63 The primary method of privatization was the method that was listed as the primary method in EBRD (1999). Voucher schemes were designated as 'concessions to insiders' and 'equal access' based on Table 5.7 in EBRD (1997). The scheme used in Azerbaijan, which is not included in the table in EBRD (1997), is also classified as having concessions to insiders. 110 they do not favor insiders - might also deter foreign direct investment. For example, although foreign investors could buy vouchers from the public in Azerbaijan, they also had to purchase 'options' from the government to be allowed to use the vouchers during auctions. This increased the cost of participating in the voucher privatization scheme to foreigners, relative to the cost to domestic investors, further deterring foreign investment. 111 CHAPTER FOUR: INTER-SECTORAL ARREARS AND FINANCING ISSUES Box 4.1 CHALLENGES IN ENTERPRISE ARREARS Areas Challenges Soft Budget * The few cash subsidies remaining in the four subject countries appear to Constraints benefit the least efficient enterprises, typically those that are state owned * Arrears to the budget, to suppliers and to employees (wages) have grown to replace cash subsidies as the core of the soft budget constraint applicable to enterprises * Government tolerance of tax arrears encourages arrears in other parts of the economy * High levels of wage arrears have a direct impact on levels of poverty Barter * The prevalence of barter permits enterprise owners and managers to hide revenues, which impairs tax collection efforts and corporate governance * Barter also contributes to a lack of transparency in the real sector, which impairs growth and competition * Government acceptance of barter in discharge of tax liabilities deprives it of fiscal flexibility and encourages further use of barter in the economy Arrears * Government-inspired efforts to eliminate arrears through netting impairs Netting financial discipline in the real sector without addressing the root causes of Exercises the arrears The concept of a soft budget constraint was first introduced to describe subsidies paid to money- losing enterprises, usually by the state, to ensure their survival.64 The idea was that, rather than allowing enterprises to close, a paternalistic government would prefer to cover a money-losing enterprise's losses, usually to avoid job losses or social unrest.f65 These payments, which could take the form of direct subsidies, loans on concessional terms or tolerance of tax arrears, would be granted to loss-making enterprises after the enterprise had been observed to be losing money.66 The mere existence of subsidies does not always mean that an enterprise faces a soft budget constraint. For example, if a governrnent requires that an enterprise set prices below long-run marginal cost but provides it with subsidies to compensate, the firm could still face a hard budget constraint if the subsidies were based on the costs of an 'efficient' enterprise, rather than the enterprise's actual costs. However, in this case, the government would have to know what the costs of an efficient firm were and would have to be willing to allow the enterprise to become bankrupt if its costs exceed those of an efficient firm. Since this is very difficult in practice, actual subsidies almost always depend on actual costs, and the presence of a soft budget constraint becomes all-but-inevitable. 64 See Kornai, 1980. 65 See Schaffer, 1998, p. 81. 66 See Schaffer, 1998. 113 As noted in Chapters 2 and 3, the existence of soft budget constraints has been correlated, in most of the subject countries, with poor loan portfolio performance and a build-up of unsustainable levels of inter-enterprise arrears that distorts pricing, constrains liquidity, and undermines productivity and competitiveness. For instance, in Albania and Azerbaijan, the state banks' NPLs reached unsustainable levels in the late 1990s, requiring costly recapitalization plans. Where losses mounted outside the banking system, such as in the Kyrgyz Republic, this was due to the existence of soft budget constraints in the state enterprise sector. In Moldova, where banks have generally limited their lending, soft budget constraints have been largely manipulated through arrears and the netting mechanisms systematically applied by the government. In all four subject countries, soft budget constraints have broadly correlated with poor performance, although many private firms have also fared poorly, indicating the problem is even larger than the issue of budget constraints. SOFT BUDGET CONSTRAINTS IN THE TRANSmON ECONOMIES In recent years, the concept of 'soft budget constraints' has been broadened to cover a wide range of practices that allow money-losing enterprises to continue to operate. These practices include some activities that appear close to the original concept of a soft budget constraint, such as direct cash subsidies, and some that do not, but which appear to allow money-losing enterprises to continue operating, distorting economic activity and obscuring market transactions and prices (e.g., directed credit, the tolerance of tax arrears, the use of non-cash payments). This section discusses the prevalence and characteristics of many of these practices, including subsidies, wage arrears, inter-enterprise arrears, tax arrears, and the use of non-cash payments in the four subject countries. Direct Subsidies Macro data from the EBRD (2000) suggest that subsidies have fallen in most of the subject countries, since the early 1990s (see Figure 4.1). The most dramnatic drop occurred in Azerbaijan, where subsidies fell from over 11 percent of GDP in 1992, to about 2 percent of GDP by 1996, and then to only 0.1 percent of GDP by 1999. However, subsidies also fell in Albania, from about 2 percent of GDP in 1993 to 0.5 percent of GDP byl999. Subsidies have In contrast, subsidies in the Kyrgyz Republic have been steady, at fallen in most around 2 percent of GDP, since 1994. There is little evidence of a countries since drop in subsidies since 1996 in the WBES data, with the number of the early 1 990s enterprises reporting subsidies increasing in some cases and falling in others. 114 Direct subsidies appear to be significantly lower in the four subject countries than they are in either the Czech Republic or Hungary (see Figure 4.2). Whereas subsidies amounted to less than 1 percent of GDP in Albania and Azerbaijan and 2.3 percent of GDP in the Kyrgyz Republic, they amounted to close to 5 percent of GDP in Hungary and close to 8 percent of GDP in the Czech Republic. This suggests that direct subsidies are probably no more significant in the four subject countries than they are elsewhere in Central Europe and the Baltics (CEB) and the CIS. Figure 4.1 Subsidies Generally Fell as a Share of GDP between 1992 and 1999. 12 10 0 4 2- 1992 1993 1994 1995 1996 1997 1998 1999 -0-Azerbaijan -W-Albania A+KyrgyzRepublic Source: EBRD (2000) The evidence from the WBES suggests Figure 4.2 Subsidies are not Unusually High in the that the level of direct subsidies is Four Subject Countries. similar in the four subject countries, other CIS countries and EU accession 9 candidates. On average, enterprises 8 from the EU accession candidates were 7 more likely to report receiving 7 subsidies or tolerance of tax arrears than enterprises in any of the four ~5- subject countries except Albania (see (9 4 lil ll Figure 4.3). Whereas 14.8 percent of o 3 - enterprises in the EU accession 2 candidates reported receiving subsidies 1 in 1999, only 3.9 percent of enterprises o - , _, *, *, *, lll, 1ll, - in the Kyrgyz Republic, 13.1 percent of _ X m 2 g enterprises in Azerbaijan, and 13.2 -° D ° percent of enterprises in Moldova 6 QD Y) o: I ° t W t°reported receiving subsidies. Data Source: EBRD (2000) Note: Data for Moldova was not available. 115 However, there were some Figure 43 Subsidies were No More Common in the discrepancies between the macro data Four Subject Countries than in the CIS or the EU on subsidies and the evidence from the Accession Candidates in 1999 WBES. Most notably, although te reported subsidies in Albania and :2 25%- Azerbaijan were lower than in the _ __ 2% Kyrgyz Republic (see Figure 4.1), far = _ more enterprises in these countries 0v- 15%- ~ t _ 5reported receiving subsidies (see Figure 10% _ 4.3). There are several plausible ** 5% * _ | | _ explanations for the differences. First, if * * _l * | _ a small number of large enterprises 0 0% - receive large subsidies, then total O cV NX C., D to subsidies might be higher in that < a 0 country (i.e., the Kyrgyz Republic) than w < 6 in another country where a large number IN01996 01999 of small enterprises receive smaller subsidies. Second, since the data from Data Source: WBES the WBES include enterprises that claim Note: For comparison between years only includes enterprises that existed that tax arrears were tolerated, the in 1996. difference might be because tax arrears, which are omitted from the previous measure of subsidies, are more common in Albania and Azerbaijan than in the Kyrgyz Republic. Finally, the difference might be because the macro level data depend on different definitions of subsidies across countries, or that some subsidies that are counted in the Kyrgyz Republic are omitted in Albania and Azerbaijan. In general, it is difficult to accurately measure subsidies in a comparable way. Although, as discussed above, subsidies do not inevitably mean soft budget constraints, evidence from the WBES suggests that they do reflect soft budget constraints in the four subject countries. State-owned enterprises were more likely to receive subsidies than private enterprises were in all of the four (see Table 4.1). Since private enterprises are more efficient than similar state-owned enterprises67, this suggests that, at least in part, subsidies continue to go to inefficient enterprises, delaying needed restructuring. 67 See Djankov and Murrell, 2000 and Sherif et al, 2000. 116 Table 4.1 Enterprises Reporting Subsidies or Tolerance of Tax Arrears in 1999 Albania Azerbaijan Kyrgyz Moldova EU accession CIS Republic candidates Private 9% 12% 2% 14% 8% 10% State-Owned 48% 29% 15% 15% 46% 19% Data Source: WBES Low levels of One other interesting note is that the percentage of state enterprises reporting subsidies or tax arrears tolerance is higher in non-CIS subsidies in CIS countries than in CIS countries. This includes the five EU candidate countries may countries, where state enterprises still appear to receive benefits of this be due to the sort. By extension, the reason for low levels reported in CIS countries may be due to the lower fiscal capacity to subsidize, although this lower fiscal does not explain the tolerance of tax arrears that is more prevalent in capacity to CIS countries. However, notwithstanding these differences, the low subsidize overall level of direct subsidies partly reflects the relatively weak fiscal situation in the four subject countries during the 1990s. In 1998, the governments of all four subject countries had fiscal deficits that were greater than 4 percent of GDP, and two of the subject countries, Albania and the Kyrgyz Republic, had fiscal deficits that were greater than 10 percent of GDP. Since the weak fiscal situation put great pressure on the governments to reduce expenditures in non-essential areas, it is not surprising that the governments have been forced to reduce subsidies. However, this does not rule out the possibility of further reductions in subsidies - the fiscal situation in the four subject countries remains weak and scope remains for further reductions. ARREARS Enterprises were far more likely to report having overdue payments than they were to report receiving subsidies in all of Enterprises are far the subject countries.68 For example, whereas only 18 percent more Ikel t of enterprises in Albania reported that they received direct I Y ° subsidies (see Figure 4.3), over 40 percent reported having report overdue overdue payments to at least one of five groups of creditors payments than to (see Figure 4.4). Similarly, only 15 percent of enterprises in report receiving Moldova reported receiving subsidies, while close to 80 percent of enterprises reported having overdue payments. subsidies 68 Arrears are defined as payments that are over 90 days overdue. 117 Enterprises with Although the number of enterprises with overdue payments varied greatly among the subject countries, enterprises in these countries overdue payments other than Albania were more likely to have overdue payments to any single than were enterprises in the EU accession candidates. Whereas group of creditors only about 40 percent of enterprises in the EU accession candidates had overdue payments to any of the five groups of creditors, over are more likely to 60 percent of enterprises had overdue payments in the four subject have overdue countries in the CIS (see Figure 4.4). Enterprises in Moldova were payments to other most likely to have overdue payments - 81 percent of enterprises had overdue payments to at least one of the five groups of groups creditors. In contrast to the other subject countries, fewer enterprises reported overdue payments in Albania than in the EU accession candidates. However, as discussed below, other evidence suggests that the difference between Albania and the other four subject countries might be less than this evidence suggests. In all of the four subject countries except Azerbaijan, enterprises were more likely to report having inter- Figure 4.4 Overdue Payments were More Common enterprise arrears than to any other in the Four Subject Countries than in the EU creditos (seeTable .2). ~Accession Candidates in 1999 creditors (see Table 4.2). hIn Azerbaijan, although enterprises were slightly more likely to report 80% _ overdue paymnents to the national e-70% government, they were more likely 4 Q 60% - * to report overpayments to suppliers - 50%v than they were to any creditors , 40%- * _ * 3 except the government. Overdue ! 30% v payments to workers were also ' 8 20% * I f * * common in some of the subject ae 10% v I I I 31 countries, with enterprises in oA , C N . ' m~~ N Moldova and the Kyrgyz Republic ,, o reporting wage arrears as the second < (D -1 5 u most common type of overdue 6 payment. This is in sharp contrast to l-Percent reporting overdue payments enterprises in the countries that are U Percent reporting serious overdue payments EU accession candidates (and in Azerbaijan), where wage arrears Data Source: WBES were the least common type of Note: Percent of enterprises reporting overdue payments either to utilities, wrreare. Somewhat surprisingly, I suppliers, local or national government, or workers. arrears. Somewhat surprisingly, overdue payments to utilities were not especially common in any of the subject countries; they were the least common type of overdue payment in Moldova and the second least common type in Azerbaijan. Again, this is a sharp contrast to the situation in the EU accession candidates, where overdue payments to utilities were the second most common type of overdue payment. This also contradicts other data sources that have concluded that electricity companies are often major creditors to enterprises and households in CIS countries. 118 Table 4.2 Percent of Enterprises Reporting Overdue Payments in 1999 Overdue Overdue Overdue Overdue Overdue Payments to Payments to Payments to Payments to Payments to Utilities National Local Suppliers Workers Government Government Albania 19.5% 24.2% 13.7% 37.3% 18.8% Azerbaijan 60.7% 63.1% 61.0% 62.5% 56.7% Kyrgyz Republic 40.8% 31.2% 39.0% 49.6% 47.1% Moldova 37.2% 54.2% 54.5% 72.8% 66.1% EU candidates 26.1% 22.2% 19.8% 38.4% 17.6% cIS 39.2% 40.6% 39.4% 51.9% 44.5% Data Source: WBES In general, enterprises with overdue payments to any single group of creditors were more likely to have overdue payments to other groups. The high correlation between different types of overdue payments across enterprises (see Table 4.3) suggests that firms do not have the incentive to behave strategically, paying one group of creditors off quickly while letting other types of debts accumulate. The correlations are highest in Azerbaijan and the Kyrgyz Republic and lowest in Moldova. Table 4.3 Correlations Between Overdue Payments to Government and Other Creditors Albania Azerbaijan Kyrgyz Moldova Republic Overdue Payments to Suppliers 0.56 0.73 0.62 0.43 Overdue Payments to Utilities 0.63 0.66 0.60 0.38 Overdue Payments to Workers 0.66 0.67 0.71 0.56 Overdue Payments to Local Government 0.76 0.82 0.79 0.73 Data Source: WBES Note: All correlations are significant at a 1% level or higher. Correlations are simple correlations for categorical variables (4=no overdue payments, 3=modest overdue payments, 2=manageable overdue payments and 1=substantial overdue payments). In the three CIS subject countries, insider- and In the CIS countries, insider- state-owned enterprises were more likely to and state-owned enterprises report overdue payments than outsider-owned private enterprises (see Table 4.4). Given the are more likely to report substantial evidence that private enterprises in the overdue payments than transition economies perform better than state- outsider-ownedprivate owned enterprises and insider-owned enterprises do, this provides evidence that non-payment is enterprises primarily concentrated among poorly performing 119 enterprises.69 In contrast to the results for the three subject countries in the CIS, state-owned enterprises in Albania were less likely to report overdue payments than private enterprises were. Table 4.4 Firms Reporting Overdue Payments by Ownership in 1999 Albania Azerbaijan Kyrgyz Moldova Republic Insider-owned private --- --- 74% 100% Outsider-owned private 45% 59% 54% 78% State-owned 41% 75% 79% 81% Foreign 38% --- --- Data Source: WBES Inter-Enterprise Arrears Inter-enterprise arrears are more common in the four subject countries than in the EU accession candidates. Overall, inter-enterprise arrears do not appear to be larger (as a share of GDP) in the EU accession candidates, than Inter-enterprise they are in some OECD countries70. Consistent with this, arrears are more foreign-owned enterprises in the EU accession candidates were common than in no less likely to report overdue payments than were other the EU accession enterprises. Although it is not possible to estimate inter- enterprise arrears as share of GDP from the data collected in countries the WBES, the data suggest that inter-enterprise arrears are more common in the CIS subject countries than they are in the EU accession candidates (see Figure 4.5). In contrast, the level of inter-enterprise arrears in Albania was similar to those in the EU accession candidates. Some analysts have argued that inter-enterprise arrears are quite different from tax or wage arrears,71 suggesting that "Overdue trade credit [i.e., inter-enterprise arrears] has not usually been a problem because firms have learned to apply very basic credit control mechanisms to each other. Surveys of firms in [countries in CEB and the CIS] present clear evidence that suppliers do not continue to ship to customers who have not paid cash for goods already received... To consider overdue trade credit as so-called 'involuntary credit' as is sometimes done misses the point that it takes firms only a matter of months at the beginning of transition to learn that late payment is endemic. Firms in [countries in CEB and the CIS], as in Western 69 See Djankov and Murrell (2000) and Sherif et al (2000) for evidence on the effect of ownership on enterprise performance in the transition economies. 70 See Schaffer, 1998. In addition to data from the Czech Republic, Hungary and Poland, he also presents data from Romania, Kazakhstan and The Russian Federation. However, he does not present data for any of the four subject countries. 71 See Schaffer, 1998. 120 economies, tolerate late payment by customers and do so simply because they want the business." However, in practice, enterprises with Figure 4.5 Interenterprise Arrears Were More arrears to their suppliers also tended to Common in Four Subject Countries than in the EU have large arrears to other creditors (see Accession Candidates in 1999. Table 4.5). This suggests that inter- enterprise arrears are quite similar to , 80% other forms of arrears in the four subject 2! 70% countries, and that, in general, a large X 60%- _ part of the stock of arrears is held by a 0 distressed enterprises.72 Even if inter- ov 50% -_ c0 _ _ enterprise arrears do not result in soft 3 40%- - - = _ ~ budget constraints per se, they still 20%- _ _ distort economic activity. For example, 2 20% _ r when close to half of all enterprises have 4 10% _ _ _ U overdue payments due, it can become o % _z _ _ _r , _ _-_ difficult to distinguish between good and *<- W., = > - bad credit risks, increasing the risk in D D y>s qZ a normal business relations. This, in turn, w < undermines the willingness of the IMSubstantial *Anyl banking sector to provide credit, which stifles development of the financial sector because of the consequent D cunwillingness of enterprises to place funds with banks for safekeeping (since they do not obtain loans) or to comply with the terms of credit agreements when a bank does decide to lend on commercial terms. In all four subject countries, enterprises were more likely to report being owed overdue payments by their customers than they were to report owing overdue payments to their suppliers. Although this might seem puzzling at first, there are several plausible explanations for the difference. First, enterprise customers differ from their suppliers. If enterprises further down the supply chain are more likely to have overdue payments, then enterprises will tend to report being owed overdue payments more often than owing overdue payments. In addition, most enterprises have many customers, all of whom might have overdue payments with the enterprise. If most enterprises either owe overdue payments to either none of their suppliers or to all of their suppliers that extend credit, then an enterprise with multiple customers will more likely be owed overdue payments than owe overdue payments to other enterprises. In fact, that the two numbers are so close suggests that enterprises do not extend credit to their customers at random. For example, if every enterprise has a 50 percent chance of having arrears to all of its suppliers and a 50 percent chance of owing nothing to any of them, then an enterprise with five 72 Since the data for the WBES were collected in the summer of 1999, many years after the start of the transition even among the countries covered in this study, it is unlikely that most of the stock of arrears was built up in the very early years of the transition. 121 customers would have only a 3 percent chance of not being owed any arrears. Although this might explain the discrepancy between enterprises reporting arrears, it is a less convincing explanation for the similar difference between enterprises reporting 'substantial' arrears and those reporting being owed 'substantial' arrears. For example, in Albania, enterprises were more likely to report being owed substantial amounts (13 percent of enterprises) than owing substantial amounts (3 percent of enterprises).73 A plausible explanation for the discrepancy might be that enterprise managers are more willing to admit being owed substantial amounts during interviews than to admit owing substantial amounts. This could be the case if managers believe that having overdue payments to suppliers reflects poorly on their managerial ability (relative to having customers that owe them money). As noted earlier, fewer enterprises in Albania reported owing overdue payments to their suppliers than in the other four subject countries (see Figure 4.5). However, enterprises in Albania did not appear to be significantly less likely to report being owed overdue payments than enterprises in either Azerbaijan or the Kyrgyz Republic (see Table 4.5). This suggests that the low level of overdue payments in Albania might partially reflect that managers in Albania are less likely to admit having overdue payments to suppliers than managers in the other four subject countries are, rather than that inter-enterprise arrears are less common in Albania.74 Whether looking at overdue payments to suppliers or overdue payments owed by customers, arrears appear to be more common in Moldova than in the other four subject countries. Table 4.5 Arrears Owed to Suppliers and to Customers in 1999 Overdue Payments to Overdue Payments owed by Correlation Suppliers Customers Albania 37% 57% 0.45 Azerbaijan 63% 64% 0.88 Kyrgyz Republic 50% 59% 0.58 Moldova 73% 84% 0.46 Data Source: WBES In the subject countries, enterprises that have overdue payments to their suppliers are also more likely to have overdue payments owed to them. The correlation between overdue payment owed and overdue payment due is larger in the four subject countries than in the EU accession candidates. There are several plausible explanations for the high correlation. First, it is possible that poorly managed enterprises are both less able to pay their own suppliers and to collect payments that are due from firms they supply. However, it is also possible that non-payment spreads through the system, affecting both viable and non-viable enterprises. 73 It is important to note that the question asked "is the amount of overdue accounts receivable ... from customers substantial?" not "is the amount of overdue accounts receivable from ANY customer substantial." If the second question were asked, then the previous comment would apply to this question as well. 74 A simnilar discrepancy is noticeable for enterprises reporting substantial overdue payments. In Albania, 13 percent of enterprises reported having substantial overdue payments owed to them by customers, compared to only 8 percent in Azerbaijan, 21 percent in the Kyrgyz Republic and 31 percent in Moldova. 122 For example, non-payment by a non-viable enterprise to a potentially viable enterpnrse might mean that the potentially viable enterprise is not able to pay its suppliers. In this way, non- payment can work through the real economy and the banking system, making it difficult for enterprises and banks to assess the risk of extending trade credit to any other enterprise. Inter-enterprise arrears are more common in Moldova than in any of the other four subject countries (see Table 4.5). One of the main reasons that enterprises allow other enterprises to build arrears is that they anticipate that the government will eventually step in and arrange a clearing operation.75 This might explain much of the build-up of arrears in Moldova, where non-cash payments made up over one-third of government revenues between 1996 and 1998 (see Table 4.6). These 'netting operations', where tax and penalty receivables are netted out against expenditure arrears, include multi-party arrangements agreements in which inter- enterprise arrears are cancelled out.76 As long as enterprises are able to pass inter-enterprise arrears onto the state in this way, they have only limited incentives to try to reduce arrears to suppliers. Moreover, because this mode spreads and weakens liquidity, barter is often used as a substitute for cash, creating greater scope for pricing distortions through the actual netting process. Table 4.6 Non-cash Revenues for Government in Moldova 1996 1997 1998 Netting operation revenues (% of total revenues) 35.2 36.8 35.8 Netting operations for national government (% of revenues) 33.7 34.7 34.1 Netting operations for local government (% of revenues) 37.4 42.3 41.0 Source: IMF (1999) 75 See Schaffer, 1998. 76 The example of such a multiple party arrangement given in IMF (1999) involves a case where a farm buys fuel from a fuel company, but is unable to pay for it. In tum, the fuel company is then unable to pay excise taxes to the state, which, in turn, is unable to pay the farm for food that the farm supplies to the government. In the netting operation, the government cancels its tax claim on the fuel company, which cancels its claims on the farm for fuel, which, in turn, cancels it claim on the state for food. 123 Wage Arrears Wage arrears are an especially great concern in the CIS Wage arrears are an because of their links with poverty. In contrast to most countries in Central Europe and the Baltics, where especially great concern households with unemployed heads are far more likely in the CIS because oJ to suffer from poverty than households with a working their links with povertv head, poverty was also common among households with working heads in most countries in the CIS. One of the reasons for this is the high level of wage arrears.77 The importance of wage arrears to poverty is also linked to the fact that low wage earners are especially likely to be owed arrears.78 Wage arrears are significantly higher in the three CIS subject countries than in the EU accession candidates. About 57 percent of enterprises in Azerbaijan, 47 percent of enterprises in the Kyrgyz Republic, and 66 percent of Figure 4.6 Wage Arrears Were More Common in enterprises in Moldova reported wage the Four Subject Countries than in the EU arrears, compared to only 18 percent of Accession Candidates. enterprises in the EU accession candidates. The high level of wage 70% arrears in Moldova appears consistent 60% with the high levels of other types of 50% - arrears reported in this country. 50%O -o Although significantly fewer enterprises 40% reported having wage arrears in Albania 30% - than in the other subject countries, the 20% U 5 results from the previous section, which 10% * suggested that enterprise managers in r I U 5 1 I r Albania are less likely to report arrears 0% _ , , , than managers in the other subject m X 1 t0 W a a countries, suggests wage arrears might CD ° 2 W! 'a= -8 be underestimated in Albania. * Substantal overdue payments *Any overdue payments The prevalence of wage arrears among insider- and outsider-owned private enterprises is similar in the four subject Data Source: WBES countries (see Table 4.7). In the two I countries with significant numbers of insider-owned private enterprises - the Kyrgyz Republic and Moldova - insider-owned enterprises were more likely than outsider-owned private enterprises to report wage arrears in the Kyrgyz Republic, but less likely in Moldova. Given that employee-owned enterprises are 77 See World Bank 2000, pp. 71-72. 78 World Bank (2000) notes that whereas 22 percent of poor wage earners suffered due to wage arrears in Tajikistan, only 8 percent of non-poor wage earners did. 124 probably less willing to resort to wage arrears under most circumstances, the relatively similar levels of wage arrears in insider- and outsider-owned enterprises might reflect the relatively poor performance of insider-owned enterprises. Table 4.7 Percent of Firms Reporting Wage Arrears in 1999, by Ownership Albania Azerbaijan Kyrgyz Moldova Republic Insider-owned private --- --- 55% 48% Outsider-owned private 17% 53% 41% 59% State-owned 19% 71% 52% 62% Data Source: WBES State-owned enterprises were more likely to owe wage arrears than outsider-owned private enterprises were. The difference between state-owned and outsider-owned enterprises with respect to wage arrears was small in Moldova and Albania, but large for Azerbaijan and the Kyrgyz Republic (see Table 4.7). Given the importance of wage arrears in predicting poverty at the household level, the poor performance of state-owned enterprises with respect to wage arrears is a great concern. This suggests that one plausible reason for slow reformers having higher levels of poverty than fast reformers (see Box 3.2) might be the prevalence of wage arrears among state-owned enterprises. Tolerance of Tax Arrears Because distressed enterprises are the most likely to be unable to pay their taxes, this is the most direct form of soft budget constraint according to the earliest Tolerance of tax definitions of soft budget constraints.9 Consequently, the arrears means that tolerance of tax arrears means that poorly performing poorly performing enterprises can avoid needed restructuring. Further, when the enterprises can avoid government allows tax arrears to be settled through "netting operations," as has been done extensively in Moldova and needed restructuring.., several other CIS countries, tolerance of tax arrears and the the resulting resulting expenditure arrears can encourage the build-up of expenditure arrears inter-enterprise arrears. encourages the build- Tax arrears have become a major problem in many transition up of inter-enterprise economies, including several of the four subject countries. Tax arrears were estimated to be close to 15 percent of GDP arrears in Moldova by 1998, and close to 1 percent of GDP in the Kyrgyz Republic (see Table 4.8). Since personal income taxes typically make up only a small part of revenues in the transition economies, most arrears are owed by enterprises rather than by households. 79 See Schaffer, 1998. 125 Table 4.8 Tax Arrears as Percent of GDP 1995 1996 1997 1998 1999 Moldova 10.91 13.65 11.19 14.67 --- Kyrgyz Republic 0.84 0.89 1.66 1.33 0.76 Source: IMF (1999) and IMF (2000). Note: Data for Moldova include arrears to social funds. Although comparable data are not Figure 4.7 Tax Arrears Were Highest in available for Azerbaijan or Albania, Azerbaijan. evidence from the WBES suggests that tax arrears are most common in e 70% Azerbaijan. As for the other forms of t 60% arrears, enterprises in the three CIS x 50% _ countries were more likely to report 50% having tax arrears than enterprises in o 40% _ _- Albania or the EU accession candidates r_40% Fgr .) 30% (see Figure 4.7). Over 60 percent of A. 20% _ ~ ~ _ _ _ enterprises in Moldova and Azerbaijan 10% F F I * F reported having tax arrears, compared to I K E 1 Jr Ffl 24 percent of enterprises in the EU 0% , ,, accession candidates. Enterprises in _ Albania were less likely to report tax n £ y' < ; -a arrears than enterprises in the EU w u< accession candidates. Although the * Substantial overdue payments *Any overdue payments| evidence regarding inter-enterprise arrears suggests that this might Data Source: WBES underestimate the extent of arrears in Note: Includes arrears to both local and national Albania, tax arrears are generally thought governments. to be a lesser problem in Albania than in the other subject countries. In contrast to the other types of arrears, which were most common in Moldova, enterprises in Azerbaijan were slightly more likely to report having arrears than enterprises in Moldova were. This suggests that although SOCAR might be responsible for a large part of the stock of arrears, other enterprises also have problems with tax arrears in Azerbaijan. In the four subject countries with significant insider-ownership, insider-owned private enterprises were the most likely to have tax arrears. In both the Kyrgyz Republic and Moldova, insider-owned enterprises were, by far, the most likely to have tax arrears (see Table 4.9). Since insider-owned enterprises (especially employee-owned enterprises which account for most "insider-owned" enterprises in the subject countries) are probably especially unwilling to lay off surplus workers, this is consistent with the hypothesis that governments tolerate arrears to avoid job losses. In contrast to the other types of arrears, state-owned enterprises were not consistently more likely to have tax arrears than (outsider-owned) private enterprises. This suggests that government ownership might give the government greater leverage over the enterprises, resulting in reduced levels of tax arrears 126 Table 4.9 Firms Reporting Tax Arrears in 1999, by Ownership Albania Azerbaijan Kyrgyz Moldova Republic Insider-owned private --- --- 45% 86% Outsider-owned private 22% 60% 27% 50% State-owned 29% 74% 18% 43% Foreign 25% --- --- Data Source: WBES Although relatively few enterprises reported that they received subsidies or tolerance of tax arrears from the government, many more reported actual arrears. In all the four subject countries, even among enterprises with 'moderate' or 'substantial' tax arrears, very few enterprises reported tolerance of arrears (see Table 4.10). This suggests that relatively few enterprises receive formal tolerance of arrears, but that a far larger share receive at least tacit acceptance (since they have not been forced into bankruptcy proceedings) in all of the subject countries. This has been a major deterrent to bank lending, and has undermined the effectiveness of the legal framework regarding bankruptcy enforcement. Table 4.10 Enterprises Reporting Tolerance of Tax arrears and Actual Arrears in 1999 Enterprises Enterprises Percent of enterprises reporting tax subsidies with moderate or Total arrears to icungtlrce substantial tax arrears national including tolerancethtrp oeacf national of tax arrears that report tolerance of government tax arrears or subsidies Albania 150 38 26 33.3% Azerbaijan 122 77 19 22.2% KyrgyzRepublic 112 35 4 15.8% Moldova 120 65 17 16.2% Data Source: WBES Note: Total is number of enterprises with data on both subsidies and tax arrears. The high level of tax arrears in the four subject countries, combined with the low level of revenue reported to the tax authorities, have contributed to the weak state of public finances in the four subject countries. In addition to arrears, another contributing factor to the poor state of public finances is that many enterprises reported only a portion of their revenues to the tax authorities (see Table 4.11). Fewer enterprises in the Kyrgyz Republic reported 100 percent of revenues to the government than in the other two subject countries for which data were available. However, the problem appears to be most severe in Azerbaijan, where 41 percent of enterprises reported less than 60 percent of revenues were reported to the authorities. 80 Data were not available for Albania. To try to encourage truthful responses, enterprise managers were asked about typical firms in their area of activity rather than their firm's actual behavior. 127 Table 4.11 Percent of revenues reported to tax authorities in 1999. Azerbaijan Kyrgyz Republic Moldova 100 percent 25.8% 16.0% 45.5% Between 90 and 99 percent 10.9% 19.2% 11.4% Between 80 and 89 percent 8.6% 43.2% 6.5% Between 70 and 79 percent 5.5% 4.0% 8.1% Between 60 and 69 percent 7.8% 7.2% 4.9% Between 50 and 59 percent 14.1% 8.8% 8.9% Between 25 and 49 percent 16.4% 0.8% 8.9% Less than 25 percent 10.9% 0.8% 5.7% Data Source: WBES Notes: Answer was in response to question: "What percentage of the sales of a typical firm in your area of activity would you estimate is reported to the tax authorities, bearing in mind difficulties with complying with taxes and other regulations." Response was categorical. Data was not available for Albania. BARTER Non-cash payments are, in general, far more common in three of the four subject countries (other than Azerbaijan) than they are in the EU accession candidates (see Table 4.12). Non- cash payments in the three subject countries were generally comparable to those reported for the CIS countries. Non-cash payments were especially common in Moldova, where 77 percent of enterprises reported conducting some sales using non-cash payments, and 23 percent reported more than 50 percent of sales using non-cash payments. Barter was especially common among Moldovan enterprises in the farming, forestry and fishing sector, where 93 percent of enterprises reported conducting some sales using non-cash payments, and 40 percent reported conducting over 50 percent of their sales with non-cash payments.8' 81 A similar pattern can be observed in other CIS countries. For the CIS as a whole, 25 percent of enterprises in the farming, fishing and forestry sectors used non-cash payments to conduct some sales, compared with 1 1 percent of other enterprises. 128 Table 4.12 Non-cash Sales Reported by Enterprises in 1999 Azebaian Kyrgyz Commonwealth EU Azerbaijan Republic Moldova of Independent accession States candidates Conduct some sales using 23% 53% 77% 56% 41% non-cash payments More than 10 percent of 13% 42% 60% 39% 15% sales More that 50 percent of 1% 12% 23% 14% 1% sales Data Source: WBES Barter distorts economic activity and reduces the transparency of economic arrangements.82 Although some authors have suggested that the use of barter reflects managers' ability to survive in difficult economic circumstances,83 barter imposes real costs on the economy. First, it allows managers to hide revenues from minority shareholders. In the weak institutional environments present in many CIS economies, this could make it easier for insider-owners to divert profits from minority shareholders to managers or employees. Second, it can stifle competition and allow inefficient producers to remain operational. This, in turn, makes it more difficult for efficient firms to survive. Third, it can undernine tax collection, making it easier to hide revenue from the authorities. In addition to providing enterprises with an additional way of hiding transactions (e.g., by avoiding the banking system), barter also obscures market prices, making it more difficult to collect value-added taxes. Given the high levels of non-reported taxes in the four When large amounts oJ subject countries (see Table 4.11), and the large fiscal sales are conducted deficits facing the governments of these countries, these distortions can harm overall economic performance. When using barter, large amounts of sales are conducted using barter, enterprises often find it enterprises often find it difficult to make cash payments for dificult to make cash their tax obligations. In such circumstances, governments are often forced to accept non-cash payments in the payments for their tax discharge of tax liabilities, which adds to the problem by obligations... which reducing the government's budgetary flexibility.84 Finally, adds to the problem by as discussed in the section on inter-enterprise arrears, if the reducing the govemment starts to accept non-cash payments, as has been common in Moldova, it can promote the emergence government 's of arrears in other parts of the economy. budgetary flexibility 82 International Monetary Fund (1998, p. 45) discusses the problems associated with barter in detail. 83 See Linz and Krueger, 1998. 84 See International Monetary Fund, 1998, p. 46. 129 Non-cash payments are Figure 4.8 Barter Was More Common With Insider- consistently more common among and State-owned Companies in 1999. insider and state-owned enterprises. In the four subject 100%- countries with large numbers of 90%.1 insider-owned enterprises, insider- 80%-- l owned enterprises were far more 60%.,- _ . ]_ likely than either other private 50%-,- ~~~~~enterprises or state-owned 40%--- = _ | _enterprises to conduct sales using 20%-1 _ _non-cash payments (see Figure 10% _ 4.8). For example, in the Kyrgyz o 0 N e O . X Republic, close to 70 percent of 'a , C.) insider-owned enterprises s 5 XQ conducted some sales using non- cash payments, compared to 50 * Insider-owned private *Outsider-owned private percent of state-owned enterprises O_State-owned and less than 50 percent of other Data Source: WBES private enterprises. Although state- Note: Classes with less than five enterprises of that type were owned enterprises were less likely omitted. to conduct sales using non-cash payments than insider-owned enterprises, they were more likely to do so than other private enterprises in all the three subject countries. Combined with the previous evidence on subsidies and arrears, this suggests that state- and insider-owned enterprises tend to face softer budget constraints than other private enterprises. This further underlines the importance of using appropriate techniques for privatization, and the potential costs of using privatization methods that favor insiders. ARREARS, COSTS AND BANK LOSSES In all four of the subject countries, problems of arrears have surfaced as hard budget constraints have been Problems of arrears imposed, and as use of the banking system has diminished have surfaced as hard as a quasi-fiscal channel.85 In Albania, lending flows to state enterprises and farms began to diminish by the mid- 1990s. The same pattern held true in the Kyrgyz Republic have been imposed, and Moldova when directed lending practices were and as the banking severely curtailed. Only in Azerbaijan have exposures to . . . the state enterprise sector remained high. However, even system has dimilnshed as a quasi-fiscal tool 85 The introduction of monetary discipline eventually tightened up on the use of banks as lending channels for loss-making enterprises. 130 here, lending flows from banks have not been sizable. In Albania, arrears were about 4.4 percent of GDP, and generally not a problem. Arrears have been far more prominent in CIS countries. Arrears in Albania were roughly estimated to be about US$165 million in 2000 based on domestic financing figures less privatization revenues. This compares with 1996 figures of about US$293 million, or nearly 11 percent of GDP at the time, when arrears were at their peak. In general, arrears have come down substantially since 1996-97. NPLs of the banks formerly represented about 45 percent of total estimated arrears, but these have generally been cleared. It is unclear how problematic inter-enterprise arrears are relative to arrears to the government and related social funds. Due to early deficiencies in the banking sector, an informal credit market evolved prior to the pyramid schemes, largely financed by foreign exchange dealers. It is unknown the degree to which this kind of financing is technically in arrears. In Azerbaijan, the major risks of arrears overall to the economy appear to be in the oil and power sector. SOCAR (the state oil company) is the country's largest taxpayer, yet also Azerbaijan's largest tax debtor. This is reportedly due to the provision of fuel oil to Azerenergy, the state electricity company, which has low collection rates and does not pay SOCAR for its fuel oil. Arrears to SOCAR were estimated to be 6 percent of GDP in 1999. Ultimately, this plus other arrears were forgone by the budget as lost revenues. This has made it more difficult to finance other budgetary items (e.g., health, education, social insurance/protection), and has also led to gas/electricity shortages in many parts of the country. Azerenergy's low collection rates reflect a subsidy to inefficient manufacturers, usually state- owned, as well as households and government. This is not uncommon in CIS countries. Ultimately, such subsidies will need to be resolved for competitiveness and environmental/energy efficiency purposes. Alternatively, such barter and arrears will perpetuate inefficiency and serve as a drain on public finances. Notwithstanding efforts to reduce some arrears, the total cost of arrears and banking sector problems have accumulated to become nearly half of 2000 GDP levels. Arrears for wages and pensions were cleared in 1999. However, there are large accumulations of arrears in the utility and communal services sectors. According to the IMF, SOCAR's stock of tax arrears was 23 percent of 1999 GDP, about half of total. This would bring total stock of tax arrears to about 7,585 billion manats as of end 1999. Eight other state-owned enterprises, including in gas and electricity, reportedly account for another quarter of cumulative tax arrears. Total cumulative tax arrears at end 1999 approximated US$1.84 billion, about two times total assets of the banking system. Adding bank NPLs and recapitalization costs, this would bring the total cost to about US$2.4 billion, or nearly half of 2000 GDP. The Kyrgyz Republic has faced similar problems of arrears found throughout the CIS. In the enterprise sector, slow corporate restructuring, weak governance, and efforts to impose discipline in the banking sector (as a function of a more stable monetary policy) have triggered an increase in inter-enterprise arrears, estimated to be about US$97.2 million on a net basis at 131 the end of 1999.86 This was equivalent to about 9.1 percent of GDP. Overdue payables and receivables were 15.8 percent and 22.7 percent, respectively. This is a longstanding problem, as net inter-enterprise arrears have long been high in the Kyrgyz Republic. In 1995, they were 10.6 percent of 1995, and overdue payables and receivables were 17.6 percent and 20.7 percent, respectively. Where the problem has become less severe is when adjusted for exchange rates. Net inter-enterprise arrears approximated US$97.2 million in 1999, as compared with US$152.6 million in 1995. Nonetheless, with major structural and financial challenges, these arrears serve as a constraint to corporate restructuring. Most of the arrears in the Kyrgyz Republic are manifested in the form of nonpayment for electricity, agricultural output, raw materials and machinery in the small-scale and retail sectors, and oil and gas suppliers abroad. This has turned electricity producers and nonferrous metals producers into net domestic creditors to the economy, including households. As for energy suppliers from abroad, this has led to deterioration in trade relations with Uzbekistan, Kazakhstan and the Russian Federation, the country's major trade partners within CIS, and among the Kyrgyz Republic's four most important partners in total trade. As in other subject countries, efforts have been made by the Kyrgyz Republic to clear arrears. The Kyrgyz Republic cleared its expenditure arrears in 2000, equivalent to 1.7 percent of GDP in 1998. Combined with accelerated utilities privatization, pension reform, and improved budget management controls, the government hopes to reduce the magnitude of the arrears problem. However, this will take some time in light of the underlying weakness of the economy. This will also likely require a review of the government's public investment program, which is primarily financed by donors but has budgetary implications. The total cost of arrears and banking sector problems has amounted to about US$200 million in the Kyrgyz Republic. This has been about half in the non-bank sector, and half in the banking sector. These figures would be about 15-20 percent of 2000 GDP. Thus, they are higher than in Albania, but lower than in Azerbaijan. This partly reflects harder budget constraints on the real sector, earlier efforts to contain the problem, and lower levels of financing. With higher proportional social fund benefits in the Kyrgyz Republic than in Azerbaijan, this suggests that Azerbaijan's oil and power companies are providing larger subsidies to the economy than found in the Kyrgyz Republic. However, it also suggests that Azerbaijan will need to do a better job of containing the use of its energy sector as a source of major subsidization, and that competitiveness and efficiency will need to be enhanced if these enterprises are going to be used for subsidization. To date, this has not shown itself to be the case. In Moldova, arrears have been a major staple of government and enterprise operations. As of 1998, budgetary arrears were US$417 million, or about 25 percent of GDP. More recent figures have shown mid-2000 wage and pension arrears to approximate US$59 million. End 1999 debt service arrears were US$115 million. Tax arrears were about 15 percent of GDP. All of these, 86 See "Kyrgyz Republic: Selected Issues and Statistical Appendix," IMF Staff Country Report No. 00/13 1, October 2000. 132 along with energy arrears and the substantial inter-enterprise arrears problem that has been netted at distorted prices, have undermined economic growth. Discipline imposed in the Moldovan banking sector has generally separated the overall problem of arrears from the banks, although such structural and payments weakness has virtually removed the potential market for banks to lend. This is reflected in weak funding of banks, and a general lack of creditworthiness of the enterprise sector based on commercial criteria. Meanwhile, the cost of NPLs of the banks represented about US$19-29 million in mid-2000, about 1.8 percent of GDP. This is minor compared to non-bank arrears. On a comparative GDP basis, it appears that Azerbaijan has the highest accumulation of arrears, followed by Moldova, the Kyrgyz Republic, and then Albania. With nearly 50 percent of 2000 GDP in cumulative arrears, Azerbaijan's arrears are Azerbaian has the double GDP-based estimates for Moldova, 2.5 times those of the Kyrgyz Republic, and about 10 times those of highest accumulation oj Albania. arrears, followed by Moldova, the Kyrgyz Republic, and then Albania 133 CHAPTER FIVE: LESSONS FROM THE EU ACCESSION CANDIDATES GENERAL OVERVIEW As noted throughout this paper, there are vast differences between EU accession candidates and the four subject countries. The reason for looking at the former was to derive some lessons of success that would be relevant for the latter. While not an exact peer comparison, lessons from the performance and policy choices of the EU accession candidates can be relevant for other countries. However, These four countries this benefit should not be overstated. In most cases in the CIS and Balkan countries, it will take years and are in thefortuitous sometimes decades to put in place the needed conditions position of being able for sustainable growth, material reductions in poverty, to learn from the competitive enterprises as measured by global standards, EU and soundly functioning markets and institutions. Until mistakes of the that time, it is perfectly conceivable that many of the candidate countries. current problems will remain, including corruption, poor governance, weak public service delivery, low levels of public confidence in civil institutions, weak judicial capacity, de facto protectionist barriers through political and bureaucratic channels, weak management capacity, and insufficient investment in social infrastructure and human capital. Particularly in the case of CIS countries, there is still substantial overhang from the Soviet period which will require time to adjust and adapt, as painfully experienced in the 1990s. Nonetheless, not all is negative. The achievements to date of many countries should not be understated, particularly in light of the difficulties faced, the magnitude of the challenge, and the long-term optimism that many have in the ability of the market to function, despite setbacks. This has been demonstrated in the Russian Federation, where a sufficient number of people benefited in the mid-1990s from increased access to more and better quality goods to realize that their standard of living would eventually improve, even after the setbacks of the ruble and GKO crisis. While the Russian Federation is the key anchor of the CIS, it is not alone in having made progress. All three Baltic states have likewise progressed economically, enough in Estonia that it became a first-tier accession country. Latvia has likewise shown progress since 1995 (triggered by the shock of the Bank Baltija crisis), and Lithuania has followed suit. These three countries are clearly different from our subject countries, having benefited from substantial aid from and links to Scandinavia and the West, and not having been quite as encumbered by seven full decades under socialism. However, they demonstrate that small and open economies that pursue reforms can improve their productivity, penetrate foreign markets, attract substantial financial flows, and put their economies on sustainable growth paths. Again, our four subject countries are different from more successful countries. The profile of the four countries studied is one of lower output and incomes, weaker institutional capacity, and lower aggregate investmnent (possibly apart from Azerbaijan, where investment flows have been 135 limited to the petroleum sector) than found in the EU accession candidates and other success stories. They also were wholly transformed during the Soviet era, or in the case of Albania, by extreme isolation and paranoia. This gutted much of the human capital that existed, and clearly weakened the institutional capacity to govern according to international standards. However, entrepreneurship has surprisingly endured, despite these controlling influences. More recently, discipline in monetary and financial services has been introduced in all four countries after shocks, three of four (with the Kyrgyz Republic being the exception) have access to investment or remittance flows that have helped to provide some underlying cash flow, and all four have a rural sector that provides an informal social safety net. While starting from a weak base, these four countries are in the fortuitous position of being able to learn from the mistakes of the EU accession candidates, and to accelerate the process of reform on the basis of some of these lessons. One of the key lessons is that, for the EU accession candidates, the prospect of EU membership has been an incentive to make difficult policy choices that might otherwise not have been made. For instance, Poland has had a contentious political environment throughout the 1990s, yet membership in NATO and the EU appear to have kept the country on the path of reform. By contrast, such incentives were not sufficient for the Slovak Republic, which later became a second-tier candidate. Over time, it is possible that Albania might be invited to join the EU, although this is not likely for some time. The CIS countries do not The prospect of now have this option. As such, all four countries should focus on EUmembership international norms to achieve competitiveness. Adoption of EU standards, to the extent that this occurs, should be within this larger has been an context. incentive to make While formation of regional alliances may not be feasible beyond difficult policy what already exists (e.g., Stability Pact for Albania, CIS for the choices. other three countries), partial solutions could conceivably come in the form of joint banking supervision, regional capital markets development, and trade, customs and/or currency unions. While currency unions may not be ideal or recommended, commitment to price stability and adequate foreign exchange cover are minimum requirements for stable exchange rates. The alternative, as has evolved in Moldova and Azerbaijan, is increasing dollarization of the economy. Nonetheless, other forms of regional coordination are possible, feasible and recommended. EU accession candidates' efforts to coordinate with the EU in many of these areas can provide useful lessons for replication. Not all five of the EU accession candidates have necessarily reformed quickly in all areas. The Czech Republic has experienced deep problems of governance resulting from cross-ownership and incomplete privatization in the financial and enterprise sectors. Likewise, Slovenia has been slow to privatize and to open its economy to foreign investment in the manufacturing sector. Notwithstanding these problems, both countries have broadly restored their pre-transition levels of output, and are on track to eventually become members of the EU. However, it is plausible to argue that there has been an opportunity cost associated with some of the sluggishness of the reform process in several of the EU accession candidates. It is also plausible to argue that these countries have had a bit more cushion for these opportunity costs, and that the low level of output and incomes in the four subject countries do not afford them the same cushion for mistakes. This has generally resulted in deeper problems of confidence in civil institutions, and 136 less adequate functioning of markets in support of economic growth. For this reason, a key lesson for the less fortunate four subject countries is to bite the bullet, finalize the privatization process, and rationalize the role of government and the public sector in the process with an emphasis on professionalization and improved service delivery. The synchronization and phasing of reforms may be more important than speed. In the Czech Republic, rapid and mass privatization combined with an open investment environment were all driven by sound policy objectives. However, the limitations of bankruptcy procedures and weak accounting and disclosure standards also slowed the pace of enterprise restructuring. This ultimately culminated in balance of payment problems in 1996, partly driven by structural weaknesses, and these The synchronization have persisted to some degree through the banking and andh asio] enterprise sectors since. In retrospect, strengthening andphasing oj incentives for enterprise and bank restructuring early on reforms may be more might have been more useful than rapid privatization important than speed. through vouchers. This certainly seemed to have been the case in Poland, where early reform efforts in the banking sector were later revised to shore up banking supervision, while banks themselves restructured their balance sheets and recapitalized, and as enterprises restructured their debts to banks. Since the mid-1990s, the Polish economy has continued to grow at high real rates, partly as a result of structural reforms implemented after initial "shock therapy" in 1990-91. Stabilizing the macroeconomic framework has been indispensable to restoring investor confidence, domestic and foreign. Slovenia is a good example, having reined in hyperinflation inherited from the Yugoslav period while maintaining fiscal balance. This performance was similar to the approach taken by the Baltic countries, which also inherited hyperinflation from the breakdown of the Soviet system. However, unlike the four subject countries, Slovenia had an export- oriented manufacturing sector in place, and close proximity to regional markets. Nonetheless, Slovenia's relative competitiveness and higher purchasing power would have diminished had macroeconomic stability not been restored. All four countries studied have learned this lesson in part, showing monetary discipline. Efforts to increase fiscal discipline on the consolidated government accounts are currently being put in place, although the four subject countries have not fared as well on this front. Arrears arrangements show leakage and a lack of discipline, and this will slow the ability of the CIS countries studied to achieve competitiveness and sustainable growth. Aggressively addressing structural weaknesses has been necessary to sustain real growth. Here, the EU has served as a catalyst, as failure to address these issues would demonstrate less capacity to assume EU membership obligations. The Slovak Republic is an exarnple of a country that was originally predicted to be first-tier, yet got off track. Part of this had to do with the perception of the EU and others that the Slovak Republic was not adequately attending to its structural weaknesses. Rather, it opted for a high level of political patronage in the privatization process, which had implications for the management of fiscal policy as well. The central bank maintained a disciplined approach, and the Slovak Republic's more recent performance indicates the country has gotten back on track with some of its structural reform agenda. This has been noticeably true in the banking sector. 137 Estonia is an example of a country that has been vigilant in addressing structural weaknesses from the start, achieving high real growth rates from 1995-2000 apart from 1999, which had much to do with exposure to the earlier the Russian Federation crisis. Likewise, Poland has experienced high real growth rates since the mid-1990s, benefiting from an acceleration of structural reform after the earlier experience of shock therapy. This has resulted in 1999 performance being 28 percent above levels achieved one year before the transition process began, by far the best performance among transition countries. Hungary has likewise experienced sustained and consistent levels of high real growth since 1997, reflecting an acceleration of structural reforms resulting from strategic privatization efforts launched in 1995- 96. Slovenia has also experienced consistent real growth in the last several years as most socially-owned enterprises have been privatized, and a disciplined macroeconomic framework has been maintained. The lesson here is that there is a difficult adjustment phase, but gradual approaches extend the difficulties even if this approach is designed to ease the difficulty. Meanwhile, the more deliberate and determined the effort, the greater the likelihood that impressive growth rates can be achieved after this initial "shock" or adjustment. Again, these results for the EU accession candidates may be different for Albania and the CIS countries studied. Nonetheless, this also means the four countries studied have less cushion to work with, and need to generally rationalize the role of the state as There is a they cannot afford a more expansive role. difficult adjustment Part of the effort in building up integrity and confidence at phase, but gradual the structural level has been demonstrating increasing approaches extend institutional capacity. In the financial sector, this has the difficulties . . . emerged from credible capacity in banking supervision, strict listing requirements and disclosure standards in the securities markets, and appropriate regulatory oversight and consumer protection in the contractual savings market. Banking supervision has been a weakness in most transition economies. Poland is probably the best example of a country that has followed a consistent and deliberate strategy to conduct a comprehensive approach to banking supervision, including earlier use of on-site inspections than most countries. Supervision in Poland has become increasingly risk-based over the years, and is now evolving on a more inter-sectoral basis to address issues of potential contagion effects, complex institutions, and consolidated supervision. This approach is in evidence elsewhere, particularly as the EU accession candidates seek to harmonize their legislation, regulations and practices with recommendations from BIS, IOSCO and IAIS. However, most CIS and Balkan countries will need substantial time to develop comparable capacity. That could argue for a more "rules-based" approach to start combined with intensified efforts to develop financial sector institutions that are more representative of the economy and typical scale of enterprises. This has already occurred in Albania, the Kyrgyz Republic and Moldova with micro-finance and small savings and credit institutions. As the scale of these operations is small, the collapse of any individual institution would not have the debilitating effect on the economy that a major deposit-taking institution would have. Albania and Azerbaijan have already experienced pyramid schemes, so supervisors need to be on their guard for such behavior. Likewise, the reputation of all four subject countries for corruption is poor, and money laundering, fraud and other comparable tendencies need to be rooted out. 138 However, it is conceivable that a different prudential framework is in order geared to smaller institutions that pose less risk to financial sector stability. As enterprises and the economy grow, it would then be expected that banks and other financial institutions would grow, and that supervision would become increasingly "risk-based" over time. These concepts are not mutually exclusive, just different in terms of how they approach and manage the complexity of risks. General foreign investment resource flows may serve to obscure deeper structural issues. Hungary long had the highest levels of foreign direct investment in the early to mid-1990s, yet pursued programs that sought to gradually restructure and rehabilitate loss-making state-owned banks and enterprises. To Hungary's credit, it changed course around 1995 with adoption of a revised Privatization Law, and since then has moved forward to establish a more competitive economy. However, it is unclear if Hungary would have pursued as gradual an approach to reform in the early 1990s had it not been able to attract external investment. The Czech Republic likewise has had a favorable investment climate from the beginning, yet ran into structural problems in 1996-97. There is a lesson here for Azerbaijan, which is that structural reforms need to be addressed even if there are high levels of foreign direct investment. While investment flows are helpful, they do not solve all structural problems in the economy. Investment flows FINANCIAL SECTOR LESSONS are helpful, but they The EU accession candidates (apart from Slovenia) have been do not explicitly committed to privatization of banks for the last five automatically solve years. This has been most pronounced in Hungary and Estonia, where each country's banking system has been effectively structuralflaws. privatized and majority-controlled by foreign investors. Poland and the Czech Republic are now following along the same path. These trends have brought each country more developed risk management systems, a broader array of services, expansion into retail lending, higher lending levels, and reduced costs of intermediation. The reduction in lending rates has also been a function of getting portfolio problems under control. From a regulatory standpoint, this has reduced the number of supervisory concerns that exist at the bank level, and permitted monetary and regulatory authorities to consider next step issues and risks such as complex institutions, cross-border coordination, consolidated supervision, and the role of market participants in asserting financial discipline in the governance and management of institutions. While such targets are premature for the four subject countries and prime-rated investment in the banking sector is less likely to materialize, such developments can serve as a road map for the next decade. Albania is on the verge of having a wholly private and majority foreign-controlled banking sector, as has occurred in most of the EU accession candidates (apart from Slovenia). Already, private banks have improved payment and transfer services. Privatization of NCB and Savings Bank will hopefully improve the quality of intermediation at the retail level over time. Combined with micro-finance activity, it is conceivable that Albania will have a functioning and effective banking system in the next decade. Lessons from the EU accession candidates may be particularly important for Azerbaijan, with its two remaining state banks, and IBA in particular, given its potential to attract major investment. 139 Most of the EU accession candidates have been deliberate and phased in their approaches to non-bank development, and in increasing the freedom of banks to be active in these areas. There are exceptions. The Czech Republic permitted a high degree of cross-ownership in the enterprise sector through bank-owned Investment Privatization Funds (IPFs). Likewise, IPFs were permitted to buy into banks. More recently, cross-ownership has been evident in all five countries in the financial services, with banks often having stakes in insurance, securities brokerages, pension funds, and other financial operations. However, their ownership in non- financial enterprises has been limited. Moreover, all five countries have recently (in the last few years) committed themselves to ensuring there is adequate regulatory oversight and supervisory capacity to prevent investments and exposures in non-bank entities from coming back to undermine the financial condition of major banks. Above all, the five EU accession candidates have sought to ensure deposits are safe, and that public confidence is strong. This appears to be the case in all five countries, which suggests that banking fundamentals need to be in place before there is aggressive expansion by banks into non-bank activities. This does not pre-empt bank exposure to leasing, factoring, mortgages and other activities, but it presupposes the existence of adequate systems, controls, and risk management standards to protect against losses and portfolio erosion. Over time, as market infrastructure and transparency develop, banks are expected to increase their participation in these markets. The lesson for all four countries studied is that liberalization of financial systems should be phased to ensure stable conditions prevail, particularly with regard to deposit safety. Liberalization oj financial systems Foreign direct investment from prime-rated institutions hasnys played an important role in most EU accession candidate should be phased to banking systems, enhancing levels of competition and ensure stable providing a far broader array of products and services to the conditions prevail, marketplace. Hungary is probably the best-known example of particularly with rapid movement towards a foreign-owned banking system. reg ard tit Apart from OTP, virtually all major banks in Hungary have regard to deposit attracted strategic investment, usually from one or more large safety global investors. Estonia has followed suit with investment from Scandinavia. Recently, Poland has become majority foreign-owned, as symbolized by the acquisition of Bank Handlowy by Citigroup and PKO SA by UniCredito. The Czech Republic has privatized its major banks, primarily with strategic foreign investment over the last two years, after unsuccessfully restructuring many domestic banks through a consolidation program. It should be noted that the business environment for foreign banks was open and conducive from the beginning of the transition, but that major bank restructuring tended to focus on non-strategic approaches first, as in Hungary and Poland. This was likewise the case in Estonia, although Estonia was always intent on liquidating loss-making banks, consolidating what was salvageable, and ultimately moving forward as quickly as possible with strategic investment to integrate with Scandinavian and EU countries. Only Slovenia has resisted this approach, although it has also maintained an open environment for foreign banks and insurers to operate in that market. 140 The lesson for the four subject countries is that strategic foreign investment is a vehicle for stabilizing and modernizing the banking sector, and increasing confidence and intermediation. However, their Efforts to create an markets are currently not as attractive for strategic investors. open and conducive The one exception should be Azerbaijan with its investment flows, oil wealth, and population base (of 8 million) that is environmentfor roughly comparable to that of Hungary and the Czech investors are Republic (each with 10 million), and four-six times that of necessary, but not Slovenia (2 million) and Estonia (1.5 million). However, apart from HSBC, Azerbaijan has not succeeded in attracting sufficient. prime-rated strategic investment into its banking sector. Even here with HSBC, the focus has been on fee generation, and not in any major balance sheet exposures. The EU accession candidates have strengthened governance and management in banking. This has been driven by regulators and markets, and been done largely on the basis of professionalizing the composition of boards, improving compensation for management and staff, introducing better MIS, enhancing the autonomy of the internal audit function, relying on outside advisory services, and forging closer relations with global institutions. In all five EU accession candidates, the banking system has shown improvement in these areas. In some cases, privatization of former state banks has been a major catalyst. This has included an opportunity to appoint new and more qualified supervisory and management board members, introduce set- aside provisions for employees (e.g., 5 percent of shares) upon the sale of the bank, and attract the needed capital resources for modern systems, management and information processing. Following Basle Committee recommendations, most country regulatory authorities have promoted a more autonomous internal audit function for board members to be better informed of their banks' performance, and to be made more aware of potential risks early on. Beyond this, many of the banks have relied on their external auditors and other specialists to enhance efficiency and competitiveness. All of this has been essential in modernizing principles of governance and management. These have occurred most where prime-rated investment has taken stakes in major banks, and where share listings on domestic exchanges have become more subject to institutional investor scrutiny. These trends can likewise serve as a road map for the four countries studied over the next decade. Market proximity is still important for the development of financial services, notwithstanding electronics and the age of the internet. However, this is because of the contribution made in increases in business volume and purchasing power as an incentive for financial firms to operate, rather than the actual location of specific banks. All five EU accession candidates have benefited from trade and investment flows from neighboring EU countries, both in the real sector and the financial sector. This has helped to integrate economies, boosting the number and value of transactions. While trade with the EU is increasing among the four countries studied, the potential for joint ventures with West European manufacturers is less feasible, export levels are lower, and incomes are generally much lower. For the time being, corporate market development prospects appear thin in most countries apart from Azerbaijan. Until Albania, Moldova and the Kyrgyz Republic are able to improve their economic prospects, modernization of their financial sectors may take longer. This suggests that laudable efforts to create an open 141 and conducive environment for investors are necessary, but not sufficient. This has shown itself to be the case in Moldova's insurance sector. In the five EU accession candidates, laws and regulations have varied, yet increasingly converged with EU norms and international standards. The candidate countries have done so in response to EU incentives, but also because they have become more closely integrated through trade and investment flows with the EU, and more broadly with the global economy. This appears to be a consensus position among all countries studied, as laws and regulations consistent with principles of market-based banking are broadly in place. Where there is a gap is in institutional capacity and market development. In some cases, restrictions on foreign investment are anachronistic and counter-productive, such as in Azerbaijan. Bankruptcy and liquidation have been a problem in most countries, although Estonia's willingness to use liquidation as a tool of consolidation for financial sector development is an exception. Hungary sought to address this issue early on in 1992, with automatic triggers. However, enterprises that were in default on their debts to banks managed to strip assets through deft use of holding companies. As cases automatically went to court, the courts were quickly overwhelmed. In general, judicial incapacity has been a challenge, leading to time- consuming adjudication that has often been unconstructive to creditors. This suggests that Estonia's approach might actually be most useful for the four countries studied. However, learning from Hungary and other countries, bankruptcy and liquidation need to be treated on a consolidated basis. This requires more than just clear laws and regulations. It also requires better property registries and financial information than currently exist in the four countries. Land privatization remains a sensitive issue in the candidate countries that may be easier to settle elsewhere. However, in general, land privatization has been slow to develop as a major basis for effective and sustainable secured lending. This has been a sensitive issue in all countries, with its own peculiarities country by country, as well as within the broader context of EU negotiations. This may be an issue that could be more readily dealt with in the four countries studied than in the EU accession candidates. This is partly due to restitution-related issues that could emerge in some of the EU accession candidates that are cross-border in nature. Among the four countries studied, land privatization and the use of land as collateral for loans appears to be held up less by cross-border issues, and more by tradition and culture. For Lnoand instance, the effective seizure of housing is generally Land privat ization and unacceptable in most cultures. Leasehold arrangements have the use of land as permitted investment from abroad in agriculture, and all the collateralfor loans EU accession candidates have promoted investment in urban appears to be held up land/premises. However, the high cost of real estate and premises in the five candidate countries points to distortions less by cross-border in the market. The limited development of the real estate and issues, and more by mortgage market in the five countries also shows that tradition and culture. property ownership issues remain sensitive. The EU accession candidates still show a need for micro- enterprise and SME financing, notwithstanding the amount of donor support already provided. Poland has recently addressed this with a credit union movement that may be an effective substitute for traditional donor and rural finance efforts. Poland still has two major state banks, 142 one of which (BGZ) is traditionally linked to the farrn sector and agricultural cooperatives. While it has been strengthened in recent years, this has been at cost to the system in terms of regulatory forbearance, the deposit insurance fund, and the possibility of deterring investment into the sector from abroad. There has also been a supervisory cost, as banking supervision has a special department dedicated to these banks that, together, account for only 5 percent of total banking system assets. Nonetheless, a credit union movement has emerged since the mid-1990s that is complementary to the banking system, and in some ways presents a potential substitute for the cooperative banks. With proper management, investment policy, and systems in place, this could well serve as a basis for other community-based financing needs in rural and urban areas. Given that the EU accession candidates still have substantial portions of their enterprises and work forces that are unable to qualify for bank financing, this model could help address some of these constraints. Such efforts would be something similar to some of the savings and credit activities found in the Kyrgyz Republic and Moldova. These efforts should be supported, as should micro-finance in Albania. However, they should also be supervised to guard against their misuse. Accounting standards remain two-tiered in the EU accession candidates, and building up local professional capacity will take time. All the EU accession candidates have intemational accounting firms playing an active role. However, this is primarily for banks, large financial services firms, multinationals, and domestic blue-chip firms. There is still a gap in the real sector, and this has made banks reluctant to lend aggressively to SMEs, or for investors to commit equity through OTCs. This suggests that developing such capacity is a stubborn challenge that takes many years. Considering the usefulness of sound accounting standards and practices for financial sector competitiveness, professional development of local capacity in the accounting and audit fields should be priority areas for development in the four countries studied. Constraints in this field drive up costs and slow market development. However, as with other institution-building efforts, this should be viewed as a long-term effort. Banking supervision has proceeded along differing paths in the EU accession candidates, with Poland generally recognized as having the most developed institutional capacity. In Poland, the National Bank undertook a deliberate program of capacity building in 1992-95 to coincide and be coordinated with efforts to restructure banks and problem debt(or)s. This was largely in response to an insufficient prudential framework introduced during the first round of banking sector reforms that permitted a large number of undercapitalized private banks to compete. The same problems emerged in CIS countries, although banking supervision has not developed as thoroughly or been as effective. Since 1995, Poland has carried on with its capacity building, and is increasingly focused on risk-based supervision. Shortcomings until recently have included the absence of consolidated supervision, although this is expected to change soon. Supervision in Poland has required banks to comply with strict prudential requirements that have shored up bank capital and liquidity, served as a catalyst for banks to develop modern infornation systems, and strengthened Poland's reputation among the international financial community as a serious banking market. This model could be used as a road map for the next decade in the four countries. Albania could also look at the strengths and weaknesses of Hungary's approach, given the degree of foreign ownership in the Hungarian banking system. The EU accession candidates have all tightened up on capital calculations to account for earlier weaknesses in accounting information, loan classification and provisioning practices, and the 143 over-statement of earnings and capital. This has included a tightening of some of the risk weights applied in capital adequacy ratios in an effort to obtain a more accurate picture of underlying capital. Several of the supervisory agencies have factored in earlier weaknesses of collateral, asset valuation, off-balance sheet items, and market risk assumptions to better gauge capital adequacy. This has been Small, "pocket" reinforced by the entry of major global financial players, the interest of portfolio investors from abroad in bank stocks, and the banks are general role played by institutional investors. Here, the fusion of dangerous to securities markets with the banking sector is important. In all five financial sector countries, bank stocks are a major portion of market capitalization. . This has provided them with access to additional financing (debt stablity. and equity) that is complementary to the syndicated market they have been able to access. Part of the reason is the governance benefits that have come from market listings. Combined with the initiative of bank regulators in having a more prudent assessment of regulatory capital than earlier applied, this has increased market confidence and bolstered bank equity values. The lesson here for the four countries studied is that there are long-term benefits to prudence. However, the four subject countries clearly lack the opportunities enjoyed by the EU accession candidates. Thus, strict standards for banks are needed, but less rigorous standards could be applied to non-deposit-taking and micro-finance institutions to stimulate basic intermediation at the household and micro-enterprise level. Slovenia has pursued a fairly unique approach with regard to minimum capital, showing the importance of exceeding EU and international standards as a sound basis for economic growth and development. Meanwhile, Estonia has pursued a useful consolidation approach that has helped achieve critical mass in bank capital where it might otherwise have been lacking if a large number of banks had remained in operation. Slovenia introduced comparatively high minimum capital requirements in 1995 to DM 60 million, at the time about US$45 million. This is about 10 times the EU minimum of Euro 5 million. This triggered consolidation within Slovenia as part of the general five-year bank rehabilitation program (1993-97). Estonia has likewise pursued a conscious strategy of reducing the number of banks, and establishing a firm bank-based foundation for universal banking. Both provide lessons to the countries studied, above all in the CIS countries. The examples point to the importance of consolidation for a critical mass of resources for meaningful levels of intermediation. In the case of Slovenia, this required comparatively high levels of minimum capital. In Estonia, this addressed the inability of many of the banks that were remnants of the old Gosbank system to meet more modest capital levels. Ultimately, in the case of Estonia, Swedish investment entered because there was a perception of enough of a market to commit to. Beyond that, the shrinking number of banks also sent a signal that the authorities were not going to tolerate the traditional "pocket" bank behavior that prevailed in so many CIS and other transition countries. While Albania, the Kyrgyz Republic and Moldova do not have a large number of banks, most are very small and have low capital. In Azerbaijan, this is also true, but with many more banks than are needed (despite the large number of banks that have already gone out of existence). Apart from Albania, the countries studied need to put serious focus on bank consolidation, recognizing that small pocket banks are dangerous to financial sector stability. 144 Deposit insurance in the candidate countries has generally followed EU directives, which leave a fair amount of local discretion in terms of coverage and fee assessments. The effort to introduce deposit insurance schemes has been phased in the EU accession candidates along with improving supervision, the demonstrated ability of the banks to comply with regulations, and the gradual consolidation of the system. These have been in place since the early to mid-1990s, and have evolved as an improvement of the earlier implicit state guarantee. However, costly recapitalization and restructuring needs in the banking systems have also served as an incentive for better management and supervision. Major bailouts have an economic cost, and most of the five EU accession candidates have had to marshal their resources for other needs. The role of the market in shifting deposits from bad banks to good has been important. Hungary is a good example, particularly during the run on Postbank in 1997. Hungary permitted the market to function, with other bank stocks falling for a few days on the Budapest Stock Exchange, followed by recovery. In the meantime, deposits shifted from Postbank to good banks, and the system strengthened as a result. The deposit safety issue was also reinforced by better management at OTP, the former savings bank, than found in many other savings banks in transition countries. Regular reporting and disclosure to the public, stiffening competition for deposits, a general decline in deposit concentration, and the prompt closure of unsalvageable banks (or their negative assets) have generally reinforced public confidence in recent years because of the underlying strength of other institutions subject to competition. While this has all benefited the candidate countries, deposit insurance is not a viable option at the moment for the CIS countries. In fact, the existence of a deposit insurance scheme often provides a false sense of security to depositors and potential for moral hazard among bank managers. Numerous prerequisites are needed for deposit insurance to be feasible, including governance, accountability, accurate and developed information systems, professional liquidity management practices, resolution experience (or at least a viable resolution framework), and market capacity for purchase and assumption. These do not exist in the CIS countries. Albania will need to manage expectations as it considers movement forward on the introduction of deposit insurance. Modernization of payment and settlement systems has helped the EU accession candidates make their banking systems more efficient. This has improved tools available for liquidity management, eased concerns about account confidentiality (particularly as the fiscal function of banks has diminished), and made it more possible for the five candidate countries to make an effective contribution to international efforts to contain money The existence of a laundering and other financial crimes. Meanwhile, development deposit insurance of large value payment systems and RTGS have served as scheme often provides another factor in large investors making a commitment to the EU accession candidates, particularly Poland, Hungary and the a false sense of Czech Republic. Improvements in financial sector infrastructure security to depositors, have also been beneficial in the functioning of the capital and the potentialfor markets of these countries. Improvements in the four countries studied will add convenience and efficiency, and this will moral hazard among provide opportunities for banks to increase their resource base bank managers. and intermediation capacity. 145 Securities supervision has likewise evolved differently, although all five EU accession candidates have aspired to achieve viable main exchanges for larger firms while having parallel exchanges for smaller firms. All five countries have reasonable accounting and disclosure requirements, capital standards, board and management tests, and (self-) regulatory capacity for investors to have confidence in the functioning of the main exchanges. There have clearly been problems of extraordinary volatility in the prices of some main market stocks, but the markets generally have mechanisms to freeze trading should such volatility reach unacceptable levels. More often than not, problems in the quality of the markets relate to inadequate accounting information provided by companies, or in the non-transparency of trading on parallel exchanges. Among the four countries studied, there may be the temptation to activate parallel and OTC markets without strict requirements. In fact to date, Azerbaijan and the Kyrgyz Republic have been commendable in maintaining strict standards at the expense of foregone listings. In the long run, the markets will have a better reputation if such standards are maintained. This has helped to sustain reasonably high levels of investment through the EU accession candidates' main exchanges, and has increased liquidity, turnover and market capitalization. In the meantime, the four countries should really focus on more fundamental intermediation needs at the base level of the economy, while improving the business enviromnent for investment and strengthening standards of governance. OTC markets should be supported, but on the basis of transparency and disclosure standards that provide confidence in transactions. The EU accession candidates have been very open about encouraging foreign insurance firms to participate in contractual savings and reinsurance, even where large domestic insurers are present. The introduction of additional competition in the insurance market has provided a more diverse array of products and services, and been timely with the introduction of securities market development and pension reform. The EU accession candidates are well beyond the stage when motor vehicle insurance accounts for most premium revenue. In particular, the five countries have benefited from the presence of major insurers as institutional investors in local exchanges, and providers of life insurance and property and casualty products. Their presence was important for several countries during flooding a couple of years ago (mainly the Czech Republic and Poland), and their growing interest in these countries is a positive signal for market development. The candidate countries are now strengthening regulatory capacity. As recommended for the four subject countries, the candidate countries have generally been in compliance with IAIS-recommended guidelines and EU directives, including the separation of life and non-life, Thefour subject consumer protection standards, investment policy countries need to guidelines, and solvency and liquidity requirements. However, the four countries studied need to attend to attend to fundamentals of safekeeping and governance for the public fundamentals oj to have confidence in financial institutions. This will become particularly important as all countries embark on safekeeping and needed pension reform. governance for the public to have confidence in financial institutions. 146 PRIVATE SECTOR LESSONS Subsidies remain stubbornly high in several of the EU accession candidates, undermining both fiscal performance and corporate governance. Although the EU accession candidates have made impressive progress since the start of the transition, subsidies remain high in the Czech Republic and Hungary, amounting to 7.7 percent of GDP and 4.8 percent of GDP, respectively. Although subsidies are lower in the other three countries, they have not been fully eliminated. The evidence from the WBES is broadly consistent with the macro evidence. In the four countries studied, high fiscal deficits point to the need to bring subsidies further under control. However, given that arrears have been more prominent in the CIS countries, this is more likely the way to focus on the problem. Although the EU accession candidates have made greater progress in enforcing hard budget constraints than the four subject countries did, there is strong evidence that it has been harder to enforce hard budget constraints for state-owned enterprises than for private enterprises, even in these countries. This confirms that, in addition to avoiding actions that soften budget constraints (e.g., netting operations and tolerating tax arrears), governments should see privatization as a necessary step in the process of hardening budget constraints. Telecommunications privatization in several of the EU accession candidates appears to have successfully promoted system expansion and improvement in those countries that have reformed. Although coverage in the telecommunications sector was high in the EU accession candidates by regional standards, coverage tended to increase faster in those countries that introduced private sector participation. This has been important for overall GDP growth, as expansion of the telecommunications sector has enabled other businesses to develop and for transactions to accelerate. Telecommunications capacity has been of particular importance to the banking sector, which has leveraged off of these improvements to modernize and expand retail banking services. This, in turn, has tended to increase funding and intermediation, serving as a building block for the modernization of banking as a whole. Improved capacity has also helped larger enterprises Governments should modernize their treasury management and reporting see privatization as a processes, serving as an internal tool for strategic planning necessary step in the and resource management. process of hardening The currency board arrangement in Estonia successfully budget constraints. reduced inflation, stabilized the exchange rate, and restored macroeconomic stability. In contrast to most other countries that were formerly part of the Soviet Union, enterprises in Estonia did not see macroeconomic instability as a major problem. In fact, by 1998, inflation was even lower in Estonia than in the other EU accession candidates, and enterprises generally rated both inflation and exchange rate instability as lesser concerns than enterprises in the other EU accession candidates.87 Before Estonia implemented the 87 Enterprises in Slovenia rated inflation as a lesser problem than enterprises in Estonia, but exchange rate instability as a greater problem. Enterprises in the other countries rated both inflation and exchange rate instability as greater problems than enterprises in Estonia did. 147 currency board, inflation was far higher in Estonia than in any of the other EU accession candidates. Given the high importance that enterprises in the transition economies that have not restored macroeconomic stability give to inflation and exchange rate stability, Estonia's experience is especially encouraging. However, it can also be noted that Estonia combined tight monetary policy with fast structural reform and hard budget constraints. This has differentiated Estonia's performance from that recorded in CIS countries, where fiscal or quasi-fiscal weakness has undermined the benefits of monetary tightness. Enterprises in the EU accession candidates see high taxes as the largest obstacle to enterprise operations and growth. In each of the EU accession candidates, enterprises rated regulation and taxation as the largest obstacle to enterprises' performance. Among the regulatory constraints, high taxes were rated as the largest obstacle. This emphasizes the importance of taking steps, such as reducing tax arrears and improving tax collection, that will allow countries to reduce tax rates for those enterprises that do pay taxes on all of their revenues. In this regard, CIS countries have fared poorly. Even in the strong institutional environments of the EU accession candidates, foreign-owned enterprises appear to perform better than other enterprises. Even in these countries, where market reforms and institutions have developed more quickly than in the CIS, foreign-owned enterprises usually perform better in a number of areas (including productivity, employment growth and restructuring) than state-owned and private domestic enterprises. This confirms the importance of FDI in the transition economies. Foreign direct investment was a major success story in Hungary in the early years of the transition. Several factors have encouraged the large inflows of FDI to Hungary. First, Hungary worked hard to open its economy to foreign trade and investment. It was the first transition economy in the region to join the WTO, and was the first to negotiate an EU association agreement. By 1994, EBRD (2000) rated Hungary a 4.3 for foreign exchange and trade liberalization - a score indicating it had reached the level of Foreign-owned liberalization typical of an advanced industrial economy. In enterprises comparison, the Czech Republic, Poland and Slovenia did not ape p rf receive similar ratings until 1996. Hungary's cash-based appear to perform privatization program was also designed to encourage foreign better than other investment. These policies resulted in large inflows of capital, with enterprises. Hungary receiving more FDI per capita than any other transition economy between 1993 and 1998. This was also broadly distributed, covering industry, services and the financial sector. The lesson for the four subject countries is to make the environment for FDI as conducive as possible, consistent with standards of fitness and propriety applied to owners. Given the competitive disadvantages of the four subject countries, this will require a long-term strategy. In the meantime, accelerating privatization and intensifying anti-corruption campaigns will help energize the effort. 148 In contrast, 'Slovenia's use of management-employee buyouts and laws that required worker representation on company boards discourage FDI. In contrast to Hungary, desgite a well- trained work force and the highest per capita income among transition countries , there was less FDI in Slovenia between 1993 and 1999 than in any other of the EU accession candidates. This reflects several policy choices. Most notably, Slovenia mainly relied on management- employee buyouts (MEBOs). In addition, until June 2000, when the constitutional court decided that the law was unconstitutional, workers were required to hold at least half of the positions on the supervisory boards of all companies with more than 1,000 employees. This discouraged restructuring and deterred FDI in large companies. The four subject countries are encouraged to use the Hungarian and Estonian privatization models, and to avoid the Slovene approach. Despite its attractive location, Poland's relatively slow progress with large-scale privatization and its failure to completely restore macroeconomic stability deterred FDI in the early years of the transition. At least compared to the other EU accession candidates, Poland made relatively modest progress with large-scale privatization in the early years of the transition. Whereas large-scale privatization was largely The four subject complete in Hungary, the Czech Republic and Estonia by 1995, countries are Poland lagged behind. In addition, although Poland had reduced inflation to less than 100 percent per year by 1991, it encouraged to use remained at over 20 percent per year until 1996. The the Hungarian combination of limited privatization and incomplete and Estonian macroeconomic stabilization deterred FDI in the early years of privatization the transition. Consequently, per capita FDI was lower in Poland than in any of the other EU accession candidates models, and to between 1993 and 1998.89 This pattern reversed with large avoid the Slovene increases in the mid-late 1990s, after inflation had been reduced and privatization had been accelerated. In 1999, Poland approach. received more FDI than Estonia, Hungary or Slovenia and it passed Slovenia with respect to cumulative FDI by 1999. This should encourage the four subject countries to continue to focus on price stability, to reduce fiscal deficits, and to accelerate privatization. 88 In 1999, per capita income in PPP-adjusted international dollars was $15,748 in Slovenia, $12,332 in the Czech Republic (1998), $10,813 in Hungary, $7,980 in Poland, and $7,726 in Estonia (World Bank, World Development Indicators.) 89 Note: FDI increased in 1998 and 1999, allowing Poland to pass Slovenia by the end of the decade. 149 CHAPTER SIX: RECOMMENDATIONS AND CONCLUSIONS FINANCIAL SECTOR DEVELOPMENT RECOMMENDATIONS Table 6.1 Priority Recommendations for Financial Sector Reform Area Recommendations Banking Environment * Strengthen corporate governance and discipline in banks with improved incentives for appropriate behavior * Improve enforcement of business legislation, including, specifically, bankruptcy and creditors' rights laws, to improve financial discipline in the enterprise sector * Strengthen governance and management in the enterprise sector to correct the culture of non-payment Banking Institutions * Encourage prime-rated investment in the banking sector * Develop micro- and SME-financing models as entry-level experience for enterprise borrowers Banking Supervision * Continue to develop analytical skills of supervisors * Phase in risk-based prudential requirements as institutional capacity expands with regulatory authorities and bank boards and management * Implement conservative capital adequacy requirements, reflecting local risks and conditions Non-bank Financial * Proceed cautiously, pending development of the banking sector Institutions and suitable supervisory framework General Banking Environment As an initial matter, it should be emphasized that all four subject countries have made progress in financial sector reform - in many cases significant progress -under exceptionally difficult circumstances. All four have had to deal with hyperinflation in the early 1990s, stabilize their newly introduced currencies, and establish a legal and incentive structure for comtnercial banking after decades of central control. During this period, a satisfactory legal and prudential framework for Al/four countries banking has been developed, banking supervision capacity have made progress has been introduced, the quality of financial information is in financial sector getting better, and bank governance has begun to improve after repeated losses and shocks were recognized as reform under unsustainable. In Albania, there will be no state banks after exceptionally difficult 2001. Only four remain in the CIS countries, in all, circumstances reflecting the privatization and liquidation of nearly 300 institutions in the four subject countries in little more than 151 a decade. Thus, financial sector reform has been anything but a failure in the four subject countries. Having said that, numerous challenges remain to build on progress achieved. General efforts should continue to focus on adherence to international norms and standards, with attention paid to a more representative array of financial institutions that reflect the scale and needs of the economies concemed. In the financial sector, these should converge with EU norms in Albania, and general guidelines offered by the Bank for International Settlements for all four subject countries. As a complement in the real sector, all four countries must reduce corruption to improve the investment climate, and seek to reduce the scope of government interference in licensing, inspections, and other activities that weaken the business environment. At the same time, they have to build human capital, so entrepreneurs will possess the skills to take advantage of the opportunities presented in the newly liberalized business environments. With a more comprehensive improvement in incentives and skills, all four countries would likely improve prospects for financial intermediation and financial sector stability. All four countries should focus on strengthening governance and management in banking institutions for effective financial discipline in the economy. By extension, such standards also need to be applied in the enterprise sector. These are fundamental prerequisites to reversing the culture of non-payment. In banks (and other financial institutions, as well as enterprises), building blocks for improved corporate governance would include protection of minority shareholder rights, and stiffening the qualifications and Al/four countries responsibilities of Board members. The former could be shouldfocus on guided by OECD-recommended principles, whereas the latter strengthening could focus on board members' capacity to conduct a proper steg therninge aassessment of proposed strategic plans, and proper oversight governance and of management performance relative to accepted plans. management in Adequate IT, MIS and an autonomous internal audit function banking institutions are also essential. Albania, the Kyrgyz Republic and Moldova for effective financial are all considered satisfactory with regard to the written laws for effetlveinanlal governing minority investor protection. However, there are discipline in the questions about capacity at the board level. hi all four subject economy countries, the role of internal audit is viewed as more of a rules-based check than a risk-oriented tool to identify potential portfolio problems As part of modern management, proper compensation and performance incentives should be structured and put in place to reward management and employees. This should be performance- based and measured over a multi-year time frame. Incentives should also work in both directions, with limits on compensation increases if performance does not merit reward. One of the best methods is to provide stock to employees. However, broad use of this tool will have to await the development of liquid securities markets, making this a long-term goal as well. These packages should put the onus on boards to ensure their managers are competent from a commercial standpoint, and compliant from a legal and regulatory standpoint. In keeping with efforts to strengthen the institutional environment, judicial reform should include a comprehensive program to train judges, lawyers and other court officers in bankruptcy 152 and secured transactions (based on effective legislation). This is already happening to some degree in Albania. However, given the amount of time required to develop a well functioning judiciary, Albania and the other subject countries should try to accelerate the use of out-of-court mediation and arbitration Any movement processes to resolve disputes in commercial cases. Above all, towards universal strict enforcement of bankruptcy and liquidation measures is needed to create a more creditor-friendly environment, and to banking should be reverse the non-payment culture found in the enterprise sector. phased This should be supported by up-to-date property and mortgage registries, as well as development of modem liquidation practices. As an extension of accounting sector reform, professional valuation services and certification need to be implemented. These are currently underdeveloped in all four subject countries. Anti-corruption efforts should be a major part of the program, along with adequate compensation for court officials to reduce their susceptibility to illicit payments designed to influence their decisions. Efforts to modernize the payment and settlement system are under way and will eventually help to reduce physical risk, reinforce account confidentiality, and make it more attractive for enterprises to use the banking system. Private banks have filled some of the gap to date in Albania, attesting to the benefits of more open competition and foreign investment in banking systems. In the CIS countries, enterprises have largely bypassed banks and the payments system due to the prevalence of arrears, barter and netting. However, deposit mobilization and banking relations have also been stifled by concerns about account confidentiality, and arbitrary garnishing of accounts by fiscal authorities. More efficient systems could alleviate this mistrust while helping banks with their own liquidity management. In addition, more integrated and developed payment systems in all four countries would facilitate branch integration, which would also help banks with accounting systems and information quality. Banking and Related Institutional Issues All four subject countries should actively encourage prime-rated investment into their banking sectors as the fastest and most efficient way to modernize financial systems. Albania has followed this approach, attracting some large banks from Greece and strategic investment for NCB. It is also seeking strategic investment in the Savings Bank. Moldova is likewise seeking prime-rated investment for its last bank privatization. However, the Kyrgyz Republic and Moldova are currently constrained by economic limitations. In contrast, while Azerbaijan is increasingly open to foreign Allfour countries investment in the banking sector, its performance has been poor relative to potential. At the moment, HSBC is the only should actively major global bank active in the market. encourage prime- Given low levels of lending, there may be a temptation to rated investment as move on to non-bank activities or "universal" banking to thefastest and most compensate for the lack of lending. This would be a mistake. efficient way to Efficient banking services are essential underpinnings of all effinaycto other financial activities. Any movement in this direction modernizefinancial should be phased to allow time for suitable regulations and systems 153 expanded supervisory capacity. As part of this effort, licensing requirements should include higher capital requirements for banks and non-banks. Supervisors will need to pay close attention to financial firms on a consolidated basis to ensure that non-bank activity does not impair deposit safety or core banking activities. Organizational, financial and regulatory firewalls should be part of this effort. As all of this is complex and takes time, all four subject countries should continue to focus on underlying stability in the financial system. Where efforts might be energized is more on the smaller end of the financial sector to meet some of the lending needs of households and very small businesses, particularly in rural areas, where poverty tends to be concentrated. Meanwhile, at the larger end, barriers to investment grade insurance companies should be removed in Azerbaijan to accelerate reform, develop capacity, and modernize financial services. Donors and others can assist with efforts to establish a documented track record at the micro- enterprise and SME level of financing to serve as the basis for "graduating" to the banking sector when financing demands reach those levels. This is primarily relevant for the CIS countries, given the predominance of very small enterprises in these countries. This is also relevant for Albania for the same reason, yet a significant amount of micro-finance activity in rural and urban areas has been in place for years, with favorable results. Such a "graduation" approach could likewise become important in Azerbaijan as an extension of credit union development. In Albania, the development of the central bank's credit information bureau provides a vehicle for such information. The long-term challenge for all countries would be to ensure the veracity and standards of such a documented credit performance database to assist enterprises in convincing banks of their credit worthiness once their financing demands exceed the capacity of donor-assisted projects, and once banks are in a position to increase lending. In Albania, the existence of FEFAD and related micro-finance groups in conjunction with the development of a credit information bureau provide an opportunity to integrate these in support of effective intermediation. There is also an opportunity for this to happen in Moldova, through donor-supported village associations. Given the small-scale nature of the four economies, there may be some lessons from micro- finance operations in addressing collateral issues. However, these can not be mechanically extrapolated to larger enterprises, nor can these approaches be considered a substitute for the development of sound secured transactions practices. Cash collateralization and joint and several liability provisions attached to group loans are often applied successfully in micro- finance, particularly in the rural economy and at small-scale levels of lending where groups are bound by social/farnily ties as well as economic incentives. However, it is questionable if such methods can be replicated with larger loans and clients. Cash collateralization is an extension of credit risk management, such as the underlying value of securities pledged or the ability to exercise a guarantee. Joint and several liability is frequently assessed on the basis of cross- holdings, cross-guarantees, and other obligations that could potentially weaken fundamental cash flows to service and repay loans. However, these approaches provide building blocks for small businesses to develop a docunented track record that could subsequently be passed on to larger financial firms as these businesses build equity and assets, and have larger financing needs in the future. 154 Banking Supervision and Regulatory Capital Issues Banking supervision should adhere to appropriate and relevant international norms, including higher than recommended minimum risk weights for capital adequacy measures. Over time, as the overall environment improves and risk management capacity develops, these norms can be revisited. All four subject countries should consider risk weights that are stringent for various assets as a basis for inducing higher capital levels. Exceptions could be made for prime-rated institutions, although this would likely trigger political opposition. The four subject countries should consider movement over time to a system in which banks detail their own assessments of "market value" for capital against regulatory capital for reporting purposes. However, this is a long-term goal, rather than something to be expected any time soon. To be credible, banks will have to demonstrate that Banking supervision their provisioning practices and asset valuation practices are should adhere to credible. This will require more developed accounting and appropriate norms, auditing standards, increased disclosure of timely information, aPPropriate norms, and greater market capacity for evaluation and analysis. In the including high meantime, regulators should interpret BIS recommendations as minimum risk weights the minimum required for financial market stability, and to be for capital adequacy applied on a country-by-country basis contingent on specific for capital adequacyrisk factors. measures Development of risk-based supervision should be viewed as a long-term project. It will take time to build up public confidence, let alone market capacity for modem risk analysis and management. This is true within institutions, and even more so in terms of the overall infrastructure and culture of disclosure. For market-based risk management to be effective, timely and accurate information is needed for investment decision-making. Likewise, a well developed legal tradition and established governance practices need to be in place. These require external Risk-based capacity, as well as internal oversight in the form of autonomous supervision should and skilled internal audit functions. Until information is adequate be viewed as a long- and timely, risk management parameters should be driven by regulatory requirements for financial sector stability. Over time, as term project capacity expands, individual institutional risk-taking can expand without taxing the system or creating undue stress. Thus, financial sector development over the next several years should focus on further development of institutional capacity so that the building blocks for more complex risks are in place. This should serve as a reference point for the phasing of banks' entry into non-bank activities, and the eventual development of universal institutions (as noted above). Savings and credit institutions should be encouraged to develop, but under proper regulation and supervision. This would largely be complementary to the banking system, but subject to high standards of fitness and propriety on the part of owners and managers, and regular reporting and disclosure requirements. Supervisors would need to focus on the potential for these institutions to undermine financial sector stability if they mobilize deposits, such as by promising to pay very high interest rates on deposits or other practices reminiscent of the pyramid schemes. Other countries have shown that this is a risk even after a major crisis occurs 155 and is subsequently resolved.90 They should also not be used by donors to provide loans at market-distorting interest rates. Albania and Azerbaijan have to be on their guard as a result of past pyramid schemes. The Kyrgyz Republic and Moldova have to scrutinize this risk due to the importance of rural finance in the local economy, and due to the presence of a large number of credit unions (Kyrgyz Republic) and village financing associations (Moldova) that could potentially be abused. Moldova has already experienced past risks associated with its close ties to Romania. Non-Bank Financial Services All four subject countries should phase in non-bank services to increase earnings opportunities and financial services, but do so in a manner that reinforces financial sector stability. If a universal banking model evolves, there should be clear separation between bank and brokerage/investment arms, and capital should be measured and supervised on both unconsolidated and consolidated bases, with capital adequacy/solvency tests reflecting the specific risks of each core business activity. At the moment, none of the four subject countries has shown an aggressive intent to develop a universal model. This is prudent under the circumstances, although safe and properly supervised practices should be pernitted. This will be important for governments as they all embark on pension reform, where banks may play critical custodial and investment intermediary roles. The benefits of more services may also help to attract funds currently outside the system, and provide banks with stable earnings opportunities. Allfour countries As the four subject countries move forward eventually with should phase in non- insurance refonn, they should seek to ensure that IAIS bank services to principles are observed. This is particularly important for Albania as it moves to privatize INSIG and liberalize the increase earnings market. Moldova and the Kyrgyz Republic are already open, andfinancial but have underdeveloped insurance markets due to broader services, but in a economic constraints. Azerbaijan should move to eliminate its restrictions on foreign investment in the sector, along with manner that its 30 percent cap on aggregate foreign investment in the reinforcesfinancial banking sector. sector stability Insurance sector and private pension fund development should follow the same general principles as securities market development. There should be general compliance with IAIS- recommended guidelines and, if relevant (such as for Albania in the long run), EU directives in the insurance sector. The latter require a separation of life and non-life. Particular focus should be on the development of the life insurance market, with major institutional investors ideally playing a role. For securities, IOSCO provides adequate guidelines. A key issue for the Kyrgyz 90 Romania experienced problems earlier in the 1990s with pyramid schemes, and only recently began to bring "popular" cooperative banks under the regulatory umbrella of the central bank. Many of these mini-banks were offering rates on deposits that were far higher than those offered by the conmnercial banks, presenting risks to the banking system. 156 Republic and Moldova will be to ensure there is adequate transparency to market trading on their main and OTC exchanges. All subject countries are advancing pension reform. As the four countries eventually move forward with insurance reform (as is currently the case in Albania), the legislative framework might consider the requirements for properly managed private pension funds as well. This would expand contractual savings offerings to the marketplace, and reduce the potential for longer-term fiscal strains resulting from rising social insurance costs/outlays. The benefits of this approach would be more diversified contractual savings options, the introduction of institutional investors to the market, diversification of liquidity management options for banks and companies, reduced reliance on banks for financing, potential for new/increased equity markets, and links to offshore finance and reinsurance companies which might also increase direct investment into these markets. However, this will all take time, as shown in Moldova, where its largest insurance company is foreign-owned. Rather, development of these markets should be deliberate, phased, and based on the development of adequate institutional capacity that reinforces public confidence in the financial viability of firms active in the marketplace. Eventually, mortgage finance could emerge as a major catalyst for banking, securities and contractual savings, as well as a destination for government funds. However, this will take years to develop. Legal and judicial reforms currently in process will need to be fully implemented before the market will take long-term property risk. Meanwhile, the insurance sector will need to refine property and casualty practices and procedures to make this feasible. To the extent that factoring and commercial finance operations emerge, they should be encouraged. However, as risk-takers, their operations should not be permitted to undermine deposit safety. This would mean banks active in these fields would be subject to prudential regulations governing these activities as part of the overall credit risk management function, or that non-banks would not be able to mobilize deposits or access any deposit insurance coverage when deposit guarantee funds become operational. When deposit insurance schemes are introduced, as will occur in 2001 in Albania, the authorities will need to conduct an active communications and education campaign to ensure the public is not misled about what is covered and what is not. Leasing also represents a potential area of development, as well as a risk. In many countries, banks have run up losses from leasing exposure, partly due to problems of borrowers, partly due to the inability to seize assets when borrowers default. All four subject countries will need to monitor these risks and exposures to ensure they do not cross over into the banking sector. To the extent that banks finance leasing operations as part of their core operations, these risks will have to be monitored as part of the normal credit risk evaluation procedure. 157 PRIVATE SECTOR DEVELOPMENT RECOMMENDATIONS Table 6.2 Priority Recommendations for Private Sector Development Area Recommendations Privatization * Accelerate large-scale privatization, particularly infrastructure * Expand the use of privatization techniques that attract strategic, outsider-owners, both foreign and domestic * Avoid methods that result in dispersed ownership Business Environment * Maintain reasonable macroeconomic stability * Reduce levels of corruption to create a business-friendly environment * Focus on the implementation of existing law, even at the expense of the marginal improvement of existing legislation * Improve tax administration to permit lowered rates with enhanced rates of collection, and the broadening of the tax base from regularizing the informal economy * Avoid actions, such as netting operations and accepting in-kind tax payments, which encourage non-cash transactions in the real sector Although the four subject countries have made impressive progress with small-scale privatization, large-scale privatization has proceeded more slowly. Albania has recently made progress in this area, as shown by its privatization of several strategic enterprises in mobile telecommunications (with fixed line in process), chromium and copper mining (via concessions) and refineries. Moldova has likewise privatized several strategic firms recently, namely in cement and utilities. The acceleration of large-scale privatization should be seen as the main priority related to private sector development in the CIS countries, especially in Azerbaijan. In addition to tackling the remaining industrial enterprises, programs should be also expanded to include "strategic" enterprises. In Azerbaijan, the privatization of the large state-owned oil company, SOCAR, could have a significant impact on economic performance. It is recognized that Albania and the CIS countries have comparatively The acceleration of small markets, low purchasing power, weak reputations, large-scale privatization and are competing for investment capital in an increasingly open and competitive international environment. should be seen as the Nonetheless, recent successes in Albania and Moldova main priority related to show investment can be attracted even in difficult private sector environments. For this reason, residual concerns in countries about generating low levels of proceeds for development assets, a national sense of franchise value for "flagship" enterprises, and maintaining employment networks during periods of economic transition should not serve as constraints on policy makers to move aggressively on the privatization front. Economies overall should benefit from the influx of investment and the cost-savings generated from additional efficiencies and services. 158 Employee-owned enterprises and enterprises with dispersed individual ownership appear to perform poorly, especially in the weak institutional environments common in the subject countries. Although the four subject countries have tended to use privatization methods that favor insiders, it is important to recognize the severe constraints facing these countries. Although promoting outsider ownership might be possible for some enterprises (e.g., utilities), it is unlikely that this could be relied on in the majority of cases. Since most forms of private ownership appear to improve enterprise performnance relative to continued state ownership, the main goal should be rapid privatization of the enterprises that remain state-owned. Infrastructure privatization should also be seen as a major priority. This was especially true in Albania two years ago, where enterprise managers saw the poor quality of infrastructure as a moderate constraint on enterprise operations and growth. Since then, Albania has made significant progress, particularly in the field of telecommunications. Accelerating reform and encouraging privatization will have large long-term benefits in all four countries over time. Further, infrastructure privatization is easier to tackle than more intractable problems such as improving the performance of the legal system or reducing political instability. Infrastructure privatization is especially likely to be beneficial in the telecommunications sector, where there have been persistent delays in several of the subject countries. This will also help to improve internet connectivity, which is low Infrastructure even by regional standards in the four subject countries. privatization Although privatization to outsiders will not always be possible, should also be privatization methods that promote concentrated outsider ownership and foreign participation should be favored over other seen as a major methods when possible. In these cases, governments of the four priority subject countries should try to use techniques, such as cash auctions, that favor insiders less strongly and should take other steps that will increase strategic investment. In particular, obstacles to foreign participation in the privatization process should be removed. For example, in Azerbaijan, the requirement that foreigners purchase options from the government before they can participate in voucher auctions is likely to deter foreign participation and should be ended. Although governments in the four subject countries have been considering infrastructure privatization for several years, little progress has been made in the CIS countries outside of the power sector in Moldova. (Albania has fared better than the CIS countries in the last two years, as noted above.) One reason for this is that overly optimistic expectations about enterprise valuations have led to tenders being delayed or cancelled in several countries. For example, one of the reasons why the privatization of Moldtelecom was delayed was due to concerns about a lack of interest given an unfavorable investment climate. However, it is very unclear whether the investment climate, especially in the telecommunications sector, will improve in the near term, or even in the medium term. Therefore, delays are likely to be costly. On the other hand, the situation in the global marketplace has shifted, and it may not be the right time to sell a telecommunications firm now due to retrenchment, consolidation, and financial restructuring at many of the acquiring firms that have invested in markets such as Moldova's. However, earlier delays reflect a missed opportunity in a country that sorely needs enhanced FDI and telecommunications capacity. 159 Governments should avoid actions, such as netting operations and the toleration of tax arrears, which encourage non-cash transactions and allow enterprises to have soft budget constraints. Although direct subsidies are relatively low, inter-enterprise arrears, wage arrears, and non-cash payments are more common in the subject countries, with the possible exception of Albania, than they are in the EU accession candidates. Since these practices soften budget constraints and obscure market transactions, making it more difficult to assess both taxes and credit worthiness, governments should discourage them. In addition to improving enforcement of bankruptcy proceedings, governments should also avoid netting operations and the toleration of tax arrears. If enterprises anticipate that the government will eventually step in with a clearing operation, they will be less willing to stop supplying customers who fail to pay them, and will be more willing to build arrears to their own suppliers and Governments employees. The links between soft budget constraints and state should avoid ownership further stresses the importance of accelerating the netting operations privatization process for large enterprises. and toleration of Enterprises in the two countries that have failed to restore tax arrears that macroeconomic stability, the Kyrgyz Republic and Moldova, rated encourage non- inflation as the largest obstacle to their operations and growth. Since there are also severe microeconomic constraints on cash transactions enterprise performance, this stresses the constraint imposed by and allow even moderate levels of macroeconomic instability. Although enterprises to have reducing the regulatory burden on enterprises, improving competition policy, and strengthening domestic financial systems soft budget might help to reduce inflationary pressures, efforts to reduce constraints inflation will not be successful without appropriate macroeconomic policies. In particular, it is important that these countries reduce their large fiscal deficits, improve revenue collection, and keep monetary policy tight. In the other two countries, Albania and Azerbaijan, where inflation is already low, this stresses the importance of maintaining macroeconomic stability. Albania's success in reducing inflation after the collapse of the pyramid schemes is particularly impressive in this respect. Of relevance as well has been its stable exchange rate throughout much of the 1990s, in conjunction with an open current and capital account. The Kyrgyz Republic in particular can benefit from lessons in Albania, as it found itself highly sensitive to exchange rate depreciation as a result of the Russian crisis in 1998. In contrast to the Kyrgyz Republic and Moldova, where enterprises saw macroeconomic instability as the main obstacle to enterprise operations and growth, enterprise managers in Albania saw crime and political instability as the largest problems in this country. Although the government of Albania has made impressive progress over the past few years in reducing crime, the government needs to continue with these policies. 160 Reducing corruption is important in all the subject countries, but is most important in Azerbaijan, where corruption appears to be both the most pervasive and the most disorganized. Although it is always difficult to reduce corruption, there are several steps that governments can take. First, governments can strengthen internal procedures and institutions to combat corruption and put anti-corruption watchdog bodies in place. These bodies can be effective if they are insulated from political interference, given proper incentives to discourage corruption, and given the power to prosecute public officials who violate anti-corruption laws. Second, strengthening the judiciary and other public institutions and procedures can have large benefits. Possible actions include establishing independent ombudsman offices, publicly auditing government bureaus, and putting transparent budget processes in place. These actions reduce the likelihood of corruption, and make it easier to discover corruption when it occurs. Finally, the strong correlation between corruption and the burden imposed by regulation and taxation suggests that The strong correlation simplifying tax and regulatory procedures would be useful. between corruption and the burden Overall, public confidence in public/civil institutions has a beneficial effect on the economy at large. In all imposed by regulation fairness, the CIS countries have all endured some and taxation suggests measure of cross-border conflict, most notably in that simplifying tax Azerbaijan and Moldova. Such disputes have delayed reforms (Azerbaijan) and development (Moldova). This and regulatory has complicated the environment for policy reform and procedures would be improvements in the business environment, and has clearly added to the political risk of investment in those useful markets. The four subject countries have already made significant progress bringing commercial law, laws protecting shareholders, and laws protecting creditors close to "international standards." Although further progress along these lines would be useful, in all of the subject countries, legal effectiveness (i.e., enforcement of laws) appear to be a far greater problem than legal extensiveness (i.e., how close existing laws are to international standards). Since legal effectiveness is a better predictor of capital markets development than measures of legal extensiveness, this suggests more effort should be placed on improving enforcement, rather than on simply rewriting existing laws. Several steps are being taken to do this in Albania by strengthening legal and judicial education, encouraging mediation and arbitration to settle commercial disputes, and improving dissemination of up-to-date legal information. Given the relatively low ratings most of the other the four subject countries received for legal effectiveness, similar efforts would be useful in these countries as well. Although enterprises in all of the four subject countries rated high tax rates as the largest constraint on enterprise operations and growth related to taxation and legislation, this should not imply that the governments should enact across-the-board tax cuts. The weak fiscal situation in the four subject countries makes across-the-board tax cuts highly unattractive, especially since this would undermine macroeconomic performance. However, there are several steps that governments could take to reduce the burden of taxation, without undermining macroeconomic stability or deepening fiscal deficits. Most notably, if governments succeeded in improving 161 collection efforts, especially in the informal sector, reducing tax arrears, and improving compliance - all of which are problems in the four subject countries - tax rates could be reduced without compromising fiscal performance. The key here is simplification of administration and collections, a broadening of the fiscal base to permit lower rates, the assurance that taxes paid actually wind up in the treasury, and that public expenditure is properly and prudently managed so that households and enterprises believe in the fairness and viability of fiscal policy and institutions. Since foreign and domestically owned enterprises generally had similar concerns about the business enviromnent, steps to improve the business environment will both improve the operations of domestically owned enterprises and boost foreign investment. This emphasizes the fact that governments working on improving the business environment for domestic enterprises (e.g., improving macroeconomic stability) are also important for boosting FDI. 162 CHAPTER SEVEN: OVERARCHING THEMES OF THE TRANSITION EXPERIENCE Several themes have emerged from this study that could profitably inform future efforts to shape structural reform strategies in other transition countries and, indeed, in all developing countries. In addition, a number of alternative approaches have suggested themselves for some of the more intractable problems that have emerged from a close study of the successes and failures of the four subject countries. These are summarized below. Selectivity Even a cursory review of the findings set out in this study on the financial and enterprise sectors of the four subject countries indicates how important the issue of selectivity will be in shaping their future reform strategies. There are literally hundreds of policy, implementation and institutional actions identified in these two sectors, without considering the equally complex and daunting problems in all the other sectors of their economies. This challenge is compounded in the four subject countries by the lack of capacity to implement these reforms. Even in the most advanced political and economic systems, it is unrealistic to expect fundamental change in more than two or three areas in a given year. It is unrealistic As a result, the first challenge of shaping the reform strategy for the to expect near and medium-term will be to decide what is critical, what is merely important, and what can (or will just have to) wait. fundamental change in more In this regard, selectivity has a number of dimensions. First, there than two or must be a decision as to what sectors of the economy are the most promising and the most likely to generate the greatest positive three areas in a response to reform. It is risky to be recommending this approach, given year given the failed efforts at the "industrial strategies" which were so popular in developing countries in the 1970s and early 1980s. Nevertheless, there is not sufficient capacity to do everything in all sectors at once. The effort to do so dooms even promising initiatives to failure. Second, reform must be phased, so that change moves from the most basic to the more complex, without skipping intermediate steps. One good example is housing finance. It is premature to pursue a Western-style housing finance model before the banking sector, the securities markets, land markets, land use regulations and the social welfare system as it relates to housing are reasonably in place. Similar arguments can be made about expending energy on securities markets while the banking system cannot support them. Third, difficult judgments must be made as to the political economy of alternative reform possibilities. If land privatization is simply not achievable, and alternatives cannot be found to unlock the collateral value of real estate, policy makers must consider carefully the utility of agricultural and rural reform. Does this mean those reforms are not important? It does not. It simply recognizes that it does no one good to fail at reform, no matter how important reform may be. 163 Critical Mass and Regional Association There is little chance of the four subject countries becoming competitive and commercially viable on their own, in the long run. All of them face a multitude of weaknesses that place them in a poor competitive position by global standards. Examples of international competitive advantage are limited and vulnerable. For instance, the Kyrgyz Republic is often cited for the potential of its gold sector, yet the output and earnings from this source are expected to decline by 2004 and virtually terminate by 2008.91 Azerbaijan is probably the lone exception with its petroleum resources. However, there is little evidence to date that Azerbaij an's oil wealth has effectively benefited other sectors of the economy, aside from construction, a shortcoming which is common in other countries with similar resource endowments in an unbalanced economy.92 More positively, all four countries are fortunate to have viable rural sectors as a social safety net. Albania has benefited as well from high remittance flows from family members working abroad. At a minimumn, this provides some cushion that other countries sometimes lack. All four countries also have trade ties in neighboring countries that offer the potential to increase economic growth prospects over time. This is particularly the case for Albania, where the number of Albanians in Italy, Greece and other European (EU and non-EU) countries is high. Albania has already benefited from this relationship, as shown in the high remittance flows that have served as the basis for much of the equity/cash-based investment in Albania in recent years. Moldova could potentially benefit in a similar manner later in the decade, depending on Romania's progress in negotiations with the EU. Moldovan workers are currently welcome in the Russian Federation, and would presumably benefit in Romania if the latter made progress with its reform efforts as part of the EU accession effort. Azerbaijan likewise could benefit from stability and growth in Central Asia. As an example, the Azeri population in Iran is reported to be larger than the 8 million in Azerbaijan. Only the Kyrgyz Republic appears to have little cordiality in its relations with larger neighbors, the latter of which have repeatedly put up trade barriers to stem the flow of cheaper Kyrgyz Republic goods into their domestic markets. 91 This is separate from a discussion of the long-term fundamentals of gold itself, which many analysts believe are weak and in secular decline (as with silver and other metals in the past). 92 There are many examples of mismanagement elsewhere around the globe where oil wealth has been squandered, balanced economies undermined, and incentives imprudently structured so as to create other problems that become difficult and costly to solve. Examples include (i) waste and mismanagement in Nigeria, which happened to be a major agricultural producer in the 1960s; (ii) excess centralization of government expenditure in Mexico which led to Mexico City's population expanding to unsustainable levels; and (iii) overly generous benefits in Saudi Arabia and other Gulf states that now face fiscal deficits and are concerned about such social expectations stifling private sector initiative. 164 One of the key lessons from the five "first-wave" EU accession candidates is that the very prospect of EU membership can be an incentive and a road map to identify difficult policy choices that might otherwise not have been made. The same logic may apply to the four countries studied, as well as other countries in the Balkans and CIS. Promoting efforts to establish larger trade, customs, and even currency unions might help to enlarge markets for the limited areas in which these countries have competitive prospects. While the currency issue would be more complex, forming trade and customs areas could increase the flow of goods, reduce transactions Promoting larger trade, costs and barriers that are conducive to rent-seeking behavior, and provide opportunities for economies of customs, and even scale and rationalization. Movement in this direction currency unions might would require agreed criteria and joint monitoring and help enlarge markets supervision. More importantly, these regional efforts for the limited areas in could parallel norms adopted in other free trade areas, with an eye towards eventual linkage over time. For which these countries instance, efforts by Albania with other Balkan markets have competitive should be structured to enhance legal trade that already prospects flows through the region, largely on an informal basis. The Role of Government, Arrears, and Soft Budget Constraints. Disciplined public sector management based on professional standards is essential in restoring public confidence in civil institutions, building a viable fiscal base, and creating an environment that is conducive to a competitive private sector. This relates to the selectivity theme, and requires a conceptualization of the role of government, what its priorities need to be, what is Disciplinedpublc . feasible from a financing standpoint, and areas where Disciplined public government should not encroach. Thus, one of the core sector management challenges of the transition is to manage the role of the is essential in government down from the all-encompassing and dominating restoring public presence of the past to the role of facilitator, rule-maker and confidence, building umpire of the market economy. a viablefiscal base, This point touches directly on the issue of arrears and soft budget constraints, which have been common in Azerbaijan, and creating an Kyrgyz Republic and Moldova. Disciplined management of environment governnent spending and policies that limit the accumulation of conducive to a enterprise arrears are critical to the installation of an incentive system that forces enterprises to adjust to the new conditions competitive private around them. In recent years, only Moldova has brought its sector fiscal deficits down to sustainable levels of about 3 percent. Azerbaijan has shown some progress, yet fiscal deficits have still been in the 4-5 percent range since 1998, with 2000 rates projected to be nearly 8 percent. However, growing arrears have raised questions about how sustainable past reductions in fiscal deficits will be. Albania and the Kyrgyz Republic have fared poorly on fiscal accounts, although both are targeting deficit reduction from 2000 on. 165 Arrears appear to be less of a problem for these countries, although the Kyrgyz Republic has received a great deal more concessional aid to compensate. Nonetheless, that arrears persist can also be interpreted as symptomatic of incomplete fiscal discipline, manifested in the form of fiscal and quasi-fiscal deficits that are a major proportion of overall GDP. This erodes confidence in government and civil institutions, reduces incentives for households and firms to be tax-compliant, and fails people with some of their critical savings and health needs at times when they are most vulnerable. Arrears have been found to clear relatively automatically once discipline is in place and supported by improved macroeconomic fundamentals. Thus, the actual exercise of clearing arrears through netting arrangements is generally not effective as it does not solve underlying structural weaknesses that trigger the arrears in the first place. All governments continue to have a dramatic impact on their economies with their spending policies. The main challenges are constructing a representative and fair list of expenditure priorities, developing the capacity to deliver needed services efficiently, and operating within a reasonable fiscal envelope while addressing long-term structural issues (e.g., pension reform) over time. In this regard, there is a clear role for donors to work closely in partnership with governments and civil society at large. This includes NGOs, business associations, and other parties at all levels. The role of donors should be conceptualized more as advisors, providing advice and counsel on what the return on expenditure would be, based on differing scenarios. At this point, there is enough information and analysis available for donors to work with recipients to prioritize. While there is always a risk that excess borrower control will be overly politicized and used for patronage purposes, broadening the discussion to include non-governmental groups should help dilute the risk. Privatization and Foreign Investment After more than 10 years of the transition, marked by an almost compete absence of capital investment, technological renewal or even maintenance, there is little hope for most remaining state enterprises to be privatized unless virtually given away. There are few major state enterprises in the four countries that should be viewed as "strategic" based on international competitive norms. More significantly, the retention of state interests, even minority interests, in enterprises presents an ongoing distraction from the more rationalized and suitable role of government in an economy characterized by private sector competition. Furthermore, continued intervention by the state perpetuates the risk of anti-competitive interference by bureaucrats in private business, in addition to the potential for future calls on overstressed state budgets to fund direct and indirect subsidies, forgiveness of taxes and other assistance to protect state "investments." In general, Albania appears to have recognized this lesson, while the CIS countries have shown varied degrees of willingness to divest their strategic enterprises. Beyond the direct costs and distractions is an equally important opportunity cost. By now, it is abundantly clear that private enterprises are more efficient and more profitable than state-owned firms in virtually all circumstances. By deferring and deterring privatization, these countries lose the potential for production and growth locked away in state enterprises, potential they can ill-afford to forego in light of their circumstances. 166 Thus, large-scale privatization, including the privatization of infrastructure, needs to be accelerated, especially in Azerbaijan. Privatization methods that promote concentrated outsider ownership and foreign participation should be favored over other methods. In the countries that have started to focus on cash privatization as their main form of privatization, overly optimistic assessments of enterprise value should not be allowed to slow the privatization process. Privatization of land is a critical goal to unlock perhaps the single most valuable source of wealth still remaining in public ownership. As an example, the Kyrgyz Republic has already shown some progress with agricultural leaseholds, and similar efforts to ensure title and collateral rights should be continued elsewhere as a stimulus to rural development and finance. The disposition of residual enterprise interests held by the state - leftovers from previous privatization programs - should be accelerated, even if that means the simple cancellation and surrender of state Privatization methods shares. The best approach for the four governments to take, that promote as Albania appears to be doing, is to divest the lower value assets as promptly as possible while mapping out a clear concentrated outsider timeline with transparent procedures for enterprises that ownership and carry some residual value. For enterprises and assets that foreign participation nobody wants, liquidation rather than restoration is the should befavored recommended path. over other methods Given the limited resources of the domestic business community, foreign investment is critical to turn around these economies. Azerbaijan's oil and gas sector is the best example among the four countries of a sector potentially attractive to foreign investment. Azerbaijan's banking sector may also be attractive at some point for foreign investment, although this is currently limited to one bank (net of new start-ups). Telecommunications opportunities may be limited in potential value due to the amount of investment required, the potential resistance to liberalized tariff structures, and current market sluggishness and efforts to divest and rationalize to reduce debt. Electricity utilities offer opportunities, but market prospects are also hampered by resistance to liberalized tariff structures, as well as traditional subsidization of rates, a common culture of nonpayment, the virtual non-existence of local capital markets, and heretofore limited interest in foreign markets for securities of utilities companies in these countries. Moldova is an exception on this front, having recently attracted Spanish investment into the sector. However, the success of this investment will require people to pay their bills, and for tariffs to realistically reflect costs. Albania and Moldova have made some progress in recent years in attracting foreign investment. Azerbaijan has shown little success outside the petroleum sector, and the Kyrgyz Republic has had difficulty apart from gold mining. It is possible that strategic investors could be encouraged to invest, as has occurred in each of the four countries in at least one or two sectors. However, important as this investment is, the prospects are generally limited. Albania just reached its best FDI level ever, equivalent to US$143 million, in 2000, or about US$40 per capita. Apart from Azerbaijan, the countries studied are not likely to attract much more than that on average for the foreseeable future, which means that total GDP and purchasing power will not be influenced dramatically any time soon. However, rationalizing the role of the state sector, providing a favorable business environment, reducing corruption and crime, and improving the overall reputation of the country should all increase opportunities and send the right signals. This will 167 provide all four countries with an opportunity to attract more FDI. While small by global standards, the impact on these countries is still significant. FDI carries with it benefits that exceed direct financial inflows. There is a significant benefit to domestic economies in upgrading management skills, introducing new systems, and forging trade links. While there is a threat to protected businesses, FDI is uniformly considered a useful injection of money and skills that increases competitiveness. In Albania, this has already been recognized in the banking sector, and in key industries. When the Savings Bank is privatized (presumably in 2002), the banking sector will be 100 percent private, with foreign participation in all 13 institutions. Their contribution to date has already provided customers with better electronic access for payments and transfers, and in some cases, trade finance. Likewise, Azerbaijan has opened up its oil sector to needed foreign expertise. The Kyrgyz Republic and Moldova are also open to foreign investment, with the latter having registered some major successes in the last year or two. For these countries to attract more than their relatively low per capita FDI levels, they will need to forge ahead with a more attractive, less bureaucratic, and less corrupt public sector. Reducing crime and political instability is important, especially in Albania where enterprises rated them as the most important constraints to performance and growth. Reducing corruption is important in all the countries, but appears to be most important in Azerbaijan, where corruption is reported to be the most pervasive. The four subject countries will also have to ensure that requisite infrastructure is in place to support private businesses - power, telecommunications, roads, air travel, hotels, etc. With small markets, their focus should be on outperforming the norrn to compensate for competitive disadvantages. In other words, the model should be Singapore, not their next-door neighbor. While FDI will remain low by global standards, the impact on the economic performance of these countries could be significant if they take the steps required to encourage inflows. Financial Intermediation and the Prudential Framework Given the difficult environment present in the real sector, it has been hard for banks to expand lending to the real sector in all the subject countries. The culture of default has been evident in the formal enterprise sector for years, largely due to earlier incentives that did not require market-based discipline on the part of companies. The legal framework provided banks with little or no protection for secured transactions, making unsecured transactions more difficult to contemplate or too costly for enterprises to assume. Where the state stayed involved in banking, these loan portfolios ultimately deteriorated in quality and generated losses for the economy to absorb through fiscal bail-outs, higher interest rates, and foregone competition. In Albania and Azerbaijan, the loan portfolios of state banks have added significant costs to intermediation. In response, Albania is properly moving forward with complete banking sector privatization. Azerbaijan is taking a more gradualist approach based on the restructuring and eventual privatization of its now two large state banks. Meanwhile, Moldova has made progress in building up capacity, yet it has no major banks, and many of the existing banks have problems of illiquid and non-earning assets. The Kyrgyz Republic has likewise experienced numerous problems in its banking sector, largely reflecting the fragility of its economy at large. 168 There have been two salient features that have pointed to the insufficiency (notwithstanding their necessity) of reform efforts to date. First, positive efforts to ensure financial discipline and proper management in banks have not been matched in the real sector, particularly in weak and fragile economies that have mobilized limited foreign investment and have low levels of trade with OECD countries. This is partly because there is no centralized regulatory authority for the enterprise sector as there is for banking, with minimum capital requirements, capital adequacy tests, liquidity management requirements, and fundamental reporting requirements. The only time such comparable discipline applies to the real sector is when companies are large enough to list on a securities exchange, when they seek depository receipt financing in international capital markets, or when they seek financing from large-scale lenders that sometimes syndicate these exposures to other banks in the West. Apart from that, accounting and other governance- related requirements are generally tax-oriented rather than focused on larger issues of stability. This is a gap that will remain, and points to other needs that include training, systems, and fundamental exposure to advanced market methods. However, this gap, which is in evidence in all four countries, also points to a second weakness. Reform efforts throughout the decade have focused on more formal institutions that cater to larger enterprises. While banks obviously mobilize household deposits and provide other fundamental services, their role has been limited from a lending standpoint because of the risks mentioned above. Any effort to ignore prudential requirements is perilous, either in the form of regulatory penalties or in the form of poor portfolio performance. Other more advanced forms of financing, such as developed securities markets, have been tried in places like the Kyrgyz Republic and Moldova without any real impact. What has not been sufficiently developed are more focused forms of financing, reflecting the scale of enterprise, income, investment, and management capacity that prevail in these economies. Progress has been made in Albania, the Kyrgyz Republic and Moldova with micro-finance and village credit groups. These have all received donor support, and are likely far more suitable for the financing needs of most enterprises. Meanwhile, the more direct contact of these groups with borrowers, and the reduced formality of the process also lead to better performance. However, these institutions alone are not enough to bring savings into the system and to achieve OECD-level intermediation figures. Banks have had to contend with an entire shift in incentives for themselves and for their prospective customers since the transition began. Rather than serving as vehicles of unchecked financial flows for political patronage and production targets, Morefocusedforms of banks have more recently been assigned the responsibility of financing reflecting managing risk and becoming a source of financial discipline in the economy at large. In some cases such as in Moldova, the scale of enterprise, banks have simply curtailed lending. This has made it more income, investment, feasible for banks to meet prudential requirements regarding and management capital adequacy, and to curtail their losses to weak or undisciplined borrowers. In other cases, lending continued capacity that prevail through the state banks until recently (as in Albania and need to be developed Azerbaijan), only to run up major losses and experience major portfolio erosion. Albania has corrected this by carving out bad loans, imposing a moratorium on lending by state banks, and moving to full privatization of the banking sector via strategic investment. In the Kyrgyz Republic, as elsewhere, public and private banks have generated poor results as a consequence of lending. 169 At the same time, banks have been criticized for not lending when they have taken remedial measures to impart discipline and improve the quality of their assets. One common critique is that they leave behind many projects that are sensible in a near wholesale rejection of all loan proposals, depriving the real sector of financing needed for growth. Their preference for investing in government securities as a safe way to recapitalize (to offset the effect of non- performing assets) leads to further criticism of their role as a source of quasi-fiscal financing, rather than as true intermediaries. Banks' responses are that they are restricted in their activities and risk assumption by regulatory authorities. Such restrictions appear to be justified considering the general culture of loan default that has permeated all four banking systems in the past (as reflected in high levels of NPLs). However, where problems appear to have been suppressed, as in Moldova, this has only translated into low levels of intermediation from the funding side as well, and a greater reliance on arrears than might otherwise have materialized. The problem of arrears in Azerbaijan and the Kyrgyz Republic suggests this is endemic to much of the CIS. In one sense, all four central banks have been effective in stabilizing inflation rates and, in most cases, exchange rates. While the Kyrgyz Republic experienced major depreciation of its currency after the Russian crisis, it can be said that the regulatory authorities in Azerbaijan and Moldova had limited their respective exposures to the Russian market, and that this shielded the banks from more adverse effects felt in countries like the Kyrgyz Republic. Albania has long shown extraordinary stability with inflation rates and exchange rates, partly due to high remittance flows and the cushion of its informal trade. As a contributing factor to financial sector stability, banking is one sector (along with telecommunications and a few other areas where foreign expertise is recognized as essential) where the public in many markets shows greater confidence in internationally recognized investment grade-rated banks than they do in domestic banks. This is for enhanced services, market links, perceptions of safety, and general perceptions of professionalism. In the past, depositors also felt confident about their savings banks, largely due to implicit deposit guarantees. However, these were ultimately shattered in the CIS countries. Problems of asset quality in these savings banks also added to confidence-related problems once the savings banks moved beyond their narrow focus and became more "commercial". While there are questions about whether a radical shift to market conditions and risk-based supervision is the most appropriate route to pursue, it must be accepted that the prudential framework needs to be disciplined and sound for long-term safety and proper risk management. When such a foundation is absent or abandoned, it takes years to restore performance and public trust. However, it is possible that a more objective strategy, in which the framework evolves in stages from "rules-based" to "risk-based" standards, would be more appropriate for small economies in situations like those facing the four subject countries. At a minimum, this would allow for needed training and adaptation by banks and regulators, and perhaps may not have led to such low levels of lending. The central focus here is on development of appropriate and needed risk management capacity, which depends on sound accounting and audit standards, modernized IT and MIS, functioning internal audit committees, and effective governance. However, in general, lending levels have been low for reasons apart from risk management capacity, namely inability to collect on collateral when loans are non-performing and the 170 prudential framework whose incentives are focused on adequate capital and liquidity based on sound asset quality. In the future, there may be a rationale for more customized financial framework in which credit unions, savings and credit associations, micro-finance and village financing groups, and other small-scale operators are viewed as a more essential priority. Here, the establishment of customized needs could be introduced, such as pledge laws, property registries, and specialized courts on a decentralized basis. Credit information bureaus that document loan performance could be developed to ensure that performance (good and bad) is recognized, and more seamlessly introduced into the banking system when financing requests exceed the levels provided by the smaller institutions. There may be some room for the easing of requirements on these smaller institutions as the demise of individual institutions would not be as devastating to financial sector stability as the demise of a major deposit-taking bank. What this approach also suggests is that donors should focus on the smaller-scale aspects of financial intermediation at least There may be a as much as they are focused on securities markets, contractual rationale for a more savings, and more complex, secondary forms of financing. These should certainly not be ignored, particularly as pension customizedfinancial reform remains a critical macroeconomic and structural need framework in which and insurance is necessary to foster investment. Their credit unions and development also helps to reduce the emphasis on banks for all financing needs, and allows a more appropriate segmentation smallerfinancial of the market. However, there is also an argument that these institutions are should evolve over time in a more "organic" way based on emphasized market demand, and that the phasing of reforms should be more parallel to the structure of the economy at large. That would translate into most donor efforts being in the small-scale domain, with an eventual view to graduation to more formal and comparatively larger institutions (e.g., banks) that have higher fixed costs and more complex regulatory, managerial and administrative requirements. At a minimum, donors should not encourage premature development of "universal" institutions until adequate institutional and market capacity are in place. Legal Framework The issue of transforming the legal framework with appropriate incentives is a challenge that is likely to take a generation, at least. Part of the reason is that judicial capacity-building projects were not introduced into the transition process until after some of the shocks occurred. Even where major weaknesses were identified in the legal framework, as they were early on, prospects for suitable improvement were overestimated. Accommodating earlier legal structures with the requirements of a modem economy is an ambitious endeavor that requires a comprehensive framework with appropriate legislation and, above all, trained judges, bailiffs, administrators, and elected officials. Instead, what has occurred is a frenzy of legislation and regulation driven by different pressures, combined with modest levels of training and assistance to support the effort. Importantly, this assistance has not been demand-based, so it tended to 171 reflect the views and capacity of specific donors - not to Judicial capacity- mention national conmmercial interests in some cases - rather than genuine local demand arising from local conditions. building projects were not As in all societies, it takes decades for bodies of law, decisions, introduced into precedents, and general capacity to take shape. Given that Albania was among the most isolated of communist countries for the transition more than four decades and that Azerbaijan, the Kyrgyz process until after Republic and Moldova have all had to revise and unwind Soviet- era legislation, re-working the legislation alone has taken time. some of the shocks In the case of the four countries as well, unlike several other occurred Central European and Baltic states, there is virtually no commercial tradition that was actively exercised through courts. Thus, in all four countries, the introduction of modem norms for bankruptcy, liquidation, and secured transactions is a radical development. The ability to make such reforms effective will likely take decades. These weaknesses have shown up in all four countries with Even with secured regard to financial intermediation, and represent a major factor in assets, the time banks being unwilling to lend and take risk. Even with secured andcotinvolved assets, the time and cost involved in repossessing collateral and cost Involved makes such prospects unattractive. The vulnerability of court in repossessing personnel to kickback, bribes and other forms of payment adds to collateral makes the risk. This translates into low levels of lending from banks, and this may well remain the case until a more creditor-friendly lending environment is in place. This is beginning to happen in Albania, unattractive and has actually been partly in place in the Kyrgyz Republic. However, in the latter case, enterprises have been able to avert bankruptcy by running up fiscal and utilities payment arrears. Under such circumstances, banks have had limited recourse through the courts. These are the kinds of practical problems that need to be resolved for banks to ultimately feel confident about lending to the enterprise sector. Under the circumstances, a strategy that focuses on the "batching" of legislation, transactions and specialized in-court and out-of-court capacity might be more effective in solving part of the problem. The objective here would be to narrow the focus so that commercial transactions could be implemented within the context of a more narrow yet effective legal and judicial framework. This is a lesson that can partly be drawn from Poland in 1993-94, where exceptions were permitted to the civil code for 15 months to allow out-of-court restructuring to proceed as a basis for financial restructuring of debtors and banks' bad loan portfolios. While not exactly the same situation, the four subject countries would benefit from alternative forms of dispute resolution, such as out-of-court arbitration, as well as specialized courts for common commercial transactions. This approach could be applied with specifically tailored legislation, defining clear rights and responsibilities of parties all the way to standardized contracts. Automated and mandatory pledge registries and properly notarized contracts would clarify rights and responsibilities associated with secured assets as adjuncts to legal credit agreements. International standards of asset valuation combined with clear guidelines regarding quality control and substitutability would provide some modicum of protection to those 172 receiving pledges. Standardized and statutorily approved contract language could be adopted to serve as basic norns to be followed according to specific types of transactions, much as lease agreements and Powers of Attorney are standardized in Western markets. The use of such forms and consequent disputes would entitle the parties to access to specialized "courts," with specially trained judges, to benefit from expedited procedures under the direction of trained arbitrators. While this approach would not solve all problems, it would be effective at addressing some of the impediments to lending and basic dispute resolution. In addition, land privatization, to the extent permnissible based on local cultural standards and traditions, is an important issue to encourage secured transactions. Innovative land-lease arrangements and the development of commercial property markets based on clear claims and sound mortgage legislation could also help overcome some of these obstacles. 173 BIBLIOGRAPHY Bardhan, Pranab, 1997. "Corruption and Development: A Review of the Issues." Journal of Economic Literature 35 (3), 1320-1346. Blanchard, Olivier and Philippe Aghion, 1996. "On Insider Privatization." European Economic Review, 40 (3-5), 759-66. Blomstr6m, Magnus, and Ari Kokko, 1996. "The Impact of Foreign Investment on Host Countries: A Review of the Empirical Evidence." Policy Research Working Paper 1745, World Bank, Washington DC. Borish, Michael, Wei Ding and Michel Noel, 1996. On the Road to EUAccession: Financial Sector Development in Central Europe. World Bank, Washington DC. BP Amoco, 1999. BP Amoco Statistical Review of World Energy. BP Amoco PLC, London, UK. Bruno, Michael and William Easterly, 1998. "Inflation Crises and Long-Run Growth." Journal ofMonetary Economics 41 (1), 3-26. Coats, Warren, Betsy Faulk and Walter Faulk, 1997. Building a Strong Banking Sector in Transition Economies: The Case of Moldova. European Bank for Reconstruction and Development, London, UK. Dasgupta, Susmita, Somik Lall, and David Wheeler, 2000. "Policy Reform, Economic Growth and the Digital Divide: An Econometric Analysis." Policy Research Working Paper #2567, World Bank, Washington DC. Dethier, Jean-Jacques, Hafez Ghanem, and Edda Zoli, 1999. "Does Democracy Facilitate the Economic Transition? An Empirical Study of Central and Eastern Europe and the Former Soviet Union." Policy Research Working Paper #2194, World Bank, Washington DC. Djankov, Simeon, 1999a. "Ownership Structure and Enterprise Restructuring in Six Newly Independent States." Comparative Economic Studies 41 (1), 75-95. Djankov, Simeon, 1999b. "The Restructuring of Insider-Dominated Firms." Economics of Transition 7 (2), 467-479. Djankov, Simeon and Peter Murrell, 2000a. "The Determinants of Enterprise Restructuring in Transition: An Assessment of the Evidence." Mimeo, World Bank, Washington DC. 175 Djankov, Simeon and Peter Murrell, 2000b. "Enterprise Restructuring in Transition." Mimeo, World Bank, Washington DC. (EIU) Economist Intelligence Unit, 2000. Country Profile: Albania. Economist Intelligence Unit, London, UK. (EIU) Economist Intelligence Unit, 2000. Country Profile: Azerbaijan. Economist Intelligence Unit, London, UK. (EIU) Economist Intelligence Unit, 2000. Country Profile: Moldova. Economist Intelligence Unit, London, UK. (EIU) Economist Intelligence Unit, 2000. Country Profile: Kyrgyz Republic. Economist Intelligence Unit, London, UK. (EBRD) European Bank for Reconstruction and Development, 1997. Transition Report 1997: Enterprise Performance and Growth. European Bank for Reconstruction and Development, London, UK. (EBRD) European Bank for Reconstruction and Development, 1998. Transition Report 1998: Financial Sector in Transition. European Bank for Reconstruction and Development, London, UK. (EBRD) European Bank for Reconstruction and Development, 1999. Transition Report 1999: Ten Years of Transition. European Bank for Reconstruction and Development, London, UK. (EBRD) European Bank for Reconstruction and Development, 2000. Transition Report 2000: Employment, Skills, and Transition. European Bank for Reconstruction and Development, London, UK. Fine, Gary, 1998. "Broad Ownership Schemes in Chile, Malaysia, and the Kyrgyz Republic". Fischer, Stanley and Ratna Sahay. 2000. "Economies in Transition: Taking Stock". Finance & Development, September. Frydman, Roman, Cheryl Gray, Marek Nessel and Andrzej Rapaczynski, 2000. "The Limits of Discipline: Ownership and Hard Budget Constraints in the Transition Economies." Economics of Transition 8(3), 577-601. Hashi, Iraj and Lindita Xhillari, 1999. "Privatization and Transition in Albania." Post- Communist Economies 11 (1), p. 99-125. Hellman, Joel, Geraint Jones, Daniel Kaufrmann, and Mark Schankerman, 2000. "Measuring Governance and State Capture: The Role of Bureaucrats and Firms in Shaping the Business Environment." European Bank for Reconstruction and Development Working Paper # 51, London, UK. 176 (IMF) International Monetary Fund, 1998a. Azerbaijan Republic: Recent Economic Developments. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 1998b. Republic of Moldova: Recent Economic Developments. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 1999c. Albania: Recent Economic Developments. International Monetary Fund, Washington DC. (1\4IMF) International Monetary Fund, 1999d. Poverty Reduction and Growth Facility Policy Framework Paper: 2000-2002 (for the Kyrgyz Republic). International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 1999e. Republic of Moldova: Recent Economic Developments. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 1999f. Occasional Paper 183. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 2000a. Albania: Recent Economic Developments. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 2000b. Article IV Consultations with Government of Albania. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 2000c. Article IV Consultations with Government of Azerbaijan. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 2000d. Article IV Consultations with Government of the Kyrgyz Republic. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 2000e. Article IV Consultations with Government of Moldova. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 2000f. Azerbaijan Republic: Recent Economic Developments and Selected Issues. International Monetary Fund, Washington DC. (IiMF) International Monetary Fund, 2000g. International Financial Statistics. International Monetary Fund, Washington DC. (IMF) International Monetary Fund, 2000h. Kyrgyz Republic: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC. Jarvis, Chris, 1999. The Rise and Fall of the Pyramid Schemes in Albania. International Monetary Fund, Washington DC. Jernakowicz, Wladyslaw W. and Julian Pankow, 1995. "Privatization in the Kyrgyz Republic." Russian and East European Finance and Trade 31(6), 31-72 177 Komia, Janos, 1980. Economics of Shortage. North-Holland, Amsterdam, Holland. La Porta, Rafael, Florencis Lopez-de-Silanes, Andrei Shleifer and Robert W. Vishny, 1998. "Law and Finance." Journal of Political Economy 106 (6), 1113-1155. Linz, Susan and Gary Krueger, 1998. "Enterprise Restructuring in the Russian Federation's Transition Economy: Formal and Informnal Mechanisms." Comparative Economic Studies 40 (2), 5-52. Muent, Holger, Francesca Pissarides and Peter Sanfey, 2000. Taxes, competition andfinance for Albanian enterpripses: evidence from afield study. EBRD, London, UK. "New Corruption Indexes of Transparency International: Wide Range of Scores". October 1999. Transition. World Bank, Washington DC, and William Davidson Institute, Michigan. Pistor, Katharina; Martin Raiser, and Stanislav Gelfer, 2000. "Law and Finance in Transition Economies." Economics of Transition 8 (2), 325-368. Rosenberg, Christoph and Tapio Saavalainen, 1998. How to Deal with Azerbaijan 's Oil Boom? Policy Strategies in a Resource-Rich Transition Economy. Intemational Monetary Fund, Washington DC. Saggi, Kamal, 2000. 'Trade, Foreign Direct Investment, and International Technology Transfer: A Survey". Policy Research Working Paper #2349, World Bank, Washington DC. Schaffer, Mark E., 1998. "Do Firms in Transition Economies Have Soft Budget Constraints? A Reconsideration of Concepts and Evidence." Journal of Comparative Economics 26 (1), 80-103. Sherif, Khaled, George Clarke, and Robert Cull, 2000. "Progress with Privatization: The Experience in Eastern Europe and Central Asia." Mimeo, World Bank, Washington DC. Shleifer, Andrei and Robert W. Vishny, 1993. "Corruption." Quarterly Journal of Economics 108 (3), 599-617. TACIS, 2000. Azerbaijan Economic Trends Quarterly. http://www.economic-trends.org/az/ TACIS, 2000. Moldova Economic Trends Quarterly. http://www.economic-trends.org/md/ Thomson Financial Bankwatch, 2000. Banking System Report for Azerbaijan. Thomson Financial Bankwatch, London, UK. UNDP, 1999. Azerbaijan Human Development Report. United Nations, New York. UNDP, 1999. Kyrgyz Republicstan Human Development Report. United Nations, New York. 178 UNDP, 1999. Moldova Human Development Report. United Nations, New York. Vienna Institute for Comparative Economic Studies, 2000. "Countries in Transition". Vienna Institute for Comparative Economic Studies, Vienna. World Bank, 1994. "Moldova: Moving to a Market Economy". World Bank, Washington DC. World Bank, 1995. "Staff Appraisal Report. Republic of Moldova. First Private Sector Development Program". World Bank, Washington DC. World Bank, 1996. "Country Assistance Strategy: The Azerbaijan Republic". World Bank, Washington DC. World Bank, 1997 "Staff Appraisal Report. Republic of Moldova. Second Private Sector Development Program". World Bank, Washington DC. World Bank, 1998. "ECA Privatization and Restructuring Database". World Bank, Washington DC. World Bank, 1999a. "Country Assistance Strategy: The Azerbaijan Republic". World Bank, Washington DC. World Bank, 1999b. "Country Assistance Strategy: The Republic of Moldova". World Bank, Washington DC. World Bank, 1999c. Kyrgyz Republic: Private Sector Review in the Transitional Era. World Bank, Washington DC. World Bank, 2000a. Entering the 215' Century. World Bank, Washington DC. World Bank, 2000b. Making Transition Workfor Everyone: Poverty and Inequality in Europe and Central Asia. World Bank, Washington DC. Country Sources Azerbaijan Ministry of State Property website at http://www.msp.gov.azl Azerbaijan State Committee on Statistics website at http://www.statcom.baku- az.com/index .htm Central Bank of Albania website at http://www.bankofalbania.orM 179 Central Bank of Moldova website at http://www.bnm.org Department of Privatization and State Property Administration website at http://www.privatization.mdl Kyrgyz Republic Development Gateway website at http://Kvrivz Republicinvest.orn Kyrgyz Republic State Statistics Agency website at http://nsc.bishkek.su/Eng/Home/hidex.html National Bank of Kyrgyz Republicstan, 2000, monthly reports. 180 Recent World Bank Discussion Papers (continued) No. 384 Integrating Social Conicerns into Private Sector Decisiontitakinig: A Reviewo of Corporate Practices in the Mininig, Oil, anid Gas Sectors. Kathryn McPhail and Aidan Davy No. 385 Case-by-Case Privatization in the Russiant Fedlerationi: Lessonsfrom Initernationlu7 Experiecncc. Harry G. Broadman, editor No. 386 Strategic Management for Government Agencies: An Institutionial Approach for Developing and Transition Economies. Navin Girishankar and Migara De Silva No. 387 The Agrarian Econotmiies of Central and Eastern Europe and the Commonwealth of Independenit States: Situtation 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 Evaluationi. Julia Dayton No. 390 Evolution of Aoiildtioal Services in Sub-Saluirani Africa: Trends and Prospects. V Venkatesan and Jacob Kampen No. 391 Financial Incentives for Renewable Energy Development: Proceedings of an International 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 Conferenice Sponsored by the Inistitute of Policy Studies, Singapore, and the World Bank, Novemtber 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-Saharani Africa: Beyond the Supply Con1strain1t. Joseph Bredie and Girindre Beeharry No. 396 Transforming Agricultural Research Siystems in Transition Economies: The Case of Russia. Mohinder S. Mudahar, Robert W. Jolly, and Jitendra P. Srivastava No. 398 Land Reform and Farni Restructuiring in Moldova: Progress and Prospects. Zvi Lerman, Csaba Csaki, and Victor Moroz No. 400 Russian Enterprise Reform: Policies to Further the Transitiont. Harry G. Broadman, editor No. 401 Russia,, Trade Policy Reform for WTO Accession. Harry G. Broadman, editor No. 402 Trade, Global Policy, and the Environment. Per G. Gredriksson, editor No. 403 Ghana: Gender Analysis and Policymaking for Developmnenit. 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 Power 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 Strategy for 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 Bankintg Crises: Lessons fromt the Republic of Korea and Mexico. Jose De Luna-Martinez No. 414 Liquefied Natural Gas in China: Options for 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 Apportioninlg 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 THE WORLD BANK 1818 H Street, N.W Washington, D.C. 20433 USA Telephone: 202-477-1234 Facsimile: 202-477-6391 Internet: www.worldbank.org E-mail: feedback@worldbank.org ISN 8135065- ISBN 0-8213-5065-X