State-owned Banks in the Transition:

Origins, Evolution and Policy Responses

By, Khaled Sherif, Michael Borish and Paul J. Siegelbaum

TABLE OF CONTENTS

CHAPTER ONE: BACKGROUND.............................................................................................1

General Introduction..................................................................................................................1

Purpose and Methodology.........................................................................................................4

Acknowledgements....................................................................................................................5

CHAPTER TWO: ECONOMIC STRUCTURE AND TRENDS IN THE EARLY STAGES OF TRANSITION..........................................................................................................................6

Economic Structure of Transition Countries in the Early 1990s...............................................6

Brief Synopsis of Monetary and Fiscal Trends in Transition Countries: 1989-95....................7

CHAPTER THREE: PROFILE OF STATE BANKS EARLY IN THE TRANSITION.....11

General Profile.........................................................................................................................11

Traditional Roles Played By State Banks................................................................................18

Governance, Management, and Operating Standards of State Banks.....................................19

Summary of Types of Specialized State Banks.......................................................................21

CHAPTER FOUR: BANKING SECTOR TRENDS BY THE MID-1990S...........................29

The Financial Status of Banks in 1995 and Differing Patterns of Development.....................29

State Banks and Broad Financial Intermediation Measures....................................................33

Emerging Role of Private Banks..............................................................................................44

Summary: Consistency and Divergence Among Transition Regions Through 1995..............47

CHAPTER FIVE: CURRENT STATUS AND CONTINUING PROBLEMS OF STATE BANKS SINCE 1995....................................................................................................................51

Current Ownership Status and Trends of State Banks.............................................................51

Financial Condition of the State Banks...................................................................................55

CHAPTER SIX: APPROACHES: WHAT HAS AND HAS NOT BEEN DONE?...............86

The Role of State Banks and the Effects of Sustaining the State System................................86

Problem Assets, Bank Restructuring Costs, and Approaches in Transition Countries...........97

The General Impact of Bad Loan Quality and the Pace of Reform.........................................99

Regional Funding and Intermediation Trends.......................................................................101

CHAPTER SEVEN: REMAINING ISSUES: WHAT NEEDS TO BE DONE, AND HOW TO GET THERE.......................................................................................................................105

General Findings and Conclusions........................................................................................105

Prospects and Preconditions for State Banks’ Privatization and Resolution.........................107

Recommendations for State Banks’ Privatization and Resolution........................................111

What Can/Should Donors Do?...............................................................................................119

CHAPTER ONE: BACKGROUND

GENERAL INTRODUCTION

This study examines state ownership in the banking systems of the transition economies, and how the continuation of this ownership distorts economic reform efforts. While many of the distortions found in poorly performing economies do not originate in the banking sector, the risk is always there. Admittedly, in many countries, banks have remained in state hands because privatization options have not provided acceptable solutions to difficult development problems. Nonetheless, irrespective of motivation and incentives, the experience teaches us that state banks are typically either patronage vehicles that ultimately worsen prospects for competitive market development, or ineffective shells that fail to perform a useful intermediation role once government imposes effective hard budget constraints and a modern supervisory system.

One useful strategy is to differentiate different types of state banks to ascertain their specific negative impacts on the economy. Doing so reveals that industrial and agricultural banks have been the most problematic as their roles were to finance state farms and industries that employed large numbers of people and served as the backbone of the earlier socialist economic model. Because banks with industrial or agricultural orientations focused on lending to what eventually became loss-making enterprises and farms, for increased production (and the preservation of jobs, (as opposed to the generation of profit), it is these banks that incurred the greatest levels of insolvency. Two banks that reflect such problems, culminating in their eventual liquidation, Bancorex in Romania and Bank Ukraina in Ukraine, are discussed in Annexes 6 and 7 as well as in Box 6.1.

Because of their perceived importance, combined with complications arising from the methods used to privatize them, these banks often emerged in the post-socialist economies as banks that were “too big to fail.” The result was usually costly (and often repeated) recapitalizations, regulatory forbearance, and distortions in the marketplace that prevented a more efficient banking system from emerging early on. This has been true in many Central European economies, including countries that are considered among the more successful performers (e.g., Poland, Hungary, Czech Republic, Slovenia).Whatever has been positive about these countries’ performance has not been due to recapitalization, forbearance, etc., as shown in several countries where these restructuring efforts have largely given way to privatization and efforts to mobilize strategic investment. Rather, their successes lie either in opening up their systems to market-based competition to more fully integrate with EU economies, being able to attract strategic investment into their financial systems,1 and leveraging off of their relatively favorable positions in the socialist era to evolve more quickly to competitive status.

1 Even Slovenia, which has resisted foreign investment more than the other three economies, has experienced substantial foreign investment in banks, insurance and other financial services despite continued high levels of state ownership in the banking system. This has added competition, and permitted the foreign banks and insurance companies to compete and capture market niche positions in the country where purchasing power parity income measures are the highest among transition countries. Income measures have been comparatively high for decades due to the significant trade and export position of Slovenia during the Yugoslav era.

It should be noted, however, and in fairness, that rapid introduction of strategic investment in domestic banks would not have produced miracles in the early 1990s. It often takes years before new banks gain the skills needed to prudently assume major balance sheet risk. Moreover, given the instability of the market at the time, gathering needed information and security for increased lending and investment were major challenges. In addition, some countries faced bleak prospects even if they were predisposed to pursuing policy reforms that focused on moving to market-based systems. The experience of the Kyrgyz Republic is an example of this dilemma. In other cases, such as Uzbekistan (see Annex 8 and Box 6.2), having substantial resources may serve as a disincentive to introducing reforms due to the political inconvenience involved. The presence of such resources and financing can temporarily alleviate bank portfolio problems, making the immediate political choice of avoiding reform more expedient. However, this defers addressing the long-term structural problems facing the economy, and usually results in higher costs at a later date. In either case, policy makers were left with difficult and generally unattractive options on how to resolve the fate of troubled banks with major links and exposures to structurally weak sectors of the economy. Many also believe that privatization occurred more quickly and efficiently in places like Poland and Hungary as a result of their earlier restructuring efforts. While this may be true, it is also a very expensive proposition for countries with less attractive markets, and weaker prospects for attracting investment in the first place.

In contrast to agricultural and industrial banks, savings banks were and still are among the most sensitive to market trends. They are particularly important with regard to deposit safety and long-term public confidence in the ability of the banking system to operate in a sound manner. Often, this trust has been compromised, with long-term adverse effects on efforts to build stable banking and financial systems. (For an example of this, see Annex 9 and Box 5.2 for a review of Ukraine’s Oschadny Bank, the traditional state savings bank.) At the outset of the transition, after the break-up of the monobank system, savings banks held significant local currency household deposits, making their funding base attractive. In some countries (mainly CIS and the Baltic states), this advantage quickly evaporated as hyperinflation destroyed local currency-denominated savings. Elsewhere (mostly in Central Europe), given the less unstable macroeconomic situation, the condition of savings banks was not as dismal. In some cases, such as with the Savings Bank in Albania, efforts were made to diversify the bank’s assets and activities to provide a more balanced and sustainable income stream. However, the inability to properly manage new risks quickly led to an erosion in financial condition that later had to be corrected2 at significant cost to the budget.

Savings banks have sometimes been used as sectoral safety nets, required to absorb smaller troubled banks as part of banking sector consolidation. This was the strategy pursued by the Czech Republic with Ceska Sporitelna (see Annex 10 and Box 7.1), only to lead to a worsening of the bank’s financial condition and likely diminution of value when presented to the market for private acquisition.3 While the objective of creating larger institutions in comparatively small markets was often rational, more often than not the strategy led to erosion in the financial position of the savings banks due to the absorption of bad banks, bad portfolios, additional and unneeded staff, and a culture that was non-traditional (e.g., lending-oriented, risk-seeking). These banks further suffered from limited risk management capacity, and insufficiently

2 In 1997-98, the Government of Albania suspended the right of public banks to lend as part of its effort to stabilize the banking system after the collapse of pyramid schemes.

2

sophisticated standards and practices. The prominence of Sberbank in the Russian market will need to be continuously scrutinized for such potential problems, given the continued use by government of this bank for a number of political functions (see Annex 11 and Box 4.1). As the Russian government has announced it will divest its holdings in banks that are less than 25 percent government-owned, Sberbank will need to avoid absorbing any banks, assets or shares that would weaken its financial position.

Meanwhile, there have been some success stories. A number of state banks initially ran into problems, but were subsequently restructured and put on a commercially sustainable path within a relatively short period. Unibanka in Latvia went through such a period, facing restructuring needs as a result of early loan portfolio problems (see Annex 12 and Box 4.2). Once dealt with resolutely by the authorities, the bank was able to withstand systemic problems faced in the Latvian banking sector in 1995, attract strategic investment, list on the local exchange, and ultimately become a part of a larger regional banking institution. Significantly, however, Unibanka is not a case of an ex-state bank successfully reformed, but rather a wholly new institution that was built from the viable branches and assets of separate institutions.

Other state banks that have fared comparatively well and been easier to privatize have been the foreign trade banks. These banks have had foreign currency funding, longstanding links to international trade and investment, and management capacity that was more attuned to international trends. Examples have included Bank Handlowy in Poland (now with Citigroup), and MKB, the Hungarian Foreign Trade Bank (now with Bayerishe Landesbank). Nonetheless, such positive experiences have not been universal among foreign trade banks. The example of Bancorex in Romania shows how one particular bank with international links and high profile connections to large state enterprises turned out to be the nexus for much of the loss-making and unsustainable economic behavior that characterized Romania through most of the 1990s. As Azerbaijan restructures a newly consolidated United Universal, it will need to avoid such a fate from occurring with its most dominant bank, the International Bank of Azerbaijan (see Annex 13 and Box 5.1). Other countries that still have state-owned foreign trade and export-import banks will likely need to ensure that similar problems do not occur.

More positively, foreign trade banks have generally performed better than the norm. Often, export-import banks and foreign trade banks are retained by the state for long periods because of the critical intermediary role they play in international transactions, and for needed trade finance services during periods when project finance is limited in volume. They frequently have well trained staff, and donor funding often runs through these banks to stimulate international trade. However, this should not be viewed as a justification for state ownership. Rather, these strengths should be viewed positively in increasing the attractiveness of the banks, and improving their prospects for privatization to well-managed, prime-rated, strategic investors.

3 Ceska Sporitelna was sold to Erste Bank of Austria in 2000 for $515 million-equivalent for a 52 percent stake after the government agreed to guarantee half of the bank’s loans for five years. However, this price was actually higher than originally expected, partly the result of a late attempt by IPB, the Czech unit of Nomura (Japan), to counter Erste’s bid. As part of the deal, Erste committed itself to a $114 million capital increase at Ceska Sporitelna or its subsidiaries, and to set aside $571 million for housing and small business programs and $29 million for venture capital. While higher than expected, many in the market believe Ceska Sporitelna would have attracted a far higher price earlier in the 1990s before it was used by the government as an anchor for the consolidation of several smaller banks that ultimately weakened the overall financial condition of the bank.

3

It can be noted that problems of state ownership are not restricted to transition countries. Turkey is currently engaged in a major restructuring and privatization exercise in its banking sector. Moreover, several OECD countries have a long tradition of state ownership in their banking systems. While such ownership traditions are changing in continental Europe, state banks in these and other comparatively wealthy markets have often generated losses and experienced problems when engaged in directed lending for political purposes (see Annex 14). This demonstrates how non-viable the approach to directed lending and non-commercial orientations in banking generally are, notwithstanding the wealth of nations that sometimes practice this approach. Important is this regard is that it is not just ownership that matters, but overall incentives and practices. Private banks that engage in politicized lending patterns will also face a severe threat to their financial position over time. Thus, the key is detaching the ownership, governance and management from political patronage, and putting banking decisions on a purely commercial footing with the objective of achieving high returns and ensuring the safety of the deposit base.

METHODOLOGY

This study evaluates the history and evolution of state banks. Chapter 2 looks at economic structure and trends early in the transition, and includes a synopsis of monetary and fiscal trends from 1989-95. Chapter 3 reviews the role of state banks early in the transition, with specific data and information broken out by specialized type of state bank in each country. Chapter 4 assesses banking sector trends in the mid-1990s, including the financial status of banks, the emergence of private banks, and differing patterns of growth by region. Chapter 5 reviews the current status and continuing problems of state banks since 1995, including intermediation trends, asset quality, earnings performance, and solvency issues. Chapter 6 summarizes differing approaches to state bank reform, including an evaluation of results. Chapter 7 highlights remaining issues that require attention, including ongoing risks to state ownership in the banking system, and provides recommendations on how to move beyond the current challenges to state banks in support of development of safe and sound banking systems. This includes recommendations on privatization approaches, timing and phasing issues, and broader institutional development needs to ensure a stable foundation is in place for sound financial intermediation practices to take hold.

In addition to the chapters above, the study features a series of annexes that provide further statistical detail. These include Annexes 1-4 with statistics on most remaining state banks as of 2000. Annex 5 provides a detailed breakdown of arrears in nine selected transition countries. Short summaries of selected state banks from several transition economies and their experiences are highlighted in Annexes 6-13. A brief summary of selected state banks in OECD countries and their performance can be found in Annex 14. The study relies on data from a range of sources. The specific methods and applications surrounding the data can be found in the Methodological Notes in Annex 15. Specific sources can be found in the bibliography in Annex 16. The following summarizes key sources of data.

International Financial Statistics from the IMF were used for macroeconomic data and banking statistics.

World Bank data and publications were used for macroeconomic and structural data, as well as for estimates of state banks in the mid-1990s.

4

EBRD Transition Reports were used for banking system data regarding non-performing loans, ownership shares, and numbers of banks per country.

Bank Scope (Fitch IBCA) was used to obtain fundamental financial information on state banks and their ownership structure. This applies to 67 banks for which reports were available. Data used were almost entirely from 2000 results, and included balance sheet and income statement figures, performance ratios, and ownership status.

ACKNOWLEDGEMENTS

The authors wish to thank Marcelo Selowsky, currently with the IMF and formerly Chief Economist in the Europe Central Asia Department of the World Bank, for encouraging this study to proceed and providing resources for this to happen. Luigi Passamonti and Hormoz Aghdaey served as peer reviewers and their many comments have contributed greatly to the final product. The many task managers and researchers who assisted with data requests, clarifications, and opinions also helped immensely with the effort.

George Clarke was responsible for compiling the major portion of macroeconomic data utilized, and for producing the data on arrears that have been incorporated into the text. The authors wish to thank Alexander Pankov for coordinating the data gathering effort on the state banks, and to Alexandra Gross and Anna Sukiasyan for working with Alexander Pankov on the team to compile needed data and information on state banks. They have also been the primary authors of the cases produced in the annexes on the individual state banks.

5

CHAPTER TWO:

THE EARLY STAGES OF TRANSITION

ECONOMIC STRUCTURE OF TRANSITION COUNTRIES IN THE EARLY 1990SS

When the transition process began, the former socialist economies were generally oriented towards industry. As of 1992, 47 percent of total output was from the industrial sector. Services were about 38 percent, most of it government-related. Agriculture only accounted for about 15 percent of recorded output, although this does not capture subsistence farming. By contrast, OECD countries in 1992 showed a different distribution, with 66 percent in services,4 23 percent in industry, and only 5 percent in agriculture.

Transition country patterns showed little deviation. A few countries had prominent service sectors, mainly in Central Europe. For example, Slovenia, Hungary and FYR Macedonia all had service sectors that accounted for more than half of GDP, suggesting that there were enterprises active in transport, distribution, tourism, and related activities net of direct government administration. However, in the CIS and Baltic states, services (including government) did not exceed 39 percent (Russia), and the average was about 35 percent. The Baltic states in particular showed very low levels of service sector development, particularly Latvia and Lithuania.

Meanwhile, agriculture played a limited role in general economic output, averaging less than 15 percent. Central Europe in particular showed extraordinarily low levels of agricultural output, at only 8 percent of total GDP. This may understate primary sector output, as many people in the region rely on subsistence farming as part of their safety net. Moreover, as private farming was permitted in several of these countries,5 much of the output is presumed to have gone unrecorded. Only Albania recorded more than half of its GDP in agriculture. Many CIS countries showed about one third of their economies to be based on agriculture.

The following figure highlights the economic structure of transition countries in 1992.

4 OECD countries at the time showed 19 percent of total GDP was in financial services. While reliable figures are not available for the 27 transition countries at the time (given the economic turbulence of the period), the share of financial services to the economy was considered much lower. 5 Bulgaria, Poland, and the previous Yugoslavia (currently five countries) permitted small-scale private farming during the socialist era.

6

Figure 2.1 Structure of Transition Economies in 1992 (percent)

Notes: FYR Macedonia and Turkmenistan GDP are based on per capita incomes times population.Sources: World Development Report 1994; EBRD Transition Reports, 2000 and 2001.

With such economic structure, most of the banking that occurred was geared to financing the industrial sector. Foreign trade banks attempted to sustain traditional trade links that were integral to the central planning process, but which generally imploded when the Soviet Union collapsed. Likewise, these banks financed the export of goods for hard currency when trade opened up. In the latter case, this was already happening in places like Hungary, Romania and Yugoslavia that had opened up their trade regimes well before the collapse of the Soviet Union. In some cases, these functions converged in the agricultural sector, agro-processing, or in commodities (e.g., oil in Azerbaijan and Kazakhstan, natural gas in Turkmenistan, cotton in Tajikistan and Uzbekistan).

MONETARY AND FISCAL TRENDS IN THE EARLY TRANSITION YEARS: 1989-95

Following the initial shocks experienced in the early 1990s, most Central and Eastern European countries tightened monetary policy if they had not already, and by extension, regulations limiting risk-seeking behavior in the banking system. These countries had already shown earlier signs of monetary discipline and central bank independence, and this resulted in greater stability by the mid-1990s. For example, on an unweighted basis, inflation rates averaged 15 percent in 1995.6 These were far lower than peak rates experienced just a few years before, in 1989-1992. Six of 10 countries had single digit inflation figures, and no country exceeded year-end CPI of 33 percent. Rather, the weakness was more on the fiscal side, where loss-making enterprises

6 EBRD figures presented regionally showed a mean of 20.5 percent for Central Europe and the Baltics, 39.4 percent for southeastern Europe, and 350 percent for the CIS countries.

7

were still receiving financing, either from the banks (usually state-owned), the budget, off-budgetary accounts, or arrears to state companies and energy suppliers. While lower than in earlier years, fiscal deficits were about 3.3 percent of GDP on an unweighted basis in 1995. It should be noted that the consolidated deficits were respectable by 1995, but they understated the softness of budget constraints on the state sector due to the build-up of arrears. Moreover, lending to the state sector still accounted for stocks and flows of bank lending.7 Continued state ownership in the banking sector, particularly in “large” banks8 with large exposures and long standing ties to state enterprises and farms, made this possible. State banks still accounted for more than half of all banking system assets in most CEE countries in 1995. The table below shows that state banks’ asset shares were declining in CEE countries in the early to mid-1990s, yet were still high. On an unweighted basis, state bank assets were more than half of total through the mid-1990s. Even in countries where these shares were reported to be less, this was not the case.9

Table 2.1 Central and Eastern Europe Indicators (1990-95) (percent)

Inflation Rate Fiscal Deficit/GDP State Bank Assets/Total1990-94 1995 1990-94 1995 1992 1995

Albania 237.0 6.0 -31.0 -10.3 97.8 94.5Bulgaria 338.9 32.9 -10.9 -6.4 82.2 82.2Croatia 1,149.0 3.8 -3.9 -0.9 58.9 51.9Czech Republic 52.0 7.9 -3.1 -1.8 20.6 17.6FYR Macedonia 1,935.0 9.0 -13.8 -1.2 N/A N/AHungary 32.2 28.3 -8.9 -6.2 81.2 52.0Poland 249.3 21.6 -6.7 -2.8 86.2 71.7Romania 295.5 27.8 -4.6 -2.6 80.4 84.3Slovak Republic 58.3 7.2 -7.0 0.2 70.7 61.2Slovenia 247.1 9.0 -0.3 -0.5 47.8 41.7Unweighted avg. 459.4 15.4 -9.0 -3.3 62.6 58.9Notes: Inflation rates (CPI year end) and fiscal deficit (general government balance) figures are from 1990-94 at peak; state bank shares are earliest reported if not available for 1992 or 1995; Czech figures exclude two large banks with major ownership by the National Property Fund.Source: EBRD

In the Baltic states, inflation rates were higher than in the CEE countries, where inflation rates generally did not exceed 338 percent. However, year-end CPI rates among the three Baltic states

7 Quantifying this with precision is difficult because of the prominence of large industrial enterprises that were partly privatized, or where the strategic “investor” might have been the National Property Fund of the state, even where the enterprise was classified as “private..” However, in most CEE and Baltic countries (with Estonia as an exception), state enterprises often benefited from less than hard budget constraints. 8 “Large” is based on nominal balance sheet values, not discounted for risk and quality. In reality, virtually all “large” banks would have been much smaller had they written down their assets and capital to reflect internationally accepted standards for accounting and valuation. These banks eventually faced up to this reality in the second half of the decade. However, banking systems in the CIS and in several CEE are still dealing with these issues.9 In the Czech Republic, the low state shares exclude Ceska Sporitelna and Komercni, two major state banks that would have radically shifted the ratios.

8

peaked at about 1,000 percent. Perhaps because the monetary challenge was greater, 1995 inflation rates were higher, at about 29 percent. On the other hand, the Baltic states showed a high level of discipline, bringing their unweighted average 1995 inflation rates to about a small fraction of earlier peak levels.10 Meanwhile, monetary discipline was reinforced by fiscal discipline, as all three countries kept spending within reasonable bounds, even during the hyperinflationary period. The combination of monetary and fiscal discipline imposed hard budget constraints on the state enterprise sector, and by 1995, Estonia and Latvia showed limited state shares of bank assets. In Estonia, this was carried out fairly systematically with the liquidation of loss-making banks that had been branches of the Gosbank system earlier on. Only Lithuania took several more years to reduce its state share of bank assets. It should be noted that “private” ownership alone was not sufficient for sound performance. In Latvia, the state share of bank assets was low prior to 1995. Yet its largest bank collapsed in 1995 while technically being a private bank.11 This prompted a more disciplined approach to financial services and banking supervision from that point on, a position that has been consistently reinforced by the central bank since 1995.

Table 2.2 Baltic States Indicators (1990-95) (percent)

Inflation Rate Fiscal Deficit/GDP State Bank Assets/Total1990-94 1995 1990-94 1995 1992 1995

Estonia 953.5 29.0 -0.7 -1.3 28.1 9.7Latvia 959.0 23.1 -4.0 -3.9 7.2 9.9Lithuania 1,161.0 35.5 -5.5 -4.5 53.6 61.8Unweighted avg. 1,024.5 29.2 -3.4 -3.2 29.6 27.1Notes: Inflation rates (CPI year end) and fiscal deficit (general government balance) figures are from 1990-94 at peak; state bank shares are earliest reported if not available for 1992 or 1995.Source: EBRD

Meanwhile, the CIS countries faced enormous challenges in terms of hyperinflation and a non-viable fiscal base. By all measures, the CIS countries failed to achieve macroeconomic stability to accommodate structural reform requirements. This, in turn, triggered a downward spiral that exceeded the magnitude of decline experienced in most CEE and Baltic countries. Recent measures of output relative to pre-transition levels reflect the depth of decline in CIS countries.12

Thus, their ability to recover from central planning has been much more difficult. This was evidenced first by the extraordinarily high inflation rates suffered by the CIS countries, with peak rates at nearly 5,000 percent on average. This led to the introduction of new currencies in CIS countries, including eventually in Russia with the introduction of a new ruble on January 1, 1998. Even by 1995, inflation rates were still significantly higher than in the CEE and Baltic countries and, in several cases, remained at hyperinflationary levels.. Moldova was the only CIS country whose inflation rate was less than the average of the three Baltic states that year. Among

10 This was true in the CEE countries as well, as the 1995 unweighted inflation rate was 3.4 percent of the unweighted peak average rates from 1990-94.11 See A. Fleming and S. Talley, “The Latvian Banking Crisis—Lessons Learned,” World Bank, 1996. 12 On average, CIS countries now operate at about 60 percent of pre-transition levels, as compared with the Baltic states at 70 percent and CEE countries at around 90 percent. See S. Fischer and R. Sahay, “Taking Stock,” Finance & Development, September 2000.

9

CEE countries, only Hungary, Romania and Bulgaria had higher rates that year than Moldova. Meanwhile, CIS fiscal deficits were about two times the magnitude of the CEE and Baltic states, at more than 6 percent. Part of this related to the collapse of the industrial sector, as much of the earlier fiscal revenue flow had been generated off of enterprise sales in the form of turnover taxes.13 With sales now plummeting and often unrecorded (to avoid tax payments), government revenues declined. Corruption also played a role in tax payments being made, but not finding their way to national treasury accounts. All of this reflected a very unstable environment for normal banking operations.

Against this backdrop, CIS countries often imposed hard budget constraints – in some cases by circumstance more than by choice – on the state sector. However, CIS countries were also caught in the difficult situation of seeking to maintain or revive production to preserve jobs and reactivate the fiscal base. In a tight money regime, this led to budgetary subsidies and transfers, concessionary rollovers from the banking system, and arrears to enterprises, social funds, workers and fiscal authorities (see Annex 5 for greater detail). Interestingly, the CIS countries showed that the state bank share of assets was about half that registered in the CEE countries. However, this also indicates that “private” banks in the CIS were largely used as vehicles of financing for their enterprise owners and other related parties, rather than as channels of financial discipline. These “privatized” practices reflect weaknesses in the banking sector framework and incentive structure in the CIS countries, and inherent flaws in “ownership transformation.”

Table 2.3 Commonwealth of Independent States Indicators (1990-95) (percent)

Inflation Rate Fiscal Deficit/GDP State Bank Assets/Total1990-94 1995 1990-94 1995 1992 1995

Armenia 10,896.0 32.0 -54.7 -11.0 1.9 2.4Azerbaijan 1,788.0 84.5 -15.3 -4.9 88.7 80.5Belarus 1,996.0 244.0 -2.5 -1.9 69.2 62.3Georgia 7,488.0 57.4 -26.2 -4.5 98.4 45.9Kazakhstan 2,984.0 60.0 -7.9 -2.7 19.3 24.3Kyrgyz 1,363.0 31.9 -17.4 -17.3 100.0 69.7Moldova 2,198.0 23.8 -26.2 -5.7 0.0 0.3Russia 2,506.0 128.6 -42.6 -5.9 N/A 37.0Tajikistan 7,344.0 2,133.0 -30.5 -11.9 N/A 5.3Turkmenistan 9,750.0 1,262.0 -1.4 -1.6 26.1 26.1Ukraine 10,155.0 181.0 -25.4 -4.9 N/A 13.5Uzbekistan 1,281.0 117.0 -18.4 -4.1 46.7 38.4Unweighted avg. 4,979.1 362.9 -22.4 -6.4 37.5 33.8Notes: Inflation rates (CPI year end) and fiscal deficit (general government balance) figures are 1990-94 at peak; state bank shares are earliest reported if not available for 1992 or 1995; earliest figures for Russia, Tajikistan and Ukraine are for 1996.Source: EBRD

13 See L. Barbone and D. Marchetti, “Economic Transformation and the Fiscal Crisis: A Critical Look at the Central European Experience of the 1990s,” World Bank Working Paper 1286, April 1994.

10

CHAPTER THREE: STATE BANKS EARLY IN THE TRANSITION

GENERAL PROFILE14

Overview. The reform of banking systems in post-socialist Central and Eastern Europe and the former Soviet Union15 has shown some common themes throughout the years in terms of basic reform efforts, yet with broadly divergent results. Most transition countries introduced two-tier systems around 1989 as communism (and its monobank system) collapsed. This change placed monetary policy and its implementation in the hands of the central bank, and established the basis of a second tier commercial and other banks to engage in normal banking functions. Prior to 1989, these functions had all been part of one monobank system in most socialist economies. Thus, the very institutional configuration of financial intermediation represented a challenge to transition countries. This involved defining new roles and responsibilities in the first and second tiers of the new system, and conceptualizing how intermediation could proceed at a time when traditional production, trade and investment were in a state of dislocation and, in some cases, collapse.

As part of this reconfiguration, the central banks of transition countries were generally entrusted with the role of banking supervision as a function of their monetary policy role. Thus, as laws were introduced for both the new central bank and second tier banks, fundamental prudential norms and initial measures to establish supervisory oversight of the banks were also introduced. These initially dealt with ownership, capital, lending exposures, reporting requirements, and other common components of banking legislation and regulation. However, not only did the prudential requirements prove to be insufficient. Supervisory capacity itself was undermined by poor and inaccurate accounting and financial information in the banks, weak off-site surveillance capacity among the supervisory authorities, and lack of experience with on-site examinations.

After an initial period of experimentation with low minimum capital requirements and a push to liberalize the licensing process for new banks, most transition countries encountered periods of severe instability. These fundamental banking sector problems were part of larger structural problems in post-socialist economies, reflecting macroeconomic disorder (e.g., hyperinflation, exchange rate instability), the breakdown of traditional trade patterns and distribution channels, and the severe decline in purchasing power of enterprises and individuals. Liquidity shortages triggered an increase of dramatic proportions in inter-enterprise, tax and other arrears, while debt service payments to banks declined. Arrears also became more generalized, particularly in the CIS countries, where power companies, fiscal accounts, wage earners, and pension, health, unemployment compensation and other social funds effectively became net creditors to the economy.

14 General profiles of public banks at the beginning of the transition period are based on 1992 information. 15 For purposes of discussion, the 12 countries of the CIS are viewed as one group, and the Baltic countries are segregated from the former Socialist countries of Central and Eastern Europe.

11

While there were some preliminary efforts to have banks operate on a commercial basis, in most transition countries, government officials could not resist using at least some of the banks as vehicles for directed lending. Where banks operated “privately,” this was frequently on behalf of their connected shareholders, rather than on the basis of normal commercial banking principles found in a market economy. Meanwhile, state-owned banks continued to serve as the primary banking vehicle for directed lending and quasi-fiscal financing, usually to loss-making state-owned enterprises and collective farms. In some countries, this ran parallel to the slow emergence of private banks and private sector development, including in Central Europe (e.g., Hungary, Czech Republic) where high levels of direct investment and remittance flows were serving as a catalyst for modernization and privatization. However, the bulk of intermediation occurred to/through state channels. As a result, the quality of the loan portfolios of these banks declined rapidly and, for the most part, irretrievably.

Early Structural Changes in the Banking Sector. Among the centrally planned economies, virtually all had monobank systems. There were a few exceptions where commercial banks functioned on a more decentralized basis.16 However, in general, the monobank model was fairly systematically adopted and followed throughout the socialist world.

As a subset of the monobank system, most of the transition countries were accustomed to having a small number of specialized state banks (e.g., savings, foreign trade, industrial investment, agricultural) in addition to the central bank. While this might appear to be similar to the two-tier system that was later established, they functioned more like departments of one singular banking system, rather than as independent commercial bank entities operating on their own. In this sense, there was substantial differentiation from the earlier monobank model as compared with the post-socialist two tier system introduced around 1989.

Late in the socialist period (mid-1980s), some governments started timid attempts at decentralization in the banking sector by establishing new “specialized” commercial banks. For example, in Bulgaria in 1987, the Government moved to establish seven such banks (in addition to four state banks that existed earlier), each serving a particular industrial sector. The new banks provided current account facilities, accepted deposits, lent in both local and foreign currencies, and provided “venture capital” for firms in their particular sector. However, decentralization only went so far. None of the new banks had branches, and they dealt with their customers through the local offices of the National Bank. More importantly, these banks were not given any opportunity to exercise independent judgment, and merely allocated investment funds to state-owned enterprises in their sectors according to central planners’ instructions. Thus, these experiments indicated that the traditional system was failing to meet the broad banking needs of the economy. Given continued state control in most cases, there was little substantive change resulting from these experiments in terms of how the socialist economies continued to operate. Moreover, Bulgaria was more the exception than the rule. Even as the earlier system failed to meet anything more than rudimentary banking needs, there was little effort to push for reforms until after the socialist system collapsed.

16 Bulgaria and Hungary introduced two-tier systems in 1987. Prior to that, both countries had a monobank system. Only Yugoslavia had a large number of banks during the socialist era, most of them “socially-owned” by enterprises and employees. While Yugoslavia did not have a monobank system, the banks responded to the planning prerogatives of the central authorities. In this regard, they were not significantly different from their counterparts in other centrally planned economies.

12

Apart from Yugoslavia, which established a two-tier system in 1971, Hungary was the only other real exception to the rigid socialist model. Net of Yugoslavia, Hungary was the first transition country to introduce a market-oriented two-tier banking system (in early 1987). In the Hungarian example, monetary policy was implemented by the National Bank, while credit activities were undertaken by independent commercial banks operating in a competitive market. Poland followed Hungary’s lead in January 1989, with Bulgaria, Czechoslovakia and Romania following Poland in January 1990. As noted above, Bulgaria had already begun to experiment with new organizational models in 1987.

Notwithstanding these limited exceptions, centrally planned economies generally relied on the Gosbank system, with departments that specialized in the peculiarities of differing financing needs of different sectors of the economy. This functional specialization partly explains why, by 1992, the initial transformation to a two-tier banking system was characterized by sector concentration (e.g., agriculture, industry, export-import, housing, savings) in state banks. In some cases, (e.g., Croatia, Estonia, FYR Macedonia, Poland, today’s Slovak Republic), some of the state banks were smaller and commercially diversified in their activities, yet geographically concentrated,17 just as branches from the Gosbank system remained local in their economic orientation. Often, they were owned by state enterprises. In other cases, the banks were specialized by economic sub-sector.18 In general, the large state banks created from the monobank system represented the core of the new banking system that remained state-owned and highly concentrated in practically all transition countries.

After the immediate outset of the transition process, the number of banks increased very quickly. By the early 1990s, shortly after the rapid move to ownership transformation and private entry, there were roughly 2,350 banks in the 27 ex-socialist countries of Europe and Central Asia. Of these, only 200 were considered to be “major” or “prominent” state banks.19 Russia alone had 1,306 banks in 1991, 2,456 banks in 1994, and 2,297 banks in 1995. The CIS in total accounted for 1,841 of the 2,350 banks in transition countries early in the transformation period. By 1995, the number of banks in CIS countries was 3,171 out of a total of 3,783 banks in all transition countries. Thus, the CIS countries consistently accounted for about 80 percent of the total number of licensed banks in transition economies through the mid-1990s.

17 For example, in Poland, nine of the smaller Treasury-owned banks were “specialized” geographically while being more diverse in their banking activities. These banks were separate from the initial four large state-owned banks spun off from the central bank, and specialized by function (e.g., agriculture, foreign trade, housing, and foreign currency savings). In Croatia, most of the banks were local in their orientation. 18 For example, in Bulgaria, there were specialized banks for transportation, chemicals and biotechnology, electronics and defense goods, and building and construction. In Central Asia, several banks were specialized by commodity function (e.g., cotton in Uzbekistan; natural gas in Turkmenistan).19 The estimated figure is 200. However, this figure may underestimate the number of smaller banks that remained publicly owned by the state, municipalities, local government, and separate funds (e.g., National Property Fund) that were government-influenced and controlled. For example, Russia had more than 400 state banks in the early 1990s, although most of these were not majority stakes. Sberbank and Vneshtorgbank were the two large state banks by 1992, while others were relatively small. Likewise, Yugoslavia had the “big six,” but there are more than an additional 20 smaller banks that are state-owned and/or “socially owned.”

13

BOX 3.1 SELECTED PROFILE OF CIS STATE BANKS AFTER THE MONOBANK SYSTEM

Armenia: Armenia inherited all five state-owned banks that operated on its territory before the breakup of the former Soviet Union (FSU) in 1991. These were (i) the specialized Bank for Industry and Construction (Ardshinbank), which separated in 1991 from its FSU counterpart Promstroybank; (ii) the specialized Agrobank, also separated from its FSU counterpart in 1991; (iii) the Export-Import Bank of Armenia (Armimpex Bank), reorganized on the basis of the former Vnesheconombank, which carried out all foreign exchange transactions in Armenia; (iv) Econombank, that did not specialize in any particular industry; and (v) the State Savings Bank (Sberbank Armenian Savings Bank -ASB), which separated from its FSU counterpart (Sberbank) at end-1991 and which accounted for the bulk of household deposits in the system at that time. All of these banks, except Sberbank ASB, became incorporated as joint-stock companies in 1992, although the majority of shares either remained in the hands of the state or were sold to state-owned enterprises, thus also indirectly controlled by the state. They remained the largest banks in the country from 1993 through 1996, despite the entry of a large number of commercial banks. In 1993, the state banks accounted for more than 70 percent of the banking system's balance sheet figures. However, the banks were weak and unprofitable, with the bulk of their assets non-performing as a result of large amounts of directed credit extended under government pressure to state-owned enterprises. A major restructuring of the former state banks was initiated in 1996.

Latvia: Following the breakup of the Soviet Union, Latvia found itself in much the same position as the other FSU countries. It inherited branches of the specialized Soviet banks, namely the Savings Bank (Latvijas Krajbanka), the Agricultural Bank, the Industry and Construction Bank, the Housing and Social Development Ban, and the Foreign Trade Bank. In addition to the inherited problems of large non-performing loan portfolios and management who were unused to lending along commercial lines, the branches were suddenly cut off from their former head offices. Moreover, the banks found that the authorities in Moscow were unwilling to pass on to the newly independent branches the assets needed to cover substantial portions of their liabilities. Unlike most of the other newly independent countries that converted these branches directly into nationally owned specialized banks corresponding to the old Soviet banks, the Latvian government placed all of the branches of the specialized banks (except the branches of the Savings Bank) under the direct supervision of the Bank of Latvia (the Central Bank). These branches dominated the credit business, since the Savings Bank, initially, did not make loans to either private or public enterprises. As a result, at the end of 1991, the 45 branches controlled 83 percent of all credit to business and also held three-quarters of enterprises’ demand deposits.

Russia: In 1991, the Russian government broke up the two-tier system, consisting of the Gosbank (the central bank) and five specialized banks that had existed in the Soviet Union since 1987. This reform led to the creation of some 800 new banks, taking the capital of the previous state banks. From 1992-95, the number of banks grew enormously, with nearly 2,500 banks in operation in Russia by 1994. However, the system was also characterized by significant concentration. The largest 10 banks accounted for 50 percent of total assets in 1995. Most of these banks were spin-offs of former Soviet specialized banks, or new commercial banks created with very limited capital. The impact of these nearly 2,500 commercial banks on the real economy through lending to enterprises was relatively limited. Many of the banks were heavily involved in hard currency speculation, and then diversifying their asset holdings into treasury bills and equity stakes in blue chip enterprises. Such practices were ultimately unsustainable, and the number of banks in Russia has declined considerably. By 2000, Russia had 1,311 banks, about half of its peak in 1994.

BOX 3.1 SELECTED PROFILE OF CIS STATE BANKS AFTER THE MONOBANK SYSTEM (CONTINUED)Ukraine: Ukraine’s early transition banking system consisted of the four state-owned specialized banks that were

spun off from the corresponding Soviet banks as Ukraine gained its independence from the Soviet Union in 1991. The state banks specialized in agriculture (Ukraina), industrial lending

14

(Prominvestbank), social programs (Ukrsotsbank), and household savings (Oschadny Bank). A fifth state bank, Ukreximbank, was formed in 1992 to process Ukraine’s foreign trade payments. All specialized banks but Oschadny and Ukreximbank were corporatized (and thus nominally privatized) in 1992, primarily via “ownership transformation” whereby a number of large state-owned enterprises took substantial ownership shares in the banks that serviced their sectors. During the ensuing years, the ownership structure of these corporatized banks became more complicated due to a 1993 order by the Government that all state enterprise shares in these banks should be transferred to the Ministry of Finance. This prompted the banks to devise a method of transferring ownership through the distribution of shares to the employees of client enterprises, and of the banks themselves. Thus, ownership of the former state banks became diluted among tens of thousands of shareholders, most of them individuals. In reality, most major policy and personnel decisions were still made by top managers of the state enterprises that were majority shareholders prior to the share redistribution. In the absence of a major outside entity owning a controlling stake, this meant that the state continued to exercise considerable influence in those banks’ affairs.

Together, state banks accounted for a nominal $131 billion in total assets shortly after the dismantlement of the monobank system (or by the mid-1990s). However, the real “market value” figure is virtually impossible to estimate due to inaccurate accounting techniques, overvalued properties and loan portfolios, inadequate provisions and reserves, and general market risk that ultimately triggered numerous crises and subsequent failures. Nonetheless, based on existing data and conversion of these data to US dollar exchange rates, the total asset value of these banks was on the order of about 16 percent of GDP20 by the mid-1990s, most of it in Central Europe. On an average basis, this translated into state banks having about $654 million in assets,21

although the real average could be much smaller in light of the many small banks over which state or local governments continued to exercise influence and control. The real average would also have shrunk had IAS been applied, as these accounting standards would have adjusted balance sheets (and earnings) for non-performing loans, overvalued fixed assets and secured loans, asset revaluation from hyperinflation, and related practices that presented a superior financial position than what existed.

Most transition countries had at least three specialized state banks, and by 1992 the average among the 27 countries was about seven major state banks per country.22 As noted above, in some cases, state banks were not specialized in terms of sector. Rather, they focused on more local geographic markets. However, the vast majority of transition countries had state banks that were specialized by function, with some crossover in some cases. The following table summarizes the types of specialized state banks by country around 1992.23

20 Asset and GDP measures are used with some caution. However, state banks are estimated to have had $130,758 million in assets. GDP was roughly aggregated at $804,405 million.21 $130,758 million in assets across 200 state banks. 22 This ratio applies to major state banks. As noted in the text, many countries had smaller stakes in banks that could be technically classified as state banks, or at least as government-owned banks, including at municipal and local levels.

23 It can be noted that several specialized banks played multiple roles (e.g., many industrial banks also financed foreign operations and construction activities; some savings banks made housing loans; foreign trade banks

15

Table 3.1 Profile of State Banks in Transition Economies in 1992Industrial Agricultural Savings Foreign Trade/EXIM Other

Albania X X XArmenia X X X XAzerbaijan X X X XBelarus X X X X XBosnia X X XBulgaria X X X X XCroatia X XCzech Republic X X X XEstonia X X XFYR Macedonia XGeorgia X X X XHungary X X X XKazakhstan X X X X XKyrgyz X X X XLatvia X X XLithuania X XMoldova X X X X XPoland X X X X XRomania X X X X XRussia X XSlovak Republic X X X XSlovenia XTajikistan X X X XTurkmenistan X X X X XUkraine X X X X XUzbekistan X X X X XYugoslavia X X X

Countries with noteworthy state banks were Yugoslavia, Poland, the Czech Republic, Hungary, and Bulgaria. In particular, Yugoslavia had a substantial number of large banks—eight of the 24 largest banks24 among transition countries were from Yugoslavia, accounting for $50 billion in assets in 1991. Poland also had several large banks—seven of 24—although they only accounted for about $27 billion in assets, half that of the major Yugoslav banks. Hungary had four banks accounting for $16 billion in assets. The balance was comprised of banks from Czechoslovakia and Bulgaria. Russia surprisingly did not have banks listed in the top 1,000 on an asset basis, although 1995 figures for Sberbank suggest that it might have been one of the largest banks at

financed exporting manufacturers). Likewise, several non-specialized banks provided loans to industry and agriculture, savings facilities, trade finance, and financing for housing, construction, etc. Thus, just because a particular cell in the table is empty does not mean that the country’s banks did not provide such types of financing. Rather, the table simply highlights transition countries’ banks dedicated to particular sectors of the economy, largely reflecting their earlier focus as part of the monobank system of the central planning era.

24 These are defined as transition country banks that were among the 1,000 largest in the world in 1991.

16

the time, as might have been Vneshtorgbank. These banks were comparatively large in terms of asset size, with their rankings relative to other banks around the globe highlighted in the table below.

Table 3.2 Large State Banks by Global Asset Size Standards Early in the Transition

Bank Country Asset Size Global Ranking

Beogradska Banka Yugoslavia (Serbia) $15,983 million 287Bulgarian Foreign Trade Bank Bulgaria $14,151 million 312Sberbank Russia $13,000 million (1995) N/ALjubljanska Banka Yugoslavia (Slovenia) $9,121 million 425Komercni Czech Republic $9,085 million 426OTP Hungary $8,575 million 448Jugobanka DD Beograd Yugoslavia (Serbia) $7,542 million 475PKO BP25 Poland $6,923 million N/ABank Handlowy & Warszawie Poland $6,756 million 497PKO SA Poland $5,722 million 548Privredna Banka Sarajevo Yugoslavia (Bosnia) $5,307 million 580Vseobecna uverova banka Czechoslovakia $4,839 million 606Privredna Banka Zagreb Yugoslavia (Croatia) $4,295 million 636Zagrebacka Banka Yugoslavia (Croatia) $3,852 million 683Bank Gospordarki Zywnosciowej (Food Industry)

Poland $3,150 million 757

K&H (Commercial & Credit) Bank

Hungary $2,923 million 786

Stopanska Bank Yugoslavia (Macedonia) $2,752 million 804Magyar Kulkereskedelmi (Foreign Trade) Bank

Hungary $2,675 million 818

Vojvodjanska Banka Yugoslavia $2,357 million 860Bank Przemyslowo Handlowy (Industry & Comm.)

Poland $1,826 million 918

Budapest Bank Hungary $1,793 million 925Bank Slaski Poland $1,650 million 942Mineral Bank Bulgaria $1,420 million 968Wielkopolski Bank Kredytowy Poland $803 million 993Sources: The Banker, July 1992; Moodys Banking Statistical Supplement, 1998 (for Sberbank)

TRADITIONAL ROLES OF STATE BANKS

State banks played a number of roles in society under the socialist command economy. As discussed above, the key functions were lending to state farms and enterprises, and basic deposit mobilization and safekeeping. Above all, banks were conduits for the financing of line ministries’ production plans and targets. Centralized planning routinely set output targets by industry or sector for the achievement of national and multi-year economic goals. Once these decisions were made, budgetary resources were allocated and transmitted through the banks to state enterprises and farms. Important in this regard is that banks were essentially passive

25 PKO BP had assets of sufficient size to be ranked, but was not. It is included in the list due to its asset size.

17

administrative units, rather than credit information processors and active risk-takers operating according to commercial principles. Thus, when the monobanks were split and new second-tier state banks were created, they neither had the orientation nor the skills or experience to play an active role in imposing financial discipline on enterprises. Instead, at least in the early years of their existence, they remained administrative in their orientation, processing loans and payments based on line ministry or enterprise instructions. As their enterprise clients became increasingly subject to market forces and/or were unable to rely on the state for financial help to keep them operating, this inevitably resulted in severe loan portfolio problems for the state banks. While the state and ex-state banks were operating largely under traditional assumptions and processes, new prudential norms were being introduced as governments began to tighten monetary policies and to introduce hard budget constraints to rein in the destabilizing effects of hyperinflation and unsustainable fiscal deficits. The combination of these conflicting approaches to prudence in monetary and banking matters while the real sector was uncompetitive and suffered from breakdowns in production, trade and investment resulted in massive volumes of unrecoverable loans.

On the deposit mobilization side, banks were responsible for safekeeping citizens’ savings. In this regard, many of the traditional savings banks appear to have earned citizens’ trust, as did some of the agricultural banks. This is because the savings and agricultural banks had significant branch presence, and were responsible for record-keeping and maintenance of passbook savings, the processing of pension payments, other compensation awarded to people as part of their benefits package (mainly savings banks), and subsidy and other programs initiated at times in support of primary agricultural output and agro-processing plans. However, this trust broadly evaporated in CIS countries due to hyperinflation and the loss of savings value, combined with fiscal pressures that prevented governments from implementing successful bank bail-outs. In non-CIS countries, confidence levels varied based on the degree to which these banks could accommodate withdrawals, as well as the level of deposit protection provided in the event of a bank failure or liquidity crisis.

In this regard, the break-up of socialist Yugoslavia presented a special and distinctive set of circumstances that was somewhat different from other transition countries whose savings were lost due to hyperinflation. Banking systems in all of the ex-Yugoslav countries faced a crisis in 1992 with the freezing of foreign currency savings deposits which could no longer be honored after the central government confiscated and spent the hard currency assets funding these accounts. Slovenia and Croatia issued bonds early in the 1990s to provide some cover for these account holders. The more fragile economy of FYR Macedonia issued bonds later in the 1990s. However, depositor confidence has been more difficult to restore in Bosnia-Herzegovina, FYR Macedonia, and now Yugoslavia due to the numerous crises in the Balkans during the 1990s. The relatively new government of Serbia has committed itself to honoring the frozen deposits over a period of years. However, it remains to be seen what impact that action will have on depositor confidence. Evidence in most of the countries suggests that local citizens have little trust in domestic banks, but they are willing to place their funds in foreign banks. This has been particularly evident since late 2001 and the conversion of DM and other EU-country currencies to the new Euro.

During the socialist period and early in the transition, active campaigns to increase retail deposits were generally non-existent because private savings were limited in most countries. Exceptions

18

sometimes occurred during periods of national emergency (e.g., wars, floods), but these were government directed rather than commercially driven.26 Moreover, in countries where private savings were comparatively high (e.g., Slovenia, today’s Czech and Slovak Republics, Poland, Hungary),27 resources were often kept outside the banking system28 to avoid administrative and fiscal harassment from the authorities. Thus, while state banks played their fundamental role as safekeepers, they were ill equipped to pursue commercial campaigns to attract private savings. Beyond that, their inability to protect the value and availability of deposits and pensions was exposed in the CIS region with hyperinflation, non-indexation, and the collapse of the ruble (and subsequently other local currencies) in the early 1990s. In the non-CIS countries, shocks and losses also occurred in most countries, although the magnitude of the damage was not as great (see Tables 2.2-2.4).

GOVERNANCE, MANAGEMENT, AND OPERATING STANDARDS OF STATE BANKS

Because the state banks were still social and political units at the time they were established, governance was generally exercised through board representation from the Ministry of Finance. Often, bank managers had been trained and used to operating in the public enterprise domain, be it in banking or some other field. In many cases, managers were experienced in their particular sector of focus (e.g., industrial engineers managing industrial banks). Eventually, the shortcomings of this orientation emerged in the form of poor financial performance. In most cases, state banks continued to lend as instructed or based on patronage. This meant that their “commercialization” as joint stock companies was not sufficiently accompanied by a “commercialization” of their credit management, product development, service levels, operational efficiency, or risk management practices. Ultimately, this culminated in poor loan performance, and eventual insolvency. Such problems usually were undetected due to poor accounting and auditing standards, inexperienced supervisory personnel and inadequate prudential regulations, decentralized and often incomplete information systems (e.g., branches unconsolidated with headquarters accounts), and the traditional reliance on the government for

26 The State’s “official obligations” (similar to bonds) were a major instrument to encourage savings in the Soviet Union. There were all kinds of them, many of them issued to support a specific cause – development of the air force and navy in the 1930s, for example. The biggest effort of all was mounted during WWII – similar to what was done in the West. All these obligations turned to worthless paper after 1991. This practice was also common in other socialist countries, especially in the first post-WWII decade. For example, in the early 1950s, the Government of Hungary issued a bond called "Loan for the Peace..” People were forced to spend a certain percentage of their salaries to purchase these securities. There were no maturity dates, nor was interest paid, and the vast majority of these securities were repurchased at face value at the end of the 1960s.27 The reasons for the comparatively high savings rate in these countries vary when compared with other ex-Socialist countries. Slovenia was a major exporter in the former Yugoslavia, and, like Croatia, benefited from relatively open borders that accommodated the tourist trade, part of which was serviced by the private sector (lodging, restaurants, cafés). The Czech and Slovak populations traditionally maintained high levels of savings, with some measure of confidence induced by numbered accounts that appeared to have provided most people with a sense of privacy. Poland benefited from significant remittances from family and friends living abroad, and from informal commercial trade. Like Poland, Hungary had relatively high savings due to remittances sent from family members living abroad. 28 This was less the case with the Czech and Slovak Republics, where inflation rates were kept low, accounts were often numbered for privacy protection, and savings were traditionally at high levels.

19

additional funding when liquidity shortages materialized. Once monetary policy and prudential norms were tightened, this ultimately led to the reduced use of banks as vehicles for lending to uncompetitive enterprises, or added fiscal cost when bailing out undercapitalized and illiquid state banks.

Governance standards often deviated from “best practices” as these banks were not run according to market-based norms. Performance and standards varied from bank to bank. Some were run according to reasonably professional standards focused on generating/increasing profitability, boosting capital, managing liquidity, containing risk, and attempting to build “franchise value.” Others were poorly managed and less concerned with sustainable financial viability. Boards were often lacking in information and qualifications. Internal audit functions were underdeveloped and lacking in autonomy. Management information systems were weak. The ability to scrutinize management behavior in a timely fashion was constrained by all of these impediments. Annual shareholder meetings were often formal endorsement ceremonies rather than serious evaluations of performance. The absence of market information and involved institutional shareholders further weakened prospects for active and effective governance. Meanwhile, such weaknesses made it possible for many managers to take advantage of preferential deals that reinforced traditional networks of patronage, but undermined commercial prospects for the banks.

Operating standards and practices were generally manual, with high levels of personnel and inefficient processes. This is still evident in the employment figures for many state banks when compared to private banks. One of the key subsequent challenges for state banks has been reducing cost structures, increasing productivity and efficiency, and balancing the needs of the emerging marketplace with stakeholder claims often transmitted through workers’ or employees’ councils. Once privatization initiatives were announced for state banks, many of these banks included set-aside provisions for employees (e.g., five percent of shares) as an inducement to cost containment and modernization. However, as most employees were accustomed to earlier patterns of processing and job security, state banks generally distinguished themselves as highly inefficient when compared with new banks that emerged with better systems, and better trained and more motivated staff. Some of the state banks have since moved on to improve their performance, but this has largely been carried out as part of a needed restructuring plan prior to eventual privatization.

It should be noted that such flaws in governance, management and incentives have not been restricted to transition countries. Turkey is currently working its way out of a number of problems associated with past directed and subsidized lending (influenced by government officials), insider lending (reflecting major problems of bank governance), dangerous asset-liability mismatches (currency, maturity, interest rate), and inadequate accounting standards. These characteristics have adversely affected state banks as well as many private banks, and will cost the government well over $10 billion in the end.29 Many other countries and banks have also been affected by such imprudent practices, including those in highly developed economies. Thus, characteristics that have impaired the performance of transition country state banks are shared by non-transition country banks as well, including in advanced economies. Usually, the breakdown

29 See “Turkey: Bank Reform Progress” Oxford Analytica, January 2, 2002, and “Turkey: Banking Challenges” Oxford Analytica, May 15, 2001.

20

has been in the poor financial performance of a bank due to weak credit underwriting standards (often the result of political pressure and patronage), unsatisfactory governance, and insufficient supervisory oversight and enforcement.

TYPES OF SPECIALIZED STATE BANKS

Industrial Banks

Most transition countries had at least one major state bank focused on industry, usually with a bias towards heavy industry. This was largely to sustain production and employment levels, as well as to generate some fiscal revenues. Loss-making enterprises were often propped up because they served as a source of tax revenue (from sales proceeds in the form of turnover taxes) for government. While reliable financial data are not available for the early transition years, these banks were generally troubled loss-makers from the outset. In most cases, they have been restructured, recapitalized or liquidated. Only in a few cases have they been successfully privatized. Because they were set up to finance troubled companies, their mission was often doomed from the start, serving administrative, political or patronage-based purposes rather than commercial ones.

The following table shows the industrial banks that existed in the early stages of transition (around 1992). As noted above, several specialized banks also performed other functions (e.g., savings facilities, trade finance). Particularly in the former Yugoslavia, banks were more diversified in their activities, and the economy was less centrally planned than in other ex-socialist countries. Thus, the “industrial” banks below in Bosnia-Herzegovina, Croatia, FYR Macedonia, Slovenia and Yugoslavia were not as specialized as the industrial banks found in other countries, although these banks did have an industrial orientation in their ownership structure and lending activities. In general, the banks below were the major state banks in countries that had a particular focus on large-scale industrial enterprises, and they include “apex” development banks that were also established to help allocate or direct lending from abroad to selected companies and sectors.

21

Table 3.3 Industrial Banks in Transition Countries in 1992

Central and Eastern Europe and the Baltic States Commonwealth of Independent StatesAlbania National Commercial Bank Armenia Ardshinbank (Bank for

Industry and Construction)

Bosnia Privredna Azerbaijan PromstroibankBulgaria > 7 specialized banks Belarus BelpromstroibankCroatia

Privredna; HBRD; several smaller banks that were focused on shipbuilding finance

Georgia Industriyabank

Czech Republic Investicni Banka Kazakhstan Turan BankEstonia Bank of Industry and

ConstructionKyrgyz Promstroibank

FYR Macedonia Stopanska Banka Moldova MoldindconbankHungary Magyar Hitel; Hungarian Credit

BankRussia Promstroibank

Latvia Industry and Construction Bank Tajikistan Tajikbankbusiness; Tajikorientbank (previously Promstroibank)

Lithuania No specialized bank Turkmenistan Investbank; GasbankPoland Bank Handlowy; Polish

Investment BankUkraine Promstroibank

Romania Banca Comerciala Romana Uzbekistan UzpromstroibankSlovak Republic Priemyselna Banka; Investicni

BankaSlovenia No specialized bank, although

Nova Ljubljanska played this roleYugoslavia Jugobanka-Bor; Jugobanka-Beograd;

Beobanka Belgrade; Invest Banka;Beogradska Banka; Vojvodjanska

Agricultural Banks

While agriculture played a relatively minor role in the transition economies, a significant number of people were employed in this sector, often through state farms, collectives and cooperatives. In addition, several countries permitted households to have very small production sites for subsistence farming. As in most countries throughout the world, central planners were concerned with food security issues, stockpiling, warehousing, distribution, problems associated with post-harvest losses, etc. Trade networks were also frequently reliant on the shipment of cereals and grains, and the export of processed foods for inputs and other needed goods in exchange. For example, small countries like Armenia, Moldova and Georgia were known to ship quantities of wine and brandy to Russia. Such trade arrangements were not limited to CIS countries, as Bulgaria, Hungary, Poland and Romania also shipped processed foods to Russia in exchange for

22

needed energy supplies. Agricultural banks generally accommodated the farms, collectives and cooperatives, and often manufacturers of processed foods, beverages and tobacco. Agricultural banks also sometimes served as deposit-takers, providing basic safekeeping for rural communities that might not have had easy access to other banks. Apart from Estonia and several ex-Yugoslav countries, transition countries usually had at least one dedicated agricultural bank by 1992. These banks have generally been deep loss-makers, and some have since been liquidated. Nonetheless, they have often been protected because of the political patronage that results from close ties to agricultural or farmers’ movements, and due to the extensive branch coverage these banks often offer.

Table 3.4 Agricultural Banks in Transition Countries in 1992

Central and Eastern Europe and the Baltic States Commonwealth of Independent StatesAlbania Rural Commercial Bank Armenia AgrobankBosnia No specialized bank Azerbaijan AgroprombankBulgaria Agrarian and Cooperative Bank Belarus Belagroprombank

CroatiaNo specialized bank, although some regional banks focused on agribusiness

Georgia Agroprombank

Czech Republic Agrobank Kazakhstan KazagroprombankEstonia No specialized bank Kyrgyz AgroprombankFYR Macedonia No specialized bank Moldova AgroindbankHungary Agrobank Russia Soviet

Agroprombank (later changed name to Rosselkhozbank)

LatviaAgricultural Bank

TajikistanAgroinvestbank (“Shark Bank”)

Lithuania Agricultural Bank Turkmenistan AgroprombankPoland Bank Gospodarki Zywnosciowej Ukraine Bank UkrainaRomania Banca Agricola Uzbekistan Uzagroprombank

Slovak RepublicSlovak Agrobank (Slovenska Polnohospodarska)

Slovenia No specialized bankYugoslavia Vojvodjanska Banka (agro-processing)

Savings Banks

In most transition countries, the establishment of “new” savings banks accompanied the break-up of the monobank system. In some countries (e.g., Sberbank Russia, Ceska/Slovenska Sporitelna in the Czech-Slovak Federal Republic, Sberbank in the Kyrgyz Republic and Tajikistan, Uzsberbank in Uzbekistan), these were very “narrow” institutions30 that essentially placed all their savings in cash deposits with the central bank or other state-owned banks, or in government securities to help finance the budget as tax revenues diminished and fiscal deficits grew. In some cases, such as in CSFR and Russia, these roles changed and the banks took on more diverse

30 Some of these banks may have had small amounts of loans on their books. However, by and large, “narrow” is defined to mean savings banks that focused on savings and had limited commercial lending.

23

activities characteristic of commercial banks. However, at the outset, they were fairly narrow in their focus. In other cases (e.g., PKO BP in Poland, OTP in Hungary, CEC in Romania, DSK in Bulgaria), the savings banks were used (as before) for housing finance or other fundamental household needs. In fact, savings banks often had very basic asset-liability matching strategies under central planning, whereby long-term savings were matched with long-term housing loans (all in local currency). With strict price controls and the suppression of inflation in most transition economies, there was no need for sophisticated financial engineering strategies to manage interest rate, market or foreign exchange risk.. This provided relative calm and underlying stability at the time. Postal savings banks also existed, and were sometimes part of the larger savings bank.

Savings banks were often treated with a measure of protection because of their important financing role for the government.31 This was demonstrated in Central Europe, as they have been among the last banks privatized in most countries. For example, PKO BP remains state-owned in Poland. Ceska Sporitelna in the Czech Republic was only privatized in 2000, about the same time as Slovenska Sporitelna in the Slovak Republic. Moreover, even when countries have opted for strategic privatization, the actual structure of savings banks’ shareholdings after privatization has sometimes been more diluted than found at other banks (e.g., OTP in Hungary32), a condition which impairs effective governance. In the Baltics, the Lithuanian Savings Bank was privatized as late as 2000, while the Latvian Savings was still in state hands in 2001. In the CIS, savings banks likewise remain among the last to be privatized. However, in most CIS countries, the savings banks were particularly hard hit by hyperinflation, and most individual accounts were devastated. This undermined the integrity of implicit deposit guarantees, and public confidence in these banks remains low in many cases.33

31 It can be noted that public sector ownership of savings banks is not restricted to ex-socialist countries, and that it has been prominent in many Euro-zone countries. For example, Austria, Finland, France, Germany and Sweden all have savings banks that have been at least partly owned by central or local governments for many years.32 OTP’s shareholder structure when “privatized” was (i) the State owned 25 percent + one share; (ii) two State Social Security Funds owned 20 percent; (iii) domestic investors owned 27 percent; (iv) 100 foreign investors owned 20 percent in total (up to 2.5 percent individually); and (v) Creditanstalt and Schroeders each owned 2.9 percent. 33 Sberbank of Russia may be an exception, with substantial deposits and assets (see Box 4.1 and Annex 11).

24

Table 3.5 Savings and Postal Savings Banks in Transition Countries in 1992

Central and Eastern Europe and the Baltic States Commonwealth of Independent StatesAlbania Savings Bank Armenia Armenia Savings BankBosnia No specialized bank Azerbaijan Sberbank

BulgariaDurzjavna Spestovna Kasa Belarus

Savings Bank

Croatia No specialized bank Georgia Savings BankCzech Republic Ceska Sporitelna Kazakhstan Halyk Savings BankEstonia Savings Bank Kyrgyz SberbankFYR Macedonia No specialized bank Moldova Savings Bank

(Ekonomii)Hungary Orszagos Takarekpenztar

es Kereskedelmi (OTP) and Postbank

Russia Sberbank

Latvia Savings Bank Tajikistan SberbankLithuania Savings Bank Turkmenistan SberbankPoland PKO BP and PKO SA Ukraine Oschadny BankRomania CEC Uzbekistan Uzsberbank

Slovak RepublicSlovenska Sporitelna, Postovna Banka

Slovenia No specialized bankYugoslavia No specialized bank

Foreign Trade Banks

Foreign trade banks were common and increasingly active in the countries of Central Europe, partly because trade more rapidly shifted from the former Soviet Union to the more lucrative markets of Western Europe.34 Moreover, the latter showed early direct investment interest in countries like Hungary, the Czech Republic, and later Poland. Likewise, because Yugoslavia and Romania had been more non-aligned in their international relations, they had already established trade links with Western and other markets. Hence, foreign trade banks and export-import facilities (and in some cases, banks) emerged to encourage these trade and investment links.

In the CIS markets, there was clear interest early on in establishing trade and investment ties with Russia, as well as in some of the other CIS countries where strategic resources were found (e.g., oil and gas in Azerbaijan, Kazakhstan and Turkmenistan). However, neither trade nor investment flourished in the CIS, and only Russia received any material levels of direct investment from abroad.35 In 1993, gross international trade between the EU and CIS countries was only about 36 percent of the levels recorded in Central Europe. (As of 2000, CIS levels

34 Central European trade with the EU was $120 billion in 1993, the first year for which statistics are available for all countries. Gross trade grew to about $223 billion by 2000. 35 Foreign direct investment in the CIS in 1992 was only $226 million, of which $200 went to Ukraine. By 1995, these figures had only risen to $3.7 billion, of which $1.7 billion went to Russia and nearly $1.0 billion went to Kazakhstan.

25

were only about 33 percent of those between the EU and Central European economies.) Thus, while banks existed to accommodate trade and investment links, they did not seem to function with the product range or volume that Central Europe experienced. Moreover, many of the largest CIS companies were able to obtain financing from Western banks or markets in the form of syndicated loans and by issuing depository receipts. Thus, companies such as Gazprom, Lukoil and other CIS giants were able to go directly to the international capital markets. For smaller companies lacking comparable clout, they were often unable to penetrate Western markets due to quality or scheduling issues. This only added to their difficulties in arranging financing.

Table 3.6 Foreign Trade and Export-Import Banks in Transition Countries in 1992

Central and Eastern Europe and the Baltic States Commonwealth of Independent StatesAlbania No specialized bank Armenia Armimpex (Export-

Import Bank); possibly Econombank

Bosnia No specialized bank, although Union Bank (formerly Jugobanka) tried to play this role

Azerbaijan International Bank

Bulgaria Bulgarska Vnushnoturgovska Banka

Belarus Belvnesheconombank

Croatia No specialized bank Georgia Eximbank (made private in 1992)

Czech Republic Obchodni Banka, Zivnostenska Kazakhstan Alem BankEstonia No specialized bank Kyrgyz No specialized foreign

trade bankFYR Macedonia No specialized bank Moldova VneshekonombankHungary Hungarian Foreign Trade Bank Russia VneshtorgbankLatvia Latvian Foreign Trade Bank Tajikistan Tajikvnesheconombank Lithuania No specialized bank Turkmenistan VnesheconombankPoland Bank Handlowy; Bank for

Export Development; Bank PeKao (Bank Polska Kasa Opieki)

Ukraine Ukreximbank

Romania

Bancorex and EXIM Bank

Uzbekistan National Bank of Foreign Economic Affairs

Slovak Republic Obchodna Banka; Slovak Zarucna Banka

Slovenia No specialized bankYugoslavia Eximbank

Other State Banks

Several transition countries had additional state banks to finance infrastructure and social programs. Many of these were dedicated to housing and construction. In some cases, banks tried to stimulate small loans to households for SME development. In Hungary, Konzumbank

26

represented consumer cooperatives. However, as state budgets tightened over time, support for these banks often diminished. In most cases, these banks appeared to be relatively ineffective. For example, there is limited housing finance in most transition country markets, with Poland being the possible exception. More recently, mortgage lending has increased in countries like Hungary, Bulgaria, Croatia, and the Czech and Slovak Republics. However, by and large, there is limited lending for housing construction, and even less in the way of mortgage bonds or securitization. In other cases, SME lending has often been dependent on donor funding. More recently, in the most stable markets, commercial banks have increased their lending to SMEs, households, etc. However, the sustainability of this lending has often emerged only after serious reforms have been introduced. Under state-supported schemes, such financing did not prove to be sustainable.

27

Table 3.7 Social, Housing and Related Banks in Transition Countries in 1992

Central and Eastern Europe and the Baltic States Commonwealth of Independent StatesAlbania No specialized bank Armenia No specialized bank. The early

role of ASCB is not entirely clear. Bank for Industry and Construction and the Econombank may have provided some housing loans.

Bosnia Most banks were regional or local and provided loans for social, housing and other purposes36

Azerbaijan No specialized bank, although Promstroibank may have provided construction and housing loans.

Bulgaria Stroybank (construction); Mineral Bank (SME financing)

Belarus About 13 banks were small, geographically focused, and provided loans for housing, construction, etc.

Croatia No specialized banks, but local banks made housing and construction loans

Georgia Zhilsotsbank

Czech Republic

Kazakhstan Kredsotsbank (housing)

Estonia Estonian Social Bank Kyrgyz Zhilkomhozbank (housing)FYR Macedonia

No specialized bank, although the Macedonian Bank for Development Promotion could finance social/housing, infrastructure and other activities

Moldova Moldsotsbank

Hungary Konzumbank Russia No specialized bank, although Sberbank made loans for housing

Latvia Housing and Social Development Bank

Tajikistan No specialized bank

36 Among Bosnia-Herzegovina’s banks, 17 of 23 accounted for only 21 percent of bank assets. Considering total assets were only $3.1 billion equivalent in 1997, and that these values were overstated due to weak classification and provisioning at the time, most banks were very small, albeit “diversified” and “commercial..” Such a distribution means the average of these 17 banks only had assets of $38 million.

28

Central and Eastern Europe and the Baltic States Commonwealth of Independent StatesLithuania No specialized bank Turkmenistan Turkmenbank, Gasbank,

SenegatbankPoland PKO BP (for housing) Ukraine SotsbankRomania

CEC (for housing loans); Romania Bank for Development

Uzbekistan Four sectoral banks also provided specialized support (e.g., Phat Bank, or the cotton bank)

Slovak Republic

Slovenska Zaruchna Bank (guarantees, specialized in support to SMEs)

Slovenia No specialized bankYugoslavia No specialized bank

CHAPTER FOUR: STATE BANKS IN THE MID-1990S

THE FINANCIAL STATUS OF STATE BANKS IN 1995: DIVERGING PATTERNS OF DEVELOPMENT

By the mid-1990s after several years of difficult transition, the number of major state banks was about 200, roughly the same as figures in 1992 (if the large number of Russian banks in which the state had small stakes37 is excluded). The major change that did occur was a shift towards privatization and the presence of newly-created banks. While this had already begun in the early 1990s, by the mid-1990s, there was growing recognition of the need to restructure and privatize the major state banks for banking sector modernization. This process began in Central Europe and the Baltics, largely due to the failure of many state banks, the high cost of keeping banks state-owned, and the superior financing capacity and global information that many foreign banks brought to the domestic marketplace. These were also key ingredients for many transition countries as their trade flows shifted away from the CIS and increasingly towards the EU. Likewise, the advent of increasing trade was linked to increasing foreign direct investment, which increased the interest of foreign banks in Central European and Baltic markets. All of this created more intense competition for the state banks, and began to challenge their commanding balance sheet position in most markets. By contrast, there was limited foreign direct investment that was “strategic” and “prime-rated” in CIS banking markets by 1995, although some of the major banks did have operations in some of the CIS countries38.

37 There are no precise figures for the number of state-owned banks in Russia in 1995, given the large number of banks in which the state or other public authorities had minority stakes. However, the state continued to control the banking system from 1992-95, dominating savings through Sberbank (70 percent of household deposits) and providing directed lending to state enterprises through a number of new and regional banks. In the period 1992-95, the number of banks grew enormously so that there were more than 2,200 banks operating in Russia, although the largest 10 banks accounted for 50 percent of total assets by 1995. The impact of these 2,200 commercial banks on the real economy through lending to enterprises was relatively limited because many of the banks were heavily involved in hard currency speculation, and then diversified their asset holdings into treasury bills and equity stakes in blue chip enterprises. 38 For instance, ING, ABN-Amro, Deutsche Bank, Société Générale, Citigroup, and HSBC were among the major banks operating in CIS markets in the mid-1990s.

29

However, structural reform and ownership changes also prompted different results, as many private banks were undercapitalized, poorly managed, or utilized for excessive personal gain rather than long-term commercial viability. This was particularly true in the CIS and Balkan markets, but also true in parts of Central Europe and still evident in some of the Baltic banks. Meanwhile, the larger foreign banks catered mainly to a small segment of the corporate market, and took on only limited balance sheet risk. Thus, in neither case had private banks triggered the shift in intermediation fundamentals by the mid-1990s that policy makers had hoped for earlier in the decade. However, these fundamentals did show improvement and change in several markets a few years later, and lending flows did begin to increase based on commercial criteria by late 1995 in some countries (e.g., Poland, Hungary, the Czech and Slovak Republics).39

By 1995, banking systems had already taken different paths in different transition regions. The CIS countries continued to have a far larger number of banks, although they were much smaller on average, in terms of capital and assets. The large majority of these banks operated as “pocket” banks, subservient to their enterprise shareholders and other related and controlling interests. By contrast, the Central European and Baltic states were already consolidating their systems, actively restructuring and (in most cases) recapitalizing their domestic banks as foreign investment in the sector was materializing, or on the verge of doing so.40 With the exception of Slovenia, the ex-Yugoslav states were the primary exceptions to major bank restructuring among non-CIS countries, largely due to conflagration in the Balkans. However, even poor countries like Albania were beginning to attract foreign branches and investment, while other countries whose economies were performing poorly were still able to attract investment into the banking sector (e.g., Bulgaria, Romania).

On an average (unweighted) basis, the CIS countries each had 264 banks by 1995, of which seven were state banks that accounted for about one third of total bank assets. Total banking system assets in CIS countries were about $83 billion in 1995, of which $74 billion were in Russia alone. State banks in CIS countries had about $30 billion in assets, about 90 percent of it represented by Russian state banks.

BOX 4.1 RUSSIA’S SBERBANK IN THE MID-1990S

Sberbank became a joint-stock company in 1991 during the government’s reform and restructuring of the banking sector. This reform led to the creation of some 800 new banks taking the capital of the previous state banks. Sberbank was the largest among these banks, with its major shareholder becoming the Central Bank of the Russian Federation. Sberbank’s dominant role in the banking sector persisted throughout a period of growth, and during the rapid rise in the number of banks in Russia. During that time, Sberbank continued to play the role of a traditional state savings bank, providing retail banking services throughout the country and lending to state-owned enterprises at the national and regional levels.

39 See M. Borish and M. Noël, “Private Sector Development During Transition: The Visegrad Countries,” World Bank Discussion Paper 318, 1996. 40 The Czech Republic, Hungary, Poland and Romania had already attracted direct investment from Euro-zone and other Western banks into domestic banks or start-ups. Other countries (e.g., Bulgaria, Slovak Republic, Slovenia) were also beginning to attract limited investment interest in the form of bank branches or small capital investments in banks.

30

In contrast, the banking systems in CEE had fewer banks, but their assets were larger in value. There were 537 licensed banks in all of CEE and the Balkans, less than 25 percent of Russia’s banks alone. On average, the CEE and Balkan countries each had 45 banks, of which eight were state-owned. Total system assets were about $192 billion, or more than double the assets of CIS banks. State banks in CEE accounted for about 65 percent of banking system assets on average. Thus, state banks in CEE and the Baltics had a more prominent role than their counterparts in the CIS, with assets of about $100 billion, more than three times state bank assets in the CIS.

Part of the difference between the two regions is that asset values were more broadly wiped out by hyperinflation in CIS, whereas inflation rates were not as devastating in Central Europe as they were in CIS (and to a lesser extent, the Baltic states). A second reason is that Central European governments were more willing to recapitalize major state banks as part of broader pre-privatization restructuring programs that occurred in varying degrees of speed and magnitude through the 1990s. For example, major recapitalizations occurred at least once in the Czech and Slovak Republics, Hungary, Poland, Croatia and Slovenia, and to a lesser extent in Romania.41 In addition, some of the ex-Yugoslav states (i.e., Slovenia and Croatia) floated bonds to compensate depositors who lost foreign currency savings when the National Bank of Yugoslavia froze these accounts in 1992. For these reasons, state banks had larger balance sheets in CEE than in CIS, where asset values were generally erased with hyperinflation. A third explanation is that CIS countries sometimes set up parallel structures for their commodity-based resources (e.g., Oil Fund in Azerbaijan) that are considered strategic and essential for foreign exchange earnings. A fourth explanation is that the CIS countries gravitated increasingly to a system of arrears, barter and netting as the monetary system imploded, often bypassing the banking system (see Annex 5). Thus, by the mid-1990s, CIS countries were more likely to have much of their economic and asset values in non-bank institutions, whereas the CEE and Baltic models focused on eventually building a stable banking system. To the extent that the latter regions directed lending through state banks for preferred enterprises and farms, these practices slowly unraveled, as macroeconomic pressures called for the imposition of hard budget constraints, as new regulations required stricter bank adherence to solvency and liquidity norms, as new private and foreign banks demonstrated superior capacity, and as negotiations commenced in some cases for entry into the European Union.42

On a stock basis, asset-based measures appeared reasonable in the CEE (although lower than in advanced economies), lower in the Baltics, and microscopic in the CIS countries. All together, bank assets were about 58 percent of 1995 GDP in CEE, as compared with 126 percent among OECD countries. However, these stock figures should be treated with caution, as assets in many

41 See M. Borish, M. Long and M. Noël, “Restructuring Banks and Enterprises: Recent Lessons from Transition Countries,” World Bank Discussion Paper 279, 1995.42 While the Czech Republic, Estonia, Hungary, Poland and Slovenia were not officially invited to negotiate accession until 1998, countries in the region were aware of preliminary steps they needed to take to receive an invitation. The countries invited were considered those most likely to be able to comply with EU requirements, as outlined in the 31 chapters that are negotiated. Originally, the Slovak Republic was considered a likely invitee during the first wave, but its policies were not considered acceptable. After the change in government in 1998 and a corresponding change in policies, the Slovak Republic was invited along with Latvia, Lithuania, Bulgaria and Romania. Meanwhile, Estonia was among the first transition countries to be invited. They were not considered a likely first-tier candidate early in the 1990s when the idea of EU enlargement began to accelerate. However, the EU rewarded Estonia’s persistent reform efforts with the early invitation.

31

cases were overvalued. By the time loans, securities, real estate and other assets were more properly valued, balance sheets usually shrunk among the largest and most exposed banks. These were usually state banks that were recapitalized and restructured prior to privatization, sometimes more than once. The three Baltic states showed bank assets to be about one quarter of GDP, while the CIS countries had about 18 percent of bank assets to GDP. Given the low GDP figures in the CIS, the last statistic illustrates the severe decline in asset values resulting from the collapse of central planning, and how irrelevant most banks had become in CIS economies by the mid-1990s.

By the mid-1990s, state banks were playing less of a role in terms of lending flows, although they continued to hold a disproportionate share of total bank assets in Central European transition economies. State banks’ credit figures were higher than private banks, with many of these credit figures overvalued claims on Government, rather than loans outstanding to credit worthy and viable enterprises. State banks also often had significant shares of deposits, although their capital positions were not always as strong as at the private banks, even before adjusting for risk and capital adequacy. As countries asserted increasing levels of monetary discipline to control inflation rates, the role of state banks started to become less important. Lower levels of broad money to GDP partly reflected the imposition of monetary discipline. Meanwhile, weak loan classification, audit and accounting standards suggested that bank assets were overstated. In this case, state banks held most of the assets that would later be reclassified and written down.

The differences between bank assets and broad money indicate that virtually all transition economies were making progress in bringing down inflation rates as the basis for stabilizing the overall macroeconomic environment. Among the 12 CEE countries, most had fairly sizable gaps between the asset values posted with the banks and the level of broad money (M3) circulating through the economy, the latter being less than the former. While there are several possible explanations, these gaps broadly reflected the impact on liquidity of tightened monetary policy, and enterprise and household funds still held or circulating outside the banking system to conduct transactions without paying taxes. Meanwhile, these trends coincided with structural shifts in the banking system to encourage recapitalization by retaining earnings generated from increased interest rate spreads. In the Baltics, these ratios were lower than in CEE, but there were minimal gaps. In fact, in Estonia and Lithuania, M3 exceeded bank assets, in some ways suggesting that policy moves to restore public confidence in the banking sector were achieving results.43 In the CIS, broad money was higher than bank assets, but still low in relative terms. Particularly in the case of CIS countries, households and enterprises faced severe liquidity constraints, resulting in a shift to barter and arrears for most transactions, and dollarization of much of the economy. These tendencies reflected the deep lack of confidence of the public in both the safekeeping capacity of the banks, and the underlying value of local currencies.

43 These policy moves included introduction of a currency board in Estonia, with the kroon pegged to the DM, and comparable linkage in Lithuania between the litai and the US dollar.

32

Figure 4.1 Banks' Assets to GDP in Transition Countries in 1995 (millions US dollars)

Notes: Figures for 1995 or earliest year reported after 1995. No reliable figures reported for Tajikistan, Turkmenistan, Uzbekistan or Yugoslavia for the period. Sources: IMF (International Financial Statistics); EBRD Transition Report 2001; authors’ calculations.

STATE BANKS AND BROAD FINANCIAL INTERMEDIATION

Broad Intermediation Trends

In general, financial intermediation through the entire banking system in virtually all transition economies was low in the mid-1990s, in terms of incremental lending and deposit mobilization. Net loans had only increased $36 billion-equivalent from 1992 to 1995, little more than $1.3 billion per country on average. Given the 3,783 banks operating in transition countries, this amounted to about $10 million per bank,44 suggesting that most banks were doing very little, if any, new lending. Meanwhile, net deposits increased $60 billion-equivalent, or $16 million per bank on average.45 This suggests that most banks continued to suffer from weak funding bases, as they mobilized little in the way of additional deposits, lacked access to syndicated debt markets, and had few opportunities to increase capital from earnings or new issues. This also points to risks in the inter-bank market, as banks in relatively weak condition were often borrowers in these markets.

Broken down by region, the CEE banks showed the greatest increase in loan and deposit figures. In general, loans46 increased nearly $26 billion, of which the Czech Republic showed the greatest increase. The Czech Republic and Poland accounted for about 80 percent of the region’s net increase, while Hungary, Bulgaria and the Slovak Republic showed net declines. In the CIS

44 $36,276 million/3,783 banks = $9.6 million.45 $60,203 million/3,783 banks = $15.9 million.46 Loans are defined as banks’ claims on enterprises and households. This is separate from net domestic credit figures, which include banks’ claims on Governments.

33

58%

24%18%

0%

10%

20%

30%

40%

50%

60%

70%

CEE Baltic CIS

countries, there was a net increase of about $10 billion in loans, with Russia accounting for more than $12 billion, second only to the Czech Republic among all transition countries. Several CIS countries showed net declines, including Kazakhstan, Turkmenistan and Ukraine. Loans increased nearly $1 billion in the Baltic states.

With regard to deposits, CEE countries accounted for an increase of nearly $42 billion, about two thirds of total incremental deposits mobilized by transition country banks during this period. Poland and the Czech Republic accounted for nearly half of the total. FYR Macedonia experienced a net decline. In the CIS countries, deposits increased nearly $18 billion overall, a positive development, especially considering the devastating effects of hyperinflation in the region. However, Russia was wholly responsible for the gains, while all other CIS countries, apart from Belarus, showed limited increases at best or flat deposit levels. Several CIS countries showed net declines, including Armenia, Moldova, Turkmenistan and Ukraine. Meanwhile, the Baltic banks showed a nearly $1 billion increase in deposits, slightly more than net loans. This is important given the crash of Bank Baltija in Latvia, that country’s largest bank. However, Latvia’s performance lagged that of Estonia and Lithuania, also reflecting the 1995 banking crisis Latvia experienced. The figure below highlights basic loan and deposit trends from the early 1990s to the mid-1990s.

Figure 4.2 Growth of Loans and Deposits -- 1992/3 to 1995 (millions US dollars)

Notes: Loan figures are derived from IFS net domestic credit to enterprises and households, but exclude claims on Government; figures are for 1992 and 1995, or the earliest year in which figures are available (1993 in many cases, 1994 for Albania and Belarus); deposit figures are from IFS.Sources: IMF; authors’ calculations

Net Domestic Credit

34

$77

$25

$103

$35

$79

$29

$121

$47

$-

$20

$40

$60

$80

$100

$120

$140

CEE & Baltic's CIS

CEE & Baltics CIS

1992-3 1995 1992-3 1995 1992-3 1995 1992-3 1995Loans LoansDeposits Deposits

In terms of credit, state banks had approximately $95 billion outstanding47 in credit exposure48 at the end of 1995. This translates into about $472 million49 in credit exposure for the average state bank around that time. By comparison, the average private bank only had about $31 million in credit exposure. The largest banks on average were in the CEE region, mostly in Poland and the Czech Republic. This was true of both state and private banks, with Baltic and CIS private banks being particularly small in terms of credit exposure.

As with general asset values, loan values were broadly overstated, and some overall credit values were overstated, as demonstrated when some CIS Governments eventually defaulted on domestic debt. In most transition countries in 1995, loan quality was poor, classification standards were not strict enough or properly applied, and portfolios eventually deteriorated for a number of reasons. Had sound provisioning standards been in place at the time, the net loan figures on state banks’ balance sheets would have been smaller. Nonetheless, based on the available data, state banks had these credit figures.

Given the economic structure of most of the transition economies at the time, industrial banks are thought to have had the largest aggregate and average exposures. Among the countries covered, the industrial sector accounted for 21-42 percent of the total economy, and was 33 percent on an average unweighted basis.50 Thus, state banks in particular had more loans out to state-owned industrial enterprises than to any other sector or borrower. The second largest group of exposures appears to have been to the agricultural sector, although it is not clear what percent of state banks’ credit was allocated in this direction. By the mid-1990s, services were starting to increase significantly as a share of GDP. However, most of the enterprises responsible for this development were small private companies without access to bank loans for financing. The figure below shows credit exposure by state and private banks in 1995.

47 These figures are based on state bank shares of total banking system assets, and applied proportionally to aggregate credit figures. Aggregate credit figures include claims on Government, and are not restricted to loan exposures to enterprises and households.48 This definition is on-balance sheet, and does not account for off-balance sheet items due to data deficiencies. However, more recent reviews of the financial condition of these banks need to take such items and contingencies into account for a sound accounting of their status and risk. 49 $94,482 million across 200 state banks.50 See World Development Report, 1997. Albania had the lowest proportional figure for industrial value-added, while Ukraine had the highest at 42 percent.

35

Figure 4.3 Sources of Bank Credit Exposure in 1995 (percent)

Notes: Total number of banks are for 1995 except Bosnia; total banks for Yugoslavia are estimated for 1995; state banks estimated for 1994; asset figures are for 1995, or earliest year reported after 1995; reliable figures not available for Tajikistan, Uzbekistan or Yugoslavia.Sources: IMF (International Financial Statistics); EBRD Transition Report 2001; World Bank Policy Note ("Financial Sector Reform in Transition Countries")

As noted above, “stock” measures of credit should not be confused with new loans. While balance sheet exposures remained high to state industrial enterprises and farms, many of these loans represented delinquent loans which were rolled over without restructuring and without any material increase in collateral backing or other risk reduction actions. Moreover, even when credit was directed to loss-makers, these new funds were often used to pay down arrears to employees on wage accounts and to government accounts for social benefits. In effect, the new funds often did little to provide the enterprises with the financing needed to replenish working capital, fund capital improvements or take other actions designed to foster profitable production.. Thus, much of the “stock” measure of credit for transition countries in 1995 was old and non-performing despite what the banks’ formal books said.

To the extent that “flow” measures showed an increase in credit (as in most transition countries), these proceeds were often used to reconcile accounts and pay down arrears, rather than to invest or retool. In other cases, “controlling” interests diverted funds for uses that undermined the positions of both the debtors (borrowing enterprises) and creditors (banks). This pointed to the deep structural problems of many state bank clients, as well as weaknesses in the legal framework, and corporate governance and management practices that undermined market development. More specific to the banks, such cases aggravated the solvency and liquidity problems of state banks, and eventually reduced or eliminated their willingness to assume credit risk when they did have access to loan funds. In the end, these bank losses, flowing from the problems of loss-making enterprises, became so severe that governments were unable to fund

36

58.9

27.1 33.8

41.1

72.9 66.2

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

CEE Baltics CIS

State Private

them yet unable to come up with the needed investment capital to restructure and modernize without major job losses, debt write-downs, etc.

Assets

By 1995, state banks accounted for $131 billion in total bank assets, as compared with the private banks’ $149 billion. Thus, by the mid-1990s, private banks accounted for a bare majority of reported banking system assets in transition countries. This was particularly true in the CIS and Baltic countries, where private banks accounted for about 73 percent of total banking system assets. In the CEE countries, the figures were different, as large banks still remained in state hands. CEE countries still had 55 percent of assets with state banks, while CIS and Baltic countries had only 27 percent of assets in private banks.

The average state bank in CEE was significant even in 1995, at about $985 million. By contrast, CIS state banks averaged about $328 million in assets, while state banks in the Baltics were much smaller at $155 million.51 By contrast, private banks had only a fraction of the assets of state banks, particularly in the CIS countries. The average private bank in CEE countries had only $213 million in assets, compared with $41 million in the Baltic states and $17 million in the CIS countries.52 Thus, overall, CEE banks already had achieved critical mass by the mid-1990s, whereas only state banks seemed to have sufficient size in CIS and Baltic state markets to develop significant earnings. However, these prospects were undermined by the poor loan quality discussed above (and below in Chapters 5-6), as well as by poor service, weak systems, excess head count, lack of innovation, and a limited array of financial products.

51 CEE: $99,524 million/101 = $985 million. CIS: $30,147 million/92 = $328 million. Baltic states: $1,086 million/7 = $155 million.52 CEE: $92,852 million/436 = $213 million. CIS: $52,899 million/3,079 = $17 million. Baltic states: $2,801 million/68 = $41 million.

37

Figure 4.4 Assets of Average State and Private Banks -- 1995 (millions US dollars)

Notes: “Total” for averages for banks are averages by type of bank by ownership; total number of banks are for 1995 except Bosnia (1996); state banks are for major state banks in 1994-95; asset figures are for 1995, or earliest year reported after 1995; no reliable figures reported for Tajikistan, Turkmenistan, Uzbekistan or Yugoslavia for the period. Sources: IMF (International Financial Statistics); EBRD Transition Report 2001; World Bank Policy Note ("Financial Sector Reform in Transition Countries")

Deposit Mobilization

In addition to serving as a vehicle for lending (and on-lending), state banks were responsible for the safekeeping of household and enterprise deposits, and for payments and transfers, including benefits (e.g., pension payments, food subsidies, health, unemployment, education). Total deposits in transition country banks were roughly $108 billion in 1992-93. If most of this was held with state banks (at the time, a reasonable assumption given the prominence of savings banks and foreign trade banks for international transactions where hard currency deposits were held), this would have been as high as $570 million on average per state bank.53 However, with most deposits held in local currency and the ravaging effects of hyperinflation in virtually every transition country,54 many of these deposit values and accounts were eliminated. This was

53 $108,296 million/190 =$570 million.54 Hungary and the Czech and Slovak Republics are the only transition countries whose year-to-year average CPI from 1990 on never exceeded double-digit rates. Hungary’s average CPI peaked at 34 percent in 1991. In the Czech and Slovak Republics, the highest average annual CPI rates were 61 percent in Slovakia in 1991, and 21 percent for the Czech Republic in 1993. All other transition economies experienced triple-digit average CPI rates at least one year after 1989.

38

particularly in the case of the CIS, where no indexation was established for local currency savings while hyperinflation was frequently measured in four-digit orders of magnitude.55

By the mid-1990s, deposits held with banks were still relatively low, although they had increased about 56 percent from the 1992-93 period. While the trends were favorable, there was still recognition that significant money was held outside the banking system. Broad money measures routinely highlighted major shares of GDP circulating outside the formal system, particularly in the CIS region, while bank assets to GDP remained relatively low in most transition countries. Neither state nor private banks provided major incentives for households to place their funds with banks. Confidence in the stability of the banking system was still minimal. Interest rates paid on deposits were generally negative in real terms. The absence of credit for most households and small businesses combined with general liquidity constraints in the economy required people and firms to keep cash on hand for transactions. The general desire to avoid paying taxes likewise served as an incentive for private cash or barter transactions, rather than going through formal payment channels. Meanwhile, in several countries, there were problems with major banks that emerged in the early and mid-1990s, further complicating efforts to restore confidence. Thus, in the minds of average depositors, there were many reasons not to place funds with banks.

By 1995, total deposits held with transition country banks approximated $169 billion. A high proportion of these deposits (about 71 percent) were held in Central European banks, mostly in Poland, the Czech Republic, Hungary and the Slovak Republic. Among CIS countries, only Russia had any material deposit base. However, with 15 times the population of the Czech Republic, Russia’s deposits were only 1.2 times those held with Czech banks. Likewise, with the same demographic advantage, Russia’s deposit ratio compared to that of Hungary was only 2.6 times Hungary’s deposits in the aggregate. Thus, CIS banks had generally ceased to serve any useful savings mobilization role as of 1995, with the possible exception of Russia’s Sberbank, which accounted for at least one third of Russia’s $42 billion in banking system deposits.56 As of 1995, per capita deposits were nearly $1,000 in Central Europe, but only $280 in the Baltic states and only $165 in the CIS countries. The following figure highlights the difference in deposit mobilization among the regions by taking aggregate and state bank deposit figures and calculating per capita ratios.

55 Among transition countries, 15 of 27 countries experienced average annual CPI exceeding 1,000 percent at least once. These countries were (alphabetically) Armenia, Azerbaijan, Belarus, Bulgaria, Estonia, Georgia, Kazakhstan, Lithuania, FYR Macedonia, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan.56 According to Bank Scope data, Sberbank deposits were 65.6 billion rubles at end 1995, or $14.3 billion-equivalent. However, most reports put Sberbank’s deposit share at far higher levels. This could be more recent, with the flight of many hard currency deposits out of the country after 1995 that were previously held in other banks, as well as the loss of value of local currency deposits held in other banks after the ruble collapse in 1998 (although this would have also affected Sberbank’s household deposit base).

39

Figure 4.5 Per Capita Deposit Indicators in 1995

Notes: Total population for 1995 in millions; deposit figures are for 1995; deposit figures for banks are in millions; per capita deposits are unitary.Sources: IMF (International Financial Statistics); EBRD Transition Report 2001; World Bank data

While it is hard to be precise on deposits held by state banks, it is estimated that these banks accounted for a major share of deposits due to the role played by state-owned savings banks, and because the other large state banks that held major foreign currency assets were generally not privatized until after 1995. Even CIS countries that transformed the ownership of their banks by “privatizing” Gosbank branches still generally had foreign trade banks that were state-owned, as well as savings banks. Using the state banks’ share of total assets as a proxy for deposits held with state banks, these banks had an estimated $79 billion in deposits, or about $397 million on average,57 far lower than the $570 million average two to three years earlier. This shows that private banks had already begun to capture fairly significant deposit market share by the mid-1990s, as shown in the aggregate figures. In the CEE countries, average deposits for state banks approximated $612 million. In contrast, the average state bank in CIS countries had $185 million in deposits. In the Baltics, the average deposit base of a state bank was about $98 million in 1995.58

However, it is revealing that private banks had a majority of aggregate deposits in the region by the mid-1990s, even if private banks were smaller in terms of average deposits mobilized.59 As

57 $79,477 million/200 state banks = $397 million on average.58 CEE: $61,772 million/101 = $612 million. CIS: $17,019 million/92 = $185 million. Baltic states: $686 million/7 = $98 million.59 CEE: $58,129 million/436 = $133 million. CIS: $29,943 million/3,079 = $9.7 million. Baltic states: $1,471 million/68 = $22 million.

40

$509

$89$60

$479

$191

$105

$-

$100

$200

$300

$400

$500

$600

CEE Baltic CIS

State Banks Private Banks

with other balance sheet categories, CEE banks were markedly larger than their counterparts in the Baltic states, and above all by comparison with the average CIS bank. The figure below profiles the deposit base of the state banks by country.

Figure 4.6 Deposits in State and Private Banks (percent)

Notes: Total banks are for 1995 except Bosnia (1996); state banks estimated for 1994-95; deposit figures are for 1995, or earliest year reported after 1995.Sources: IMF (International Financial Statistics); EBRD Transition Report 2001; World Bank Policy Note ("Financial Sector Reform in Transition Countries"); authors’ calculations

Capital

Data from 1995 indicate that state banks had about $47.5 billion in “net capital”60 on their balance sheets, about 46 percent of the total bank capital in transition countries. CEE countries’ state banks had about $14 billion in capital at end 1995, mainly in Poland ($8.1 billion). Russia was the other major market where state banks had fairly significant levels of capital as a percentage of total, accounting for $7.1 billion. Croatia and Romania also had aggregate state bank capital in excess of $1 billion61. Overall, these four countries accounted for $17.5 billion in state bank capital, or about 80 percent of total transition economy state bank capital.

60 1995 capital figures are derived from IFS, with EBRD ratios of state bank assets-to-total applied. This is not exact, and likely overstates state bank capital while understating private bank capital. However, because of poor accounting standards in 1995, many private bank capital figures were also likely overstated. 61 The Czech Republic may have understated its state bank capital and overstated private bank capital. The National Property Fund had large stakes in several banks that may have been considered “private” for statistical purposes.

41

58.9%

27.1% 33.8%

72.9% 66.2%

41.1%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

CEE Baltics CIS

State Banks Private Banks

The average state bank had $112 million in capital, with CEE countries having the largest state banks. In this region, the average state bank had $143 million in capital. CIS countries had $84 million in average capital per state bank, although this was skewed by figures from Sberbank of Russia. Baltic state banks were very small in general, with remaining state banks averaging only $4 million in capital.

Private banks accounted for about 54 percent of total system capital among transition countries. Most private bank capital was in the CIS countries, with $13.3 billion, or 52 percent of total private bank capital. CEE countries had $12.1 billion in private bank capital, while the three Baltic states combined only had $193 million in private bank capital in 1995.

In terms of average size, state banks averaged $110 million in capital as compared with only $7 million among private banks. The vast majority of countries showed very low levels of private bank capital. CIS and Baltic private banks only had $3-4 million in capital on average, whereas CEE banks had $28 million on average. The largest private banks in terms of capital were found in the Czech Republic and Poland.62.

All together, 13 of 27 transition countries showed banks with majority state bank capital in 1995. Thus, it is fair to say that transition countries were generally at a mid-point by 1995 in terms of the shift in bank capital from state to private. Irrespective of ownership, the table below shows that most banks in transition countries were small, with only a handful of state banks really showing relatively large capital positions by 1995. Private banks were small, and many state banks were likewise small considering that state bank figures were skewed by the capital positions of a relatively small number of large banks.

62 Technically, FYR Macedonia showed $145 million on average per private bank. However, this is viewed as a statistical error, given the earlier hyperinflation of the former Yugoslavia, the sanctions imposed by Greece, and the general difficulties the country faced early in the transition in attending to banking sector problems. While FYR Macedonia launched structural reforms in 1995, it is the authors’ view that the average private bank capital figure was overstated.

42

Figure 4.7 Allocation of Bank Capital – 1995 (percent)

Notes: Figures are for 1995 unless not available (1997 for Bosnia-Herzegovina); “capital” = “capital accounts” +/- “other items net”; state share of capital is based on state share of assets applied to capital, with private bank shares serving as a residual. Sources: IMF (International Financial Statistics); EBRD; authors’ calculations

There was little difference between state and private banks with regard to their basic, nominal reporting of capital-to-asset ratios (which should be distinguished from capital adequacy ratios63). The data shows that state banks had 16.8 percent ratios, as compared with 17.2 percent for private banks. However, because of accounting weaknesses and weak classification standards in most transition countries in 1995, the published data do not provide enough information to determine what would have been appropriate capital-to-asset ratios in that risk environment. Due to improper classification, inappropriate assignment of risk weights, and general understatement of risks (including off-balance sheet transactions and posted collateral values), many countries had high capital-to-asset and capital adequacy ratios, only to experience major financial crises and severe deterioration of solvency once adjustments were made. Thus, capital ratios were reasonable on the surface in 1995, but proved to be low for most state banks (and often for private banks) unless they had sound backing from their owners, which then required fiscal resources, access to international capital markets, and/or some measure of monetary compensation (e.g., higher net spreads to recapitalize) or regulatory forbearance. More often than not, the monetary and fiscal measures proved costly to the economy and weakened the

63 Capital-to-asset ratios are direct balance sheet measures without adjustments for risk. Capital adequacy ratios are risk-weighted, and more accurately measure the depth and quality of a bank’s solvency. While state and private banks had roughly the same capital-to-asset ratios (at about 17 percent), these ratios would have to be adjusted for the risks of losses from non-performing assets, or overvalued assets. Had this been done, the capital-to-asset ratios for state banks and many private banks would likely have been far less. Where non-performing or overvalued assets are a problem, adjustments would eventually be netted out and the result would be smaller balance sheets (e.g., reduced assets and reduced capital).

43

46%

88%

64%

54%

12%

36%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

CEE Baltic's CIS

Private State

macroeconomic framework. In the case of forbearance, this often led to a distortion in the competitive environment by at least partly shielding state banks from market discipline.

EMERGING ROLE OF PRIVATE BANKS

The role of private banks was very limited at the outset of the transition, although a number of countries had already permitted their formation. While the largest banks through the mid-1990s were state banks, the total number of banks quickly grew once the monobank system was broken up. Thus, there were more than 2,000 private banks in the ex-socialist countries of Europe and Central Asia as early as 199164, mostly in Russia and other CIS countries. Non-CIS countries had 448 banks in the early 1990s, with Poland and Bulgaria65 alone accounting for more than one third of the total banks in CEE. The three Baltic states had 61 banks.

In a few cases, these private banks were established foreign banks. They tended to have limited balance sheet exposure to the transition countries, but were involved in international payments and transactions (e.g., trade finance, donor-financed infrastructure projects). For example, Hungary, Poland and Romania had about $2 billion in official credit exposure as of 1992 (originating in the pre-transition era), and about $2.2 billion in commercial credit exposure.66

Yugoslavia also had official and commercial credit exposure dating back to the Tito era. However, in general, the private banks early in the transition period were small start-ups, or branches privatized from the earlier Gosbank system. With low minimum capital and weak licensing standards, most of these banks were little more than captive finance companies of state enterprises that insiders and shareholders used for personal gain and patronage.

In the CIS and Baltic states, ownership transformation initially occurred with the rapid spin-off of branches of the Gosbank into private hands. For example, by 1995, the average CIS country had 153 banks, although more than 70 percent of these were in Russia.67 Among these, only seven on average were state banks, the rest being small but technically private. Of the state banks, Russia only had two that were considered major. Thus, most CIS countries actually had more state banks on average than Russia, even though Russia’s Sberbank dominated balance sheet measures among CIS state banks.

Several countries “privatized” their state banks through an “ownership transformation” process. Such an approach was very common in the CIS and Baltic states, particularly in Russia and Ukraine. This process generally involved changing the legal status of small banks that had been spun off from the monobank system to joint stock companies, and then “selling” these banks to new shareholders. This type of “ownership transformation” was a proxy for bank privatization in the absence of major capital investment into these banks. In most cases, the “new” shareholders were traditional state enterprise borrowers, and the “transformed” bank took on the character of a

64 In the early 1990s, transition countries already had 2,350 banks, most of them private (figures are from 1990, or the earliest year reported thereafter).65 Poland had 87 banks as early as 1993, and Bulgaria had 75 banks as early as 1992. 66 Poland became a member of the World Bank in 1946 and the IMF in 1986. Romania became a member of both institutions in 1972. Hungary became a member of both institutions in 1982. 67 Russia had 1,306 banks in 1991. CIS countries in total had about 1,841 banks.

captive finance company rather than a commercial bank. In some cases, these banks were “new” to the extent that they sought licenses from the regulatory authorities with “new capital” based on low minimum capital requirements for entry. Poor lending decisions, bad governance, weak management, and the collapse of local currencies all pushed these banks into financial trouble within a relatively short period.

In general, by the mid-1990s, the CIS countries plus Estonia and Latvia had most bank assets in “private” banks’ hands. However, because most private banks in CIS countries continued to operate as they had before, as “pocket” banks or captives of the key state enterprises they financed, there was little change in lending activity and general banking operations as compared to earlier years. The significance of these banks lies in their ability to return systematically to the state for automatic financing when needed. This tended to occur on a patronage and political basis, and contributed to the widespread corruption that has impaired the development process since the early 1990s.68

In CEE, the number of private banks proliferated as incentives were introduced for new private banks to emerge. In virtually all countries except Slovenia, licenses could be obtained with very low minimum capital levels.69. This fostered an easy entry policy that triggered a major increase in the number of banks in transition countries. However, in this region, a majority of banks’ assets remained in state-owned institutions, even though the number of state banks was far less than private banks. Ultimately, weak governance, insufficient risk management capacity, and underdeveloped banking supervision would culminate in several subsequent banking crises or losses. Nonetheless, the policy early on focused on stimulating competition in the banking sector by encouraging the formation of private banks while state banks continued to serve as the main financing channel (to the extent possible) for traditional enterprises and farms. When state banks ceased to be effective sources of financing, these enterprises either went bankrupt, withered away, lost all value as a result of asset stripping, or diverted financing to non-bank sources by running up arrears to numerous non-bank creditors (e.g., power and utility companies, fiscal accounts, other enterprises, employees). Meanwhile, through the mid-1990s, there was growing recognition that state banks were deeply insolvent and in serious need of restructuring to avoid recurrent and costly recapitalizations. Thus, while most major banks in Central Europe remained state-owned in the mid-1990s, many if not most of these were being restructured as part of an explicit pre-privatization exercise.

68 Corruption is frequently cited as a problem during the transition period, which it has been. However, discussions with people in transition countries also indicate the corruption was widespread throughout the communist period. Thus, in many ways, continued patronage was an extension of earlier customs and alliances. However, when the decision was made to engage in ownership transformation, some of the thinking was based on the assumption that a change in ownership would change incentives, and thus create commercially viable enterprises. Experience with bank privatization in CIS countries and mass privatization in many countries indicates that a change in ownership may well be necessary, but is generally not sufficient for a wholesale change in the incentive structure required for sustainable growth. 69 To encourage mergers and consolidation of the system, Slovenia increased its minimum capital requirement for a full banking license from DM 5 million to DM 60 million (about $35 million at the time) in 1993. By contrast, all other countries generally had minimum capital of Euro 5 million-equivalent or less. Most CIS countries had well below the equivalent of ECU/Euro 5 million.

The intermediation role of private banks was limited through the mid-1990s in the transition economies. Part of the reason was that they were generally small to begin with, with limited resources to lend. Hyperinflation erased most of the value of local currency savings in the CIS, while war and cross-border disputes (including the issue of frozen foreign currency deposits) undermined deposit mobilization potential and general banking stability in the former Yugoslavia. Thus, domestic private banks in all but a few transition countries had very poor prospects for development early in the transition. Meanwhile, larger banks from abroad were seeking only the best of the corporate clients. Apart from some syndicated lending, their activities were generally characterized by off-balance sheet transactions that would generate fee income without putting significant capital at risk.

By 1995, private banks accounted for $110 billion in credit exposure in transition countries, about 56 percent of total. On average, this is equivalent to only about $31 million in credit exposure per bank.70 Breaking this down regionally, private CIS banks were particularly small, averaging only $13 million in credit exposure per bank, as opposed to the $158 million of their counterparts in Central Europe.71. By international standards, these were all exceedingly low average levels, particularly in the CIS and the Baltics. The following table highlights average Euro-zone credit exposures in 1995 compared with regional averages for private banks in transition countries. Among Euro-zone countries, Greece, Portugal and Spain are highlighted as well to show how transition countries compared with the latest entrants to the Euro-zone with the lowest average incomes.

Table 4.8 Comparative Banking Intermediation Statistics for Private Banks in 1995 (millions US dollars)

Total No. of (Private) Banks

Credit Exposure Deposit MobilizationTotal $ Average $ Total $ Average $

Greece 103,750 65,640Portugal 116,610 97,390Spain 657,660 392,780Euro-zone Total 9,059,820 5,204,820CEE 437 68,878 158 58,129 133Baltics 68 1,547 23 1,471 22CIS 3,081 39,957 13 29,943 10Transition Total 3,586 110,382 31 89,543 25Notes: Euro-zone includes all 15 countries; averages are per bank in the EU countries and the Euro-zone in total, and for private banks among transition countries at the timeSources: International Financial Statistics;

From a deposit mobilization perspective, private banks made some modest progress in attracting deposits away from the state banks by 1995, although aggregate deposit levels remained very low. On a per bank basis, private banks in the CEE countries averaged $133 million in deposits per bank, higher than CIS private bank average of only $10 million and private Baltic bank average of $22 million.72 These deposit levels were low for a variety of reasons, including limited cash on hand among households and enterprises, tax avoidance, lack of confidence in the

70 $110,382 million/3,583 private banks = $30.8 million.71 CIS: $39,957 million/3,079 private banks = $13.0 million. CEE: $68,878/436 private banks = $158.0 million.

46

banks, low real rates paid by banks on deposits, and lack of confidence in deposit guarantees (implicit or explicit).

By 1995, private banks had about $90 billion in deposits in all transition countries. Given the total transition countries’ population of 414 million, this translated into per capita deposits held with private banks of only $216.73 While CEE per capita holdings were $479 in private banks, they were only $191 in the Baltic states and $105 in CIS countries. However, relative to state banks, private banks had larger holdings per capita in the Baltics and the CIS. There was growing convergence in the distribution of deposits in CEE, with CEE private banks accounting for more than 48 percent of per capita deposits for the region. These trends show that deposits were gradually migrating to private banks. This is important to note, considering that traditional savings banks were still state-owned in 1995 in the major CEE countries.74

Thus, by the mid-1990s, private banks were making slow progress in shifting the structure of deposits away from state banks, but these deposit levels remained low (in CIS and Baltic banks), and state banks still held large shares in Central Europe.

SUMMARY: CONSISTENCY AND DIVERGENCE THROUGH 1995

After the initial break-up of the monobank system, and particularly given the degree of macroeconomic chaos and instability that accompanied the early years of the transition, there was fairly widespread recognition of the need for greater financial discipline by the mid-1990s (and sometimes sooner). As noted in table 2.2 above, the unweighted inflation rate for non-Baltic, non-CIS countries was 460 percent from 1990-94. In the Baltic countries, the inflation rates exceeded 1,000 percent, and in the CIS countries the rates approached 5,000 percent. Hyperinflation alone led to massive shock in most countries, a problem that was compounded by fiscal decline via lost revenue, and the inability of government institutions to finance services and investments previously committed. By the mid-1990s, there was widespread recognition of the need for increased macroeconomic discipline to restore growth and confidence. However, there was already evidence of regional disparities, with CIS countries experiencing far more adverse effects, while many CEE countries and the Baltic states showed growing evidence of fiscal discipline and reduced vulnerability to hyperinflation.

By 1995, hyperinflation and fiscal deficit figures were coming down, partly due to the hardening of soft lending conditions through the banks. In general, confidence was far more devastated in the CIS countries than elsewhere, with the possible exception of the former Yugoslavia, where years of hyperinflation (dating back to the socialist era), the freezing of foreign currency savings accounts, and war led to a major loss of faith in civil institutions (including banks in most cases).

72 CIS: $29,943 million/3,079 private banks = $9.7 million. CEE: $58,129 million/437 private banks = $133.0 million. Baltics: $1,471 million/68 private banks = $21.6 million. 73 $89,543 million/414 million = $216.

74 Examples include PKO BP and PKO SA in Poland, OTP in Hungary, Ceska Sporitelna in the Czech Republic, Slovenska Sporitelna in the Slovak Republic, CEC in Romania, and DSK in Bulgaria.

47

These macroeconomic developments triggered (and were reinforced by) a variety of structural challenges involving financial as well as institutional and incentive issues. From a financial resource perspective, banking systems were faced with portfolio erosion, declining lending flows, overvalued fixed assets, the loss of confidence in the safekeeping capacity of banks, general savings decline, the absence of other borrowing sources, and persistently weak capital. From an institutional and incentive perspective, banking performance in the absence of protection and support suffered from (i) poor corporate governance, weak internal systems, inadequate management, and related party abuse; (ii) an inadequate legal framework for secured transactions; (iii) weak formal debt collection and liquidation systems; (iv) inadequate accounting standards; and (v) weak information on most enterprises relative to modern underwriting requirements. Such instability made it virtually impossible to move quickly to a financial system that was stable, sound, and commercially viable. This fact was compounded by the significant cost of systems and human capital development required for such a transition to be effectively implemented.

In many CEE countries and the Baltics, the tightening of monetary policy was accompanied by a stricter prudential regulatory framework for banks, with particular emphasis on loan classification, provisioning standards, and more accurate accounting of profitability or loss recognition, retained earnings, and capital measures (including more suitable risk weights applied in capital adequacy measures). In many of the CEE countries and in the Baltics as well as some of the CIS countries, the tightening of the prudential framework was accompanied by efforts to strengthen banking supervision capacity. These efforts focused on early warning signals of financial sector instability, general off-site surveillance from regulatory reporting, and the coordination and scheduling of comprehensive on-site examinations.

Initial banking supervision measures were taken in many countries prior to 1995. However, they were not effectively or sufficiently implemented until later. This was often due to the time needed to develop institutions and personnel for a broad range of functions (e.g., design of adequate and standardized reporting forms, development of systems to detect early warning signals, off-site surveillance, on-site inspections, coordination across departments, policy and strategy), as well as difficulties supervisors often had in being able to execute their mandate. The latter was particularly the case when state banks violated prudential norms and were out of compliance, or when “private” banks with close ties to government officials obtained special favors from these officials. In other cases, these mandates were simply not given enough legal strength, at least not until a major banking crisis occurred. Above all, banking supervisors often found weak political support for their mandate. This changed in subsequent years after banking crises had occurred in most transition countries, and as international institutions moved to encourage more intensive observance of standardized norms in support of financial sector stability.75

By 1993-94, the performance of CIS and non-CIS countries was already beginning to diverge. In many non-CIS countries, recognition of solvency and liquidity problems triggered a series of recapitalizations and restructuring programs geared to restoring stability in the banking system, and getting banks on track to be profitable on a commercial basis. Poland’s restructuring

75 This effort accelerated after the East Asia crisis in 1997. Urgency was reinforced in 1998 with the collapse of the ruble and the deleterious effects this had on CIS economies.

48

program in 1993-94 was followed in the late 1990s with a surge of strategic investment and privatization. Hungary’s declining fiscal and balance of payments fundamentals by 1994 prompted an acceleration of privatization throughout the economy from 1995-96 on, including a preference for strategic investment in the banking sector. Estonia moved aggressively to liquidate weak banks in the early 1990s and to consolidate in the late 1990s, attracting foreign investment from Scandinavia and Europe as an anchor. Where problems were not identified and addressed early on (such as in Latvia in 1995, Bulgaria in 1996-97, and Albania in 1997), major collapses subsequently prompted intensified reform efforts to pre-empt a recurrence of widespread instability. In all of these cases, governments moved to restructure their banks under strict guidelines, and with a clear and specific objective to privatize, usually with some form of strategic investment.

In other countries, where reforms and performance lagged, the approach has been different. Often, state banks have been kept afloat because of their (often erroneous) perceived importance to the economy. The tables above show that state shares of banking system assets were still fairly high in most countries in 1995. However, in addition to the state shares, many countries showed a significant portion of banks that had earlier been “privatized” by ownership transformation, rather than through strategic investment. This resulted in a continuation of many of the lending practices that had gotten the state banks into financial trouble. Invariably, the results have been high levels of non-performing assets, a drain on the budget and the economy, distortions in the competitive environment, and an erosion of confidence in civil institutions.

49

BOX 4.2 LATVIA’S SUCCESSFUL RESTRUCTURING AND PRIVATIZATION OF UNIBANKA

The case of Unibanka represents a rare success story. While it initially engaged in activities that undermined the quality of its loan portfolio and put bank capital at risk, its restructuring exercise proved successful, as reflected in its ability to withstand systemic weaknesses in the mid-1990s and to attract strategic investment in the second half of the 1990s.

On September 28, 1993, the rump of 21 branches from the newly reorganized Savings Bank was structured into one state bank – the Universal Bank of Latvia, or Unibanka. Most of the bad loans were concentrated in the branches that constituted this bank (40 percent of total assets in March 1994). As part of the rehabilitation process, these loans were taken off Unibanka’s books and replaced with seven-year Government bonds in the amount of LVL 25 million ($50 million equivalent).

One of the main reasons cited by the Government for creating Unibanka was to provide an insurance policy against catastrophic failures in the private banking sector. This logic was put to serious test in the first half of 1995, as the insolvency of the country’s largest bank (Bank Baltija) triggered the systemic crisis in which about 40 percent of the assets and liabilities of the banking sector were lost, and seven banks, including three of the 10 largest banks, had collapsed. Although the crisis had a large effect on both large state banks, Savings Bank and Unibanka, the latter was not directly involved in the crisis, did not need to be closed or bailed out, and was not as badly harmed as the Savings Bank. In fact, Unibanka benefited (and the Savings Bank suffered) from a flight to quality following the crisis, as depositors reallocated assets towards banks that appeared better managed, more strongly capitalized, and less risky in the composition of their portfolios. Surveys of Latvian banking professionals consistently rated Unibanka as the safest bank in Latvia.

Privatization procedures were launched at Unibanka on October 3, 1995. The board of the Latvian Privatization Agency (LPA) approved the bank's basic privatization regulations, which provided that Unibanka would be privatized in four years. During 1995, the first stages of Unibanka’s privatization were carried out. Share capital was increased to LVL 11.5 million (about $23 million), and then a little over 50 percent of the shares were sold for privatization certificates. Twenty-two percent of shares were sold publicly; 13.5 percent were sold to customers of Unibanka; and 14.5 percent were sold to employees. The LPA held the remaining shares. In October 1995, the bank's shareholders meeting decided to reorganize the bank into the joint-stock company, Latvijas Unibanka, and a new charter was approved for the bank. In January 1996, Unibanka became the first company to be listed on the Riga Stock Exchange official list.

According to the bank's privatization regulations, its share capital was increased during the next privatization round by attracting additional capital from a strategic investor. In May 1996, Unibanka’s share capital was raised by LVL 6 million (about $12 million), and the EBRD and Swedfund International AB purchased the newly issued shares. This gave the EBRD control of about 22.6 percent of total shares and Swedfund control of about 7.5 percent of shares. Over the next three years, most of the remaining state-owned shares were sold in the international market through a GDR program, and a part of the shares was sold through special auctions at the Riga Stock Exchange. Overall, the state received LVL 66.1 million (about $113.4 million) by the time privatization was complete in late 1999, including LVL 21.3 million in cash and LVL 44.8 million in privatization vouchers.

By September 2001, Unibanka's paid-up share capital was LVL 37.1 million ($59.9 million). More than 98 percent of share capital belongs to the Swedish bank Skandinaviska Enskilda Banken (SEB). A major force in the general trend towards banking sector consolidation in the Baltics, SEB initially purchased a 23 percent interest in Latvijas Unibanka at a special auction held in the stock market in late 1998. It then steadily purchased shares from the other bank's shareholders, including the EBRD’s shares.

50

CHAPTER FIVE: CURRENT STATUS AND CONTINUING PROBLEMS OF STATE BANKS SINCE 1995

CURRENT OWNERSHIP STATUS AND TRENDS

Recognizing the damage done to the economy with the continued existence of state banks, many transition countries have moved with greater intensity to effectively privatize their banking systems. For example, Albania will have a fully privatized banking system by mid- to late 2002, just a few years after having nearly 100 percent of assets controlled by state banks. The damage done by the pyramid schemes in 1997 served as a trigger for this acceleration. While nowhere nearly as damaging to the economy or civil society, the build-up of fiscal and balance of payments deficits in Hungary by 1994 served as a catalyst for its acceleration of privatization from 1995-96 on in financial services as well as the real sector. Poland is now down to two major state banks, and Hungary and the Czech and Slovak Republics have finalized banking sector privatization of major institutions with strategic investment.76 Bulgaria and Croatia recently privatized their largest state banks, and FYR Macedonia and the Baltic states have very little state investment remaining in their banking systems.

Figure 5.1 Number of State Banks -- 1992-2001

Among CIS countries, Armenia77 and Georgia have fully eliminated state ownership. The Kyrgyz Republic only has three banks that are non-private. Moldova will soon be fully private when it sells its last shares in the Economic Bank (the former savings bank). Kazakhstan has

76 In all four countries, additional smaller banks remain state-owned. However, they are not viewed as “major” competitors to the private banking system.77 The last state-owned bank, the Armenian Savings Bank, was privatized in 2001. Prior to that, four other state banks had been rehabilitated by 1998 and subsequently privatized.

51

128 118

6744

73 92

43

43

0

50

100

150

200

250

1992 1995 2000 2001

CEEB CIS

likewise significantly reduced the share of state bank assets since 1998. The following table provides a snapshot of the evolution of state banks through the transition process, as well as plans of individual country governments as of 2001 for their remaining state banks.

TABLE 5.1 EVOLUTION OF STATE BANKS THROUGH THE TRANSITION PROCESS: 1992-2001

Current Plans and StatusAlbania Savings Bank to be privatized in 2002.Armenia ASB privatized in 2001.Azerbaijan United Universal still undergoing consolidation; IBA still state-owned.

BelarusSeveral banks remain state-owned without any formal program to move forward with privatization.

BosniaMost state banks being privatized or liquidated in 2002, although progress is slow in some cases.

Bulgaria Two of last three state banks being offered for sale in 2002.Croatia Only HRB remains state-owned.

Czech RepublicThe major privatizations took place in 1999-2000 with CSOB and Ceska Sporitelna. Four banks remain state-owned.

Estonia System fully privatized.FYR Macedonia Only MDB remains state-owned.Georgia No state banks remainHungary Two banks remain state-owned.

Kazakhstan

Government reduced its 80% stake in Halyk (as of December 1999) to a current level of 33.3% plus one share. The State initially planned to sell its remaining stake in 2001. However, this tender has since been postponed indefinitely.

KyrgyzKairat 100% state-owned; Energo Bank partly state-owned; Savings and Settlement Company, a state financial institution with a limited banking license.

Latvia Full privatization planned for Latvian Savings Bank.Lithuania Only the Agricultural Bank is state-owned. Moldova System to be fully private after sale of Bank Economii.

PolandTwo large banks remain state-owned (PKO BP and BGZ) plus two other banks that are comparatively large for the region.

Romania

Three banks remain state-owned = about 40% of assets; privatization planned for one (BCR), restructuring and eventual privatization planned for the savings bank (CEC), and reorganization (and de-licensing) planned for EXIM Bank.

Russia

> 460 banks are state-owned, and as many as 679 have shares/stakes from all public institutions (including the central bank); the state (all-inclusive) held controlling stakes in 62 and blocking shares in 8078; state plans to divest all holdings of less than 25 percent, leaving Sberbank, Vneshtorgbank, Vneshekconombank and a small number of new specialized banks (export-import bank, agricultural bank, development bank).

Slovak RepublicSeveral small banks remain state-owned, but these are not viewed as highly distortionary. Major state-owned bank left is Postovna Banka.

Slovenia Slovenia’s major bank remains state-ownedTajikistan Sberbank is the only remaining state bank.Turkmenistan Very limited data, but five state-owned banks remain.Ukraine Two state banks remain, including Oschadny (savings bank).

UzbekistanOne bank fully state-owned (National Bank of Uzbekistan), but 15 are joint-stock banks with direct or indirect state ownership.

78 One report claims that as of October 1, 2001, Russia had more than 1,300 lending institutions, of which 638 were at least partly owned by the state when including shares of the central bank in commercial banks and other lending institutions. See M. Builov, “Who Owns Russia: Russia’s banking sector. The situation today.,” The Russia Journal, January 25, 2002.

52

Current Plans and Status

Yugoslavia

Four state banks being liquidated; however, the Government just opened a state-owned Savings Bank, with plans to open a few more state banks. In addition to the "Big Six" listed for 2001, there are a number of smaller banks (around 20) in which the majority of assets are controlled by the state or socially-owned enterprises and banks. Most of them are deeply insolvent and are scheduled for liquidation in the near future.

Altogether, there has been a substantial decline in state bank assets in the transition countries. As of end 1996, state bank assets were estimated to be about $125 billion, although these figures exclude Bosnia-Herzegovina, Tajikistan, Turkmenistan, Uzbekistan and Yugoslavia. (This would likely bring the total to about $130 billion or so.) Using the same measures, by 2000, this estimate had dropped to about $88 billion,79 nearly a one-third decline. The decline has primarily resulted from privatization of banks, plus balance sheet shrinkage in many remaining state banks where provisions have been more stringently applied to troubled loan portfolios, overvalued fixed assets and real estate have been sold or more appropriately valued, deposits have shifted to other banks, government and central bank financing (via loans or deposit placement) has diminished, and capital has been adjusted for declining asset values and reversals of income and accruals.

However, several countries still retain state banks in the hope that their franchise value will increase with time and eventually generate higher privatization proceeds), or more immediately, to continue to serve as vehicles for directed credit financing. As of 2000, countries that had state bank asset shares in excess of 20 percent (without major subsequent ownership changes) included Azerbaijan, Belarus, Lithuania, Poland, Romania, Russia, Slovenia, Turkmenistan, and Uzbekistan.80 Still other countries have a smaller share of bank assets in state hands, yet run the risk of systemic problems due to the strategic nature of the remaining state banks and their use for political patronage purposes (e.g., Ukraine). The following figures highlight trends in the state share of bank assets for transition countries since 1996, along with estimated dollar values for these assets.

79 These figures are based on EBRD estimates of state bank assets as a share of total bank assets. These figures differ from tallies from the financial statements of state banks. For example, in 2000, state banks’ financial statements in transition countries showed assets approximating $108 billion. Using IFS figures as the denominator, this would mean that state banks had a slightly higher proportion of total banking system assets. However, many of the additional bank assets are in banks that are currently being liquidated, such as more than $6 billion in “major” banks in Yugoslavia. Thus, the figure is probably closer to about $100 billion, but could be closer to the $88 billion approximation once adjusted for write-offs, etc.80 Albania, Bosnia-Herzegovina, the Slovak Republic and Yugoslavia had high levels (more than 50 percent) of state ownership in bank assets in 2000, but have since moved on with privatization and liquidation that has significantly reduced these shares.

53

Figure 5.2 Share of Bank Assets Held by State Banks: 1996-2000

Notes: Most recent percentage used for earlier years when not available (e.g., Bosnia-Herzegovina, Moldova, Russia, Turkmenistan, Ukraine, Uzbekistan, Yugoslavia); Hungary 2000 is 3Q.Sources: IMF; EBRD; authors’ calculations

Figure 5.3 Volume of Assets Held by State Banks 1996-2000 (US$Eq millions)

Notes: Most recent percentage used for earlier years when not available (e.g., Bosnia-Herzegovina, Moldova, Russia, Turkmenistan, Ukraine, Uzbekistan, Yugoslavia); Hungary 2000 is 3Q.Sources: IMF; EBRD; authors’ calculations

54

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

1996 1997 1998 1999 2000

CEE Baltics CIS

$-

$10

$20

$30

$40

$50

$60

$70

$80

$90

$100

1996 1997 1998 1999 2000

CEE Baltics CIS

FINANCIAL CONDITION OF THE STATE BANKS

General Financial Performance Indicators

In 2000, state banks generally accounted for about one third of credit, assets and deposits, but had far less capital (about 21 percent of total). In terms of solvency, state banks’ capital ratios are far lower than those of private banks, although this cannot automatically be equated with lower quality or greater risk. In the CEE countries, many of the state banks are simply holding Government securities and making fewer loans, both in recognition of the high risks flowing from the impaired business climate in those countries and to more easily comply with prudential norms governing liquidity and capital. Meanwhile, after-tax earnings have been relatively low in the CEE and Baltic states. CIS countries show better returns, although this conclusion is suspect due to weak accounting standards. However, on a more positive note, there was a substantial increase in capital in Russian banks, both state and private, in 2000. If the accounting techniques used are accurate, this would suggest that Russian banks strengthened their solvency in 2000 and positioned themselves to be more efficient and competitive henceforth. This, in turn, would presumably benefit CIS indicators in general, as Russia is the dominant economy in the region.

In general, asset growth among transition country banks was primarily focused on Poland and Russia, with little net increase among the other countries. Asset growth occurred in both the state and private banks, suggesting this was partly a function of larger macroeconomic trends rather than strictly a structural development. That these two countries represented such a substantial part of the overall aggregate trend for the region also reflected the continued clean-up and consolidation that occurred in the Czech Republic and Hungary, where banks’ asset growth was negative in the aggregate in 2000.

In terms of general financial profile, state banks in 2000 had about $108 billion in total assets. (See Annex 1) This approximated 14 percent of total GDP for the countries.81 Most of the state banks had less than $1 billion in assets, although 20 state banks had more than $1 billion in assets.82 In five of these cases, the banks have either since been privatized (Komercni and VUB) or are in the process of being liquidated (Beogradska, Jugobanka, Invest Banka).

As for loans (see Annex 1), state banks showed about $46 billion in loans by end 2000, or 6 percent of recorded GDP.83 About 10 banks accounted for 71 percent of these loans,84 with more

81 $108 billion/$768 billion = 14.1 percent.82 These were (from high to low) Sberbank in Russia, PKO BP in Poland, Komercni in the Czech Republic, Nova Ljubljanska in Slovenia, Vneshtorgbank in Russia, BGZ in Poland, National Bank for Foreign Economy in Uzbekistan, VUB in the Slovak Republic, Banca Comerciala in Romania, Vneschekonombank in Russia, Vneshekonombank in Turkmenistan, Beogradska Bank in Yugoslavia, Jugobanka in Yugoslavia, Nova Kreditna Maribor in Slovenia, Invest Banka in Yugoslavia, Moscow Municipal Bank in Russia, SKB in Slovenia, Savings Bank in Albania, Postbank in Hungary, and Ceskomoravka Zarucni in the Czech Republic.83 $46 billion//$768 billion = 6.0 percent. These figures include the $6 billion or so in loans reported by the big banks in Yugoslavia. These figures are excluded from several tables due to the liquidation process in place for the major banks in Yugoslavia. If the $6 billion from Yugoslavia is excluded, the ratio is only 5.2 percent.84 $32.6 billion/$46.3 billion = 70.5 percent.

55

than $1 billion in loans posted on their balance sheets.85 In four of the cases above, the banks have either since been privatized (Komercni and VUB), or are in the process of being liquidated (Beogradska and Jugobanka). If these four banks were excluded, loan figures from the major state banks would have been reduced by about $8 billion.

On the funding side (see Annex 1), deposit figures were about $82 billion, or 11 percent of GDP.86 With total broad money for transition economies at 33 percent of 2000 GDP,87 this suggests that state banks’ deposits account for about 32 percent of total transition country broad money.88 Using the same figures, this suggests that there is substantial liquidity among the major remaining state banks relative to their exposures, with loans only 56 percent of deposits. However, this needs to be evaluated on a case-by-case basis, particularly as some banks are exposed in the inter-bank market and at risk, or dependent on Government deposits for funding. This also suggests that many banks are placing their deposits in Government securities rather than in lending activities. As discussed above and below, this is often prudent and helps banks comply with regulatory liquidity and capital norms. However, it also reflects the use of state banks as sources of financing for fiscal and quasi-fiscal activities that often weaken prospects for economic growth and competitiveness. There was even higher deposit concentration than on the asset side, as 14 banks accounted for 81 percent of total state bank deposits.89

In terms of capital (see Annex 1), state banks showed only about $10 billion in total, or 9.3 percent of assets. Most state banks are very small, considering that six state banks with more than $500 million-equivalent in stated capital accounted for 54 percent of total capital.90 This would leave a remaining $4.6 billion in capital spread across 71 banks reporting figures (including some that have since been privatized or are being liquidated). Without accounting for adjustments for classified assets or other charges, the average state bank at the time had about $65 million in capital after excluding these six major state banks.

85 These banks were (from high to low) Sberbank in Russia, PKO BP in Poland, Komercni in the Czech Republic, Nova Ljubljanska in Slovenia, National Bank for Foreign Economy in Uzbekistan, BGZ in Poland, VUB in the Slovak Republic, Vneshekonombank in Turkmenistan, Beogradska Bank in Yugoslavia, and Jugobanka in Yugoslavia. Noteworthy in this regard is the number of banks with large assets that do not have major loan exposures on their books.86 $82 billion/$768 billion = 10.7 percent.87 $255 billion/$768 billion = 33.3 percent. Figures for Turkmenistan are from 1999. All other country figures for broad money and GDP are from 2000.88 $82 billion/$255 billion = 32.1 percent.89 $67.0 billion/$82.5 billion = 81.2 percent. The major state banks with more than $1 billion in deposits at end 2000 were (from high to low) Sberbank in Russia, PKO BP in Poland, Komercni in the Czech Republic, Nova Ljubljanska in Slovenia, BGZ in Poland, VUB in the Slovak Republic, Vneshtorgbank in Russia, Vneschekonombank in Russia, National Bank for Foreign Economy in Uzbekistan, Banca Comerciala in Romania, Moscow Municipal Bank in Russia, Nova Kreditna Maribor in Slovenia, Savings Bank in Albania, and SKB in Slovenia.90 $5.4 billion/$10.0 billion = 54.0 percent. The state banks with more than $500 million in capital were (from high to low) Vneshtorgbank of Russia, Sberbank of Russia, National Bank for Foreign Economy in Uzbekistan, Komercni in the Czech Republic, PKO BP in Poland, and BCR in Romania.

56

In terms of earnings (see Annex 1), state banks reported about $329 million in after-tax earnings in 2000. However, apart from Sberbank (Russia), PKO BP (Poland), Vneshtorgbank (Russia) and BCR (Romania), earnings were meager. Only eight state banks in total reported after-tax earnings exceeding $20 million. Thirteen banks reported losses, and 33 reported earnings at or barely above the breakeven point.91 Annex 2 provides a range of financial ratios for state banks that includes return measures, net interest margins, some loan volume and quality indicators, as well as some funding figures.

Overall, state banks still possess sizable market share in many countries. (See Annex 3 for shares of GDP, and Annex 4 for market shares.) For example, most banking system assets in Albania, Azerbaijan, Belarus and Slovenia, and about half in Romania, were state-owned at end 2000. Belarus was particularly high, with its six major state banks accounting for more than 90 percent of total assets. Loan share is a little different, with Belarus being the only country where state banks account for a majority of loans. This indicates that private banks are emerging as the major lenders, and that state banks’ earning assets (to the extent they are actually generating income) are more often in Government securities, fixed assets, or other items. On the deposit side, Albania, Azerbaijan, Belarus, Lithuania,92 Russia and Slovenia had majority state bank shares, with Bosnia-Herzegovina and Romania having about half of total deposits in state banks’ hands. Meanwhile, in terms of capital, only Belarus and Romania were majority state-owned, while Slovenia was about half state and half private. It may well be that state banks also accounted for the majority of assets, loans, deposits and capital in Tajikistan, Turkmenistan, Uzbekistan and Yugoslavia.

Loans and Net Domestic Credit

State banks are less prominent in active lending to the real sector than are private banks. Data from 2000 indicate that state banks had about $82.5 billion in “net domestic credit” on their balance sheets, which essentially accounts for bank loans and investment in Government securities. Of this total, about $40 billion was estimated to be loans to households and enterprises, and $42.5 billion93 was in the form of claims on Government (mostly securities investments). All together, this amounted to about 36 percent of total net domestic credit among transition countries.94 State banks are more actively engaged in Government financing for two major reasons. First, the banks themselves may be in weak condition, with their balance sheets reflecting Government securities that are earning assets to help them recapitalize. Second, state banks that are major deposit mobilizers (e.g., savings banks) often serve as a source of financing for Governments that are using the banks for budgetary and extra-budgetary financing.

91 Barely above breakeven includes earnings from $0-$2 million.92 Lithuania is included only by combining Savings Bank figures with those of the still state-owned Agricultural Bank. The Savings Bank of Lithuania has been privatized since end 2000.93 This figure differs slightly from some of the above-mentioned loan figures, due to differences in sources (e.g., IMF, Bank Scope) and methodologies. However, the differences are not considered material for purposes of the analysis.94 $82,537/$228,717 = 36.1 percent.

57

About two-thirds of the net domestic credit total for state banks in 2000 was in the CEE countries,95 as this is where most of the largest state banks remain (apart from Russia). Russia and Poland alone accounted for half of the total, largely due to Sberbank and PKO BP. These two countries, combined with the Czech Republic, Slovenia and the Slovak Republic, accounted for 78 percent of the total. Uzbekistan also has a large volume of state bank credit.96

However, further analysis shows that less than half of the “net domestic credit” of CEE state banks is actually in the form of loans. In fact, these state banks had only 44 percent of their total credit in the form of loans, as compared with private banks in the region that show 87 percent of net domestic credit in the form of loans. This is lagging trends in the Baltic states, where the three countries had a higher proportion of overall loans to net domestic credit, and where private banks’ loans accounted for 84 percent of their total net domestic credit. In the CIS countries, only 54 percent of state banks’ net domestic credit was in the form of loans, while private banks showed about 78 percent. All together, 71 percent of net domestic assets were in the form of loans, mainly from private banks in all three regions.

From a risk management perspective, this suggests that CEE state banks are more vulnerable to Government (i.e., domestic debt) risk than to enterprise or household risk, while the reverse is true for the Baltic states and the CIS. Meanwhile, private banks in all three regions are more vulnerable to the enterprise and household sectors than are state banks, particularly in CEE and the Baltics. In the CIS, private banks hold about 22 percent of net domestic credit in the form of Government securities. Considering past government defaults in some CIS countries, this suggests that private banks need to be on their guard as well concerning their exposure to Government securities, and that this is not just a risk for state banks. The figure below shows the relative share of loans to net domestic credit at end 2000 by ownership and region.

95 $51,521/$82,537 = 62.4 percent.96 Banks that played a major role in 2000 net domestic credit figures from these countries include Komercni (Czech Republic), Nova Ljubljanska (Slovenia), VUB (Slovak Republic), and the National Bank for Foreign Economic Activity (Uzbekistan).

58

Figure 5.4 Comparative Distribution of Loans as a Percentage of Net Domestic Credit: 2000(million US dollars)

Notes: Private loan figures were derived from total loans to state enterprises and the private sector (from IFS) less loans on state banks’ balance sheets. Sources: IMF; Fitch IBCA; authors’ calculations

On average (on an unweighted basis), Central European countries had about $4.7 billion in state bank credit, while CIS countries averaged $2.8 billion. Meanwhile, the loan share of net domestic credit by region was $2.1 billion on average in Central Europe and $1.5 billion in CIS countries, consistent with the figures above that show that less than half of state banks’ earning assets are in the form of loans. The Baltic states’ figures were negligible, and have become even less since 2001 after the privatization of Lithuania’s Savings Bank.

CIS countries show a higher share of state bank financing of net domestic credit on average, at about 44 percent. The CEE countries have about one third of total net domestic credit on state banks’ balance sheets. Again, the Baltic state banks were relatively insignificant at this juncture, accounting for about 12 percent of total net domestic credit at end 2000. The following table highlights these figures, along with the share of loans.

59

Table 5.2 State Banks' Loans and Total Net Domestic Credit Exposure at end 2000(million US dollars)

Total Bank Credit ($)

State Bank % of Credit

State Bank Credit ($) o/w Loans ($)

Albania 1,312 91.4 1,199 10Armenia 229 1.3 3 3Azerbaijan 544 93.6 >509 185Belarus 1,405 101.2 1,422 1,075Bosnia 2,064 7.8 161 89Bulgaria 2,512 32.2 809 429Croatia 9,741 14.1 1,375 647Czech Republic 30,578 36.0 11,002 3,752Estonia 1,407 0.0 0 0FYR Macedonia 755 1.1 8 8Georgia 229 0.0 0 0Hungary 18,230 9.1 1,657 599Kazakhstan 2,750 24.0 661 385Kyrgyz 60 10.0 6 2Latvia 1,683 18.3 308 149Lithuania 1,983 14.8 293 201Moldova 211 12.8 27 13Poland 64,795 32.1 20,813 10,179Romania 3,865 92.3 3,566 1,015Russia 52,100 41.1 21,400 10,234Slovak Republic 12,497 35.3 4,408 2,723Slovenia 8,877 74.3 6,600 3,499Tajikistan N/A N/A N/A 3Turkmenistan 1,887 104.0 1,962 1,803Ukraine 3,963 13.1 519 322Uzbekistan 5,040 77.5 3,906 2,525Yugoslavia N/A N/A N/A N/ATOTAL 228,717 36.1 82,614 39,850CEE total 155,226 33.2 51,598 22,950Baltic total 5,073 11.8 601 350CIS total 68,418 44.5 30,415 16,550Notes: Figures are for 2000 unless not available (1999 then used as alternative); “net domestic credit” includes claims on central/local governments; “loans” are only loans to enterprises and households, and exclude securities investments; state share of credit based on state bank statements, with private bank shares serving as residual; data for Yugoslavia not used due to liquidation procedures being applied to big banks. See table 5.3 for list of banks included in data analysis. Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); authors’ calculations

60

Table 5.3 State Banks Included in the Analysis at end 2000Albania Savings BankArmenia Armenia Savings BankAzerbaijan IBA, United Universal

BelarusBelpromstroibank, Belagroprombank, Belbusinessbank, Belgazprombank, Belarusbank, Belvnesheconombank

Bosnia Investment Bank, Central Profit Bank, Gospodarska, PrivrednaBulgaria DSK, Biochim, Central Cooperative

CroatiaDubrovacka, Croatia, Croatian Bank for Reconstruction and Development, Hrvatska Postanska

Czech Republic Komercni, Ceskomoravska Zarucni, Ceska Exportni Estonia No state banksFYR Macedonia Macedonian Development Bank (estimated)Georgia No state banksHungary Magyar Fejlesztesi, PostbankKazakhstan Export-Import Bank, HalykKyrgyz Kairat, Energo BankLatvia Latvian Mortgage and Land Bank, Latvian Savings BankLithuania Agricultural Bank Moldova Banca de EconomiiPoland PKO BP, BGZ, National Economy Bank, Bank Ochrony SrodowiskaRomania Banca Agricola, BCR, CEC, EXIMBank

RussiaSberbank, Medium & Long Term Credit Bank, Vnesheconombank, Russian Bank for Development

Slovak RepublicVUB, Investicna a Rozvojova, First Building Savings, Slovenska Zarucna a rojvojova, Banka Slovakia, Exportno-Importna

SloveniaNova Llubljanska, Nova Kreditna Maribor, Postna Banka, Slovene Export Corporation, Slovenska Investicijska

Tajikistan Insufficient data availableTurkmenistan Bank for Foreign Economic AffairsUkraine Ukreximbank, Oschadny

UzbekistanState Housing Savings Bank, Asaka, Uzpromstroybank, National Bank for Foreign Economic Activity

YugoslaviaNote: Banks highlighted in bold have been privatized or liquidated since end 2000

The following table shows the comparison of state banks to private banks in terms of net domestic credit exposure. The figures indicate a significant range of exposures, with 12 of 25 countries for which data are available having state banks with less than 20 percent exposure. However, among the remaining 13, seven had exposures in excess of 50 percent. Even more importantly, many of the largest economies among transition countries had 20-50 percent state bank exposure as a percent of total net domestic credit. As noted above, this includes Russia, Poland and the Czech Republic, all of which accounted for nearly $150 billion in net domestic credit at end 2000, or about 65 percent of total97 for the reporting countries. Thus, on a weighted basis, state bank shares of net domestic credit remained significant in the transition countries at

97 $147.5 billion/$228.7 billion = 64.5 percent.

61

end 2000. Moreover, state banks have the greatest amount of exposure in countries that are often considered to be advanced in the reform process. While some of the largest banks have been privatized since end 2000, such as VUB in the Slovak Republic and Komercni in the Czech Republic, state banks continue to hold significant credit stocks. As noted above, these are often in the form of Government securities, rather than loans.

The table below also shows that the average state bank had $765 million in net domestic credit at end 2000, as compared with only $67 million for the average private bank.98 This compares with 1995 figures that show the average state bank had $480 million in net domestic credit exposure, suggesting that state banks’ average lending and investment in Government securities has increased significantly since the mid-1990s due partly to the major decline (from 200 to about 108) in the number of major state banks in transition countries. Meanwhile, the average private bank had only $31 million in credit exposure in 1995, suggesting that the average private bank remains small, but has grown in lending and investment activity.

On a regional basis, there is some consistency in growth trends. In the CEE countries, the average state bank’s credit exposure has increased since 1995 (from $709 million to $846 million), while the average private bank has shown an increase in credit exposure (from $158 million to $243 million). CIS state banks have shown an increase in credit exposure (from $253 million in 1995 to $707 million in 2000), while the average private bank remains small. In the Baltic states, average credit exposure has increased slightly at state banks among the last few remaining state banks, while private banks have grown.

98 These averages assume 108 state banks and 2,181 private banks at end 2000. In fact, there were more than 108 state banks. These represent the major state banks. The disparity in total state banks would primarily result from the nearly 700 banks in which non-private authorities in Russia have shares in banks or other lending institutions. The number of private banks is estimated from the total recorded in the EBRD Transition Report for 2001 less the 108 major state banks.

62

Figure 5.5 Banks' Net Domestic Credit Exposure at end 2000 (percent)

Notes: Figures are for 2000 unless not available (1999 then used as alternative); net domestic credit includes claims on central/local governments; state share of credit based on state bank statements, with private bank shares serving as residual; in some cases, statistics appear inconsistent, such as in Belarus and Turkmenistan, where negative loans are shown for private banks; no figures available for Tajikistan state banks, or bank figures. Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); EBRD; authors’ calculations

Loan figures show growth in all regions, both in the aggregate and on average. Although not broken out by bank ownership, the following table shows aggregate loans increased from $139 billion in 1995 to $163 billion in 2000. Average loan exposure per bank at the end of 2000 was $232 million in CEE, $100 million in the Baltic states, and only $26 million in CIS.

Table 5.4 Summary of Loan Exposure Trends (millions US dollars)1995 2000

Total Loans Average per Bank Total Loans Average per BankCEE 101,179 188 112,898 232Baltics 1,919 26 4,107 100CIS 35,472 11 46,345 26Total 138,569 37 163,350 71Notes: Loan figures were derived from total loans to state enterprises and the private sector (from IFS). Sources: Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); EBRD; authors’ calculations

Total Assets99

99 Three sources of data are used for bank assets (IMF, EBRD, Fitch IBCA), of which the last two are used for state banks’ figures. These figures are not always identical, and marginal differences appear in some of the tables as a result. However, as with other balance sheet measures, the differences in the figures are not considered material relative to the analysis.

63

Data from 2000 indicate that state banks’ had more than $97 billion in total assets on their balance sheets. These assets account for about 31 percent of total banking system assets in the 25 countries for which data are available. Thus, in terms of comparative indicators, state banks in transition countries are relatively small in general. However, they remain influential in terms of banking development in their individual markets.

Of these assets, about 86 percent were estimated to be in the form of net domestic credit, 100

including securities investments (see figures above). About 61 percent of total state bank assets were in Central European countries,101 mostly Poland and the Czech Republic. These two countries accounted for $36 billion in state bank assets. When added to Russia’s $27 billion in state bank assets, the three countries accounted for 66 percent of total state bank assets.102

Many countries have high proportions of state bank assets to total. Among CEE countries, Albania, Romania and Slovenia had more than half of total assets in state banks at end 2000. Meanwhile, among CIS countries, Azerbaijan and Belarus had more than half their banking system assets in state hands. Tajikistan, Turkmenistan and Uzbekistan also had virtually all bank assets in state hands.

100 $82,537 million/$96,426 million = 85.6 percent.101 $58,600 million/$96,426 million = 60.8 percent.102 $63,253 million/$96,426 million = 65.6 percent.

64

BOX 5.1 BALANCE SHEET CONCENTRATION IN THE INTERNATIONAL BANK OF AZERBAIJAN

Tackling the issue of state-owned banks has proved to be one of the Government of Azerbaijan’s most difficult and complex tasks. Despite early attempts to recapitalize, restructure and privatize state-owned banks since 1996, state ownership in the banking sector remains high, dominated by the International Bank of Azerbaijan (IBA). This bank was left with a near monopoly when the government consolidated its three other troubled state-owned banks into a single entity—United Universal Bank—in 2000. While this step marked significant progress in reducing public ownership, the banking system remains highly concentrated and underdeveloped.

IBA has 75 percent market share of banking sector assets, and 40 percent of retail deposits. In 2000, bank assets stood at $614 million, and its loan portfolio grew by 22 percent. The bank has close links to many government departments and state organizations, including the important Oil Fund, and acts as an intermediary for government-guaranteed credit lines to Azerbaijan.

In 2001, the government reiterated its commitment to privatize IBA and issued a presidential decree to that effect. At present, the Ministry of Finance owns 50.2 percent of the bank’s shares. The EBRD, which has been providing support for the IBA in the form of a credit line targeted to small and medium enterprises, has indicated an interest in taking on a 20 percent equity stake. The remaining state shares are to be auctioned at a later date.

Despite its dominant position, IBA faces many governance and management problems and is plagued by inefficiency. Its profitability was weak in 2000, with after-earnings of only $9 million. The slight increase in profit resulted from the net interest income earned on the placement of funds of the Azeri Oil Fund. These revenues were not expected to recur in 2001. Further, the level of overdue loans rose in 2000 (as did loan loss provisions), and IBA’s loan loss reserve cover was only 12.5 percent of gross loans at end 2000.

While the government’s recent moves related to IBA and United Universal represent progress, privatization is just one element of the financial sector reforms that are necessary. The broader challenge is to make banks more central to economic activity in Azerbaijan in terms of deposit mobilization, lending, and an increased array of services.

As a share of total banking system assets, CIS countries show a higher share of total, at about 46 percent. The CEE countries have about 26 percent of total bank assets in state hands. The Baltic states’ public banks had only 8 percent of total assets at end 2000. The following table highlights these figures.

65

Table 5.5 State Banks' Assets at end 2000 (million US dollars)Total Bank Assets ($) State Bank % of Assets State Bank Assets ($)

Albania 1,993 61.7 1,230Armenia 348 2.6 9Azerbaijan 1,010 64.7 653Belarus 2,207 77.1 1,702Bosnia 2,774 10.2 284Bulgaria 4,622 20.1 931Croatia 13,521 10.9 1,475Czech Republic 49,265 25.9 12,757Estonia 3,162 0.0 0FYR Macedonia 1,279 1.1 14Georgia 322 0.0 0Hungary 24,714 7.8 1,931Kazakhstan 3,302 23.3 769Kyrgyz 96 9.4 9Latvia 4,017 9.2 369Lithuania 3,025 13.8 417Moldova 323 10.8 35Poland 87,744 26.6 23,315Romania 7,607 60.0 4,564Russia 61,573 44.1 27,181Slovak Republic 15,252 32.2 4,911Slovenia 12,847 56.0 7,188Tajikistan N/A N/A 9Turkmenistan 2,075 100.0 2,075Ukraine 5,799 13.6 790Uzbekistan 4,432 100.0 4,432Yugoslavia N/A N/A N/ATOTAL 313,309 31.0 97,050CEE total 221,618 26.4 58,600Baltic total 10,204 7.7 786CIS total 81,487 46.2 37,664Notes: Figures are for 2000 unless not available (1999 then used as alternative); reliable figures for Yugoslavia not available; there are slight discrepancies in these figures with total state bank assets presented below, suggesting some smaller state banks are included in the table below but not in this table. See table 5.3 for list of banks included in data analysis.Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); authors’ calculations

The following table compares total bank assets in state banks with those in private banks. The figures show that the CEE countries have significantly larger private banks in total and on average, even though they also have larger aggregate state bank assets. Even though Central European state bank assets are 1.55 times greater than those in CIS countries, private bank assets for Central Europe are nearly four times private banks’ assets in CIS. Meanwhile, the Baltic states had the highest private share of bank assets at end 2000, at about 92 percent of total.

66

As for specific countries, about one-third had state bank assets in excess of half of total. However, the largest aggregate figures for state bank assets were in Russia, Poland and the Czech Republic. These three countries combined had $63 billion in state bank assets, which was equivalent to 20 percent of total banking system assets in all reporting transition countries. This shows the level of concentration of banking system assets, as these three countries combined accounted for 63 percent of total assets.103

The table below also shows that the average state bank had about $900 million in total assets at end 2000, as compared with only $99 million for the average private bank. This shows an increase from 1995, when state bank assets averaged $664 million. The increase has primarily come from the CIS region, where average state banks had $335 million in assets in 1995, compared with $876 million in 2000. Sberbank of Russia is largely responsible for these changes, along with the general decrease in the number of state banks as a result of privatization, consolidation and failure.104 CEE and Baltic countries’ state banks showed roughly the same level of assets, with the Baltic state banks showing a bit more of a proportional increase.

Meanwhile, private banks are considerably smaller than state banks in both Central Europe and CIS countries, whereas they are larger on average in the Baltic states. Private banks continue to grow in the CEE region, and the statistics would shift dramatically if the Komercni privatization in 2001 were included in the analysis, just as the statistics will shift even more when PKO BP of Poland is eventually privatized. Meanwhile, CIS private banks remain very small, at only $26 million in assets, about 10 percent of the average Baltic private bank.

Figure 5.6 Banks' Assets in 2000 (percent)

103 $198.6 billion/$313.3 billion = 63.4 percent.104 The average asset size of CIS state banks would be considerably smaller if the nearly 700 banks in which the Russian government, central bank, or other public sector agency had a minority share were included in the denominator.

67

Notes: Figures are for 2000; estimates in state/private bank shares made for FYR Macedonia, Turkmenistan and Uzbekistan; figures for Yugoslavia are not included due to expected write-offs of about $6 billion-equivalent as a result of liquidation procedures for the major banks. Sources: IMF (International Financial Statistics); EBRD Transition Report 2001; authors’ calculations

Deposits

Data from 2000 indicate that state banks’ had about $77 billion in deposits105 or about 37 percent of total banking system deposits for reporting countries. About 62 percent of the state banks’ deposit total was in Central European countries.106 Russia and Poland alone accounted for 56 percent of the total, attesting to the importance of Sberbank and PKO BP in transition country deposit mobilization. These two countries combined with the Czech Republic, Romania, Slovenia and the Slovak Republic accounted for 86 percent of total state bank deposits at end 2000.

As a share of total average country deposits, CIS countries show a higher share of total, at about 58 percent, much higher than their credit figures. The Central European countries have about one third of total deposits in state banks, about the same as state bank credit shares. The Baltic states’ public banks were similar to their counterparts in Central Europe in terms of matching deposit and credit shares. However, as a percent of total, the Baltic states had only about 12 percent of total deposits in state banks at end 2000. The following table highlights these figures.

105 Differences in this figure with other figures show up in Lithuania, Slovenia and Yugoslavia. In Lithuania, the higher figures include the now private Savings Bank. In Slovenia, there may be some double-counting in terms of deposits. In Yugoslavia, the lower figure excludes deposits due to the liquidation process under way. 106 $47,538 million/$77,153 million = 61.6 percent.

68

Table 5.6 State Banks' Deposit Figures at end 2000 (million US dollars)Total Bank Deposits($)

State Bank % Deposits State Bank Deposits($)

Albania 1,605 73.3 1,176Armenia 167 5.4 9Azerbaijan 546 88.8 485Belarus 1,156 111.2 1,285Bosnia 834 18.9 158Bulgaria 2,785 28.1 782Croatia 8,085 8.9 720Czech Republic 32,796 30.2 9,915Estonia 1,590 0.0 0FYR Macedonia 531 0.0 0Georgia 156 0.0 0Hungary 17,814 7.4 1,314Kazakhstan 1,981 32.7 648Kyrgyz 68 10.3 7Latvia 1,447 19.4 281Lithuania 1,946 16.5 322Moldova 170 15.3 26Poland 62,837 31.7 19,944Romania 6,145 58.9 3,619Russia 39,903 58.0 23,156Slovak Republic 11,265 34.8 3,922Slovenia 8,277 72.3 5,988Tajikistan N/A N/A 8Turkmenistan 434 100.0 434Ukraine 3,387 17.1 578Uzbekistan 2,384 100.0 2,384Yugoslavia N/A N/A N/ATOTAL 208,309 37.0 77,153CEE total 152,974 31.1 47,538Baltic total 4,983 12.1 603CIS total 50,352 57.6 29,012Notes: Deposits = "customer and short-term funding"; Turkmenistan and Uzbekistan total deposits estimated for 2000. See table 5.3 for list of banks included in data analysis.Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); authors’ calculations

The following figure shows the comparison of state banks to private banks in terms of deposit mobilization figures at end 2000. The figures indicate about one-third of transition countries still had very high levels of state bank deposit concentration, exceeding 50 percent of total. These countries included Russia, with nearly 60 percent of total banking system deposits in state banks, mainly Sberbank.107. Another five countries had between 20-50 percent of total deposits in state

107 Sberbank alone accounted for about three quarters of domestic currency deposits, and about half of hard currency deposits in Russia in late 2001.

69

banks, including Poland, the country with the highest level of bank deposits among all transition countries.108.

As noted above, there is significant country concentration in terms of deposit mobilization. About six countries account for 86 percent of total state bank deposits. Aggregate figures show that Poland, Russia, the Czech Republic and Hungary account for 74 percent of total transition country deposits109 among reporting countries.

The table below also shows that the average state bank had $714 million in deposits at end 2000, as compared with the average private bank that had only $60 million in deposits. These differences are most dramatic in the CIS region, where the average state bank had $675 million in deposits, as compared with the average private bank with only $12 million. This largely reflects the near monopoly savings banks have had in CIS countries, particularly in local currency. It is also important to recognize the significant growth in average deposits for CIS state banks, which had less than $200 million in 1995. Again, as with other indicators, these averages would decline significantly if the full number of banks in which the Russian state authorities had shares were included in the denominator. Meanwhile, state banks in CEE countries remain large on an average deposit basis, with $779 million in deposits, more than three times the average CEE private bank. It is noteworthy that both state and private banks in CEE have experienced deposit growth on average since 1995, with the average private bank nearly doubling deposits. It is also noteworthy that state banks in the CEE region have increased average deposits about 1.4 times the average private bank.110 In the Baltic region, private banks were smaller than in CEE countries, but larger on an average deposit basis than private banks in the CIS countries. Baltic state banks were smaller on average than state banks in both CEE and CIS.

Figure 5.7 Banks' Deposits at end 2000 (percent)

108 Poland accounted for about 30 percent of total transition country deposits as of end 2000.109 $153.4 billion/$208.3 billion = 73.6 percent.110 The average CEE state bank increased deposits $161 million from 1995 to 2000. The average private CEE bank increased deposits $114 million. Thus, $161/$114 = 1.41 times.

70

Notes: Deposits = "customer and short-term funding"; Turkmenistan and Uzbekistan total deposits estimated for 2000; figures for Yugoslavia not used due to liquidation procedures in place for major banks; statistical issues recognized in the case of Belarus, where negative deposits appear for private banks.Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); authors’ calculations

As for loan-to-deposit ratios, end 2000 figures suggest that most ratios are conservative and prudent except with regard to private banks in the CIS countries. Here, loans account for about 1.4 times deposits. Should there be any erosion in loan quality among these banks, this would endanger the safety of deposits. Private banks in the CEE countries and the Baltics show loans at about 85 percent of deposits, which appears to be about as high a ratio that is prudent given that classified loans in these countries are estimated to have averaged about 16.3 percent at end 2000.111. This would have equated with 12 percent of total deposits in the CEE and Baltic banks.112

In general, loan-to-deposit ratios in state banks appear fairly conservative. This means their credit risk is fundamentally a mix of Government and company risk, as the balance of net domestic credit is in Government securities. Meanwhile, private banks need to focus on their exposures to enterprise credit risk, as most of their earning assets are exposed to companies (and households) as opposed to Government. As noted above, private banks in the CIS also need to be on their guard for domestic debt risk, as their overall credit exposure to Government securities is about 22 percent. All together, Central European banks have the lowest loan-to-deposit ratios at 74 percent, with the Baltics and CIS banks having ratios about 10-20 percent higher on average.

Table 5.7 Comparative Distribution of Loans and Deposits: 2000 (million US dollars)State Banks Private Banks All Banks

Loans Deposits Loans Deposits Loans DepositsCEE 22,864 47,538 90,034 105,436 112,898 152,974

111 $19 billion/$117 billion = 16.3 percent. NPLs were higher in CEE countries than in the Baltic states.112 $19 billion/$158 billion = 12.1 percent

71

Baltics 350 603 3,757 4,380 4,107 4,983CIS 16,742 29,012 29,603 21,341 46,345 50,352Total 42,282 77,153 123,395 131,156 163,350 208,309

Loan-to-Deposit Ratios by Region and in Total (in percent)State Banks Private Banks All Banks

CEE 48.1 85.4 73.8Baltics 58.0 85.8 82.4CIS 57.7 138.7 92.0Total 54.8 94.1 78.4Notes: Loan figures not available for Tajikistan; figures for Yugoslavia not used due to liquidation of banks in process; private loan figures derived from total loans to state enterprises and the private sector (from IFS) less loans on state banks’ balance sheets.Sources: IMF; Fitch IBCA; authors’ calculations

Capital

Data from 2000 indicate that state banks had about $10.4 billion in “net capital”113 on their balance sheets, little more than 20 percent of total bank capital in transition countries. This figure is less than state banks’ comparable shares of credit, assets and deposits, and points to the possibility that these banks are undercapitalized relative to their private bank peers. Alternatively, the lower capital figures relative to other balance sheet measures may partly reflect more conservative risk-weighting due to the higher proportion of Government securities, or implicit guarantees by the state that these banks would be rescued should they face serious solvency or liquidity issues. This has already occurred in many transition countries, and continues to pose a macroeconomic risk and potential competitive distortion in markets where their roles are particularly active.

CEE countries’ state banks had about $6.2 billion in capital at end 2000, mainly in Poland and the Czech Republic. Russia was the other major market where state banks had fairly sizeable capital as a percentage of total. Overall, these three countries accounted for $5.9 billion in state bank capital, or 56 percent of total.

On a percentage basis relative to total capital in domestic systems, Belarus, Romania, Turkmenistan and Uzbekistan showed high levels of state bank capital to total. However, on an aggregate basis, this only amounted to about $1.9 billion, with Uzbekistan representing the largest proportion. Other countries with 20-50 percent levels of state bank capital to total were in Central Europe, and included Bosnia-Herzegovina, Croatia, Hungary, the Slovak Republic and Slovenia. Several of these Central European countries have reduced their state shares since 2001.

As a share of total average country state bank capital, the CEE and CIS countries all averaged about 21-22 percent of total at end 2000. The Baltic states’ $53 million in state bank capital accounted for only five percent of total system capital at end 2000. The following table highlights these figures.

113 Net capital is derived from IFS, and subtracts (or adds if positive) “other items net” from “capital accounts” for banking/deposit-taking institutions.

72

Table 5.8 State Banks' Capital at end 2000 (million US dollars)Total Bank Capital

($)State Bank % of

CapitalState Bank Capital($)

Albania 292 11.3 33Armenia 27 0.0 0Azerbaijan 184 13.0 24Belarus 378 85.0 321Bosnia 298 32.2 96Bulgaria 1,027 11.2 115Croatia 1,976 23.0 454Czech Rep. 5,335 21.4 1,144Estonia 408 0.0 0FYR Macedonia 404 1.0 4Georgia 100 0.0 0Hungary 2,404 21.0 504Kazakhstan 794 9.6 76Kyrgyz 11 12.9 1Latvia 305 6.9 21Lithuania 307 10.4 32Moldova 90 4.5 4Poland 12,758 14.6 1,857Romania 594 116.2 690Russia 15,317 18.6 2,849Slovak Rep. 2,699 21.8 588Slovenia 1,693 40.2 681Tajikistan N/A N/A N/ATurkmenistan 21 100.0 21Ukraine 1,225 4.9 60Uzbekistan 851 100.0 851Yugoslavia N/A N/A N/ATOTAL 49,496 21.0 10,408CEE total 29,480 20.9 6,153Baltic total 1,020 5.2 53CIS total 18,997 22.1 4,202Notes: Figures are for 2000 unless not available (1999 then used as alternative); “capital” = “capital accounts” +/- “other items net”; state share of capital is based on state share of assets applied to capital, with private bank shares serving as a residual; reliable data on bank capital for Tajikistan and Yugoslavia not available. See table 5.3 for list of banks included in data analysis.Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); EBRD; authors’ calculations

The following figure shows the comparison of state banks to private banks in terms of capital. The figures show private banks accounted for about 79 percent of total system capital among transition countries at end 2000, with the Baltic states having the highest proportion of private capital. The vast majority of countries showed fairly high levels of private capital, with only a

73

few countries (noted above) having predominantly state capital. All together, private bank capital was $39 billion, about 60 percent of it in Central European banks.114

Most banks in transition countries remain small, with capital only $22 million on average per bank. This is particularly true in the CIS region, where the average bank had $11 million in capital. Baltic banks had about $25 million, and CEE banks had $61 million.

The average state bank had $97 million in capital, as compared with the average private bank, which had only $18 million. By region, state banks in the CEE region were slightly larger on average in terms of capital ($101 million), compared with CIS state banks ($98 million) and Baltic state banks ($13 million). Private banks in CEE countries were considerably larger than their private bank counterparts in the Baltic region and CIS. In particular, private CIS banks showed very little capital at the end of 2000. In fact, total private bank capital in the CIS region at the end of 2000 was only $14.8 billion, equivalent to total private bank capital for Poland and the Czech Republic. Moreover, Russia accounted for $12.5 billion in private bank capital. This means that all other CIS private banks had only $2.3 billion in bank capital, about $5.5 million per private bank115 net of private banks in Russia. Meanwhile, private Russian banks themselves were not particularly large, averaging less than $10 million in capital at end 2000.116 This is much smaller than private banks in the CEE and Baltic markets.

Figure 5.8 Banks' Capital at end 2000 (percent)

114 $23,327 million/$39,088 million = 59.6 percent.115 $2,326 million/422 = $5.5 million. There are 422 private banks in CIS countries apart from Russia.116 $12,468 million/1,309 = $9.5 million. There were a reported 1,309 private banks in Russia in 2000.

74

Notes: Figures are for 2000 unless not available (1999 then used as alternative); “capital” = “capital accounts” +/- “other items net”; state share of credit based on public bank statements, with private bank shares serving as residual; statistical problems recognized for Romania; no reliable capital figures available for Tajikistan or Yugoslavia. Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); EBRD; authors’ calculations

Straight capital-to-asset ratios indicate that private banks had nearly four times the capital ratios of state banks. Private banks’ capital-to-asset ratios were about 12.5 percent for all transition countries, while state banks showed only 3.3 percent. This cannot be automatically associated with risk or asset quality, as banks with high ratios may still be insolvent while banks with relatively modest ratios may be in sounder financial condition. However, the 3.3 percent ratios for state banks are low and would ordinarily require corrective action unless the bank was not assuming any risk. Conversely, if the bank is not assuming any risk, there are then questions about its earnings stream, its ability to increase capital from ordinary banking operations, and its long-term viability. To the extent that many state banks have low capital to assets because of heavy investment in Government securities,117 this then raises the question of whether income from securities is enough for these banks to recapitalize for long-term commercial viability and competitiveness. If so, this then points to the question of fiscal stability, and whether this is the best use of these countries’ fiscal resources. To the extent that fiscal prudence keeps such an earnings stream low, this once again raises the question of the bank’s cash liquidity, solvency, and long-term commercial viability. In any event, state banks show low levels of capital to assets in most transition countries, with the exceptions being Belarus, Romania and Uzbekistan.

Meanwhile, the private banks’ 12.5 percent capital-to-assets ratio compares favorably with general 8-10 percent BIS guidelines, although these would have to be tested for adequacy with regard to levels of risk assumed. About half the countries’ private banks’ capital-to-asset ratios are at double-digit levels, although there are many exceptions. Given the high levels of non-performing loans (equivalent to more than 50 percent of capital) at end 2000 in Azerbaijan, Belarus, Bosnia-Herzegovina, Croatia, the Czech Republic, Kyrgyz Republic, Poland, Slovak Republic, and Ukraine, it is possible that state and/or private banks are undercapitalized in these countries.

Table 5.9 Comparative Bank Capital-to-Asset Ratios at end 2000Total Banks' State Banks’ Private Banks’

Capital ($) Assets ($) CARs (%) Capital (%)Albania 292 1,993 1.7 13.0Armenia 27 348 0.0 7.8Azerbaijan 184 1,010 1.9 16.3Belarus 378 2,207 14.5 2.6Bosnia 298 2,774 3.5 7.3Bulgaria 1,027 4,622 2.2 20.0Croatia 1,976 13,521 3.4 11.3Czech Republic 5,335 49,265 2.3 8.5Estonia 408 3,162 0.0 12.9FYR Macedonia 404 1,279 0.3 31.3Georgia 100 322 0.0 31.1

117 Government securities are usually assigned a zero risk weight for regulatory capital (capital adequacy) purposes. However, several countries (including Russia and Ukraine) have defaulted on their domestic debt. This suggests that zero risk weights should not be automatic, and that capital should be higher as a result.

75

Table 5.9 Comparative Bank Capital-to-Asset Ratios at end 2000Total Banks' State Banks’ Private Banks’

Capital ($) Assets ($) CARs (%) Capital (%)Hungary 2,404 24,714 2.0 7.7Kazakhstan 794 3,302 2.3 21.7Kyrgyz 11 96 1.5 10.4Latvia 305 4,017 0.5 7.1Lithuania 307 3,025 1.1 9.1Moldova 90 323 1.2 26.6Poland 12,758 87,744 2.1 12.4Romania 594 7,607 9.1 -1.3Russia 15,317 61,573 4.6 20.2Slovak Republic 2,699 15,252 3.9 13.8Slovenia 1,693 12,847 5.3 7.9Tajikistan N/A N/A N/A N/ATurkmenistan 21 2,075 1.0 N/AUkraine 1,225 5,799 1.0 20.1Uzbekistan 851 4,432 19.2 N/AYugoslavia N/A N/A N/A N/ATOTAL 49,496 313,309 3.3 12.5CEE total 29,480 221,618 2.8 10.5Baltic total 1,020 10,204 0.5 9.5CIS total 18,997 81,487 5.2 18.2Notes: Figures are for 2000 unless not available (1999 then used as alternative); “capital” = “capital accounts” +/- “other items net”; state share of credit based on state bank statements, with private bank shares serving as residual; statistical problems are recognized for Romania; no reliable figures available for Tajikistan or Yugoslavia.Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); authors’ calculations

Asset Quality

In terms of asset quality, bad loans in transition economy banks were a troubling $26 billion at end 2000,118 equivalent to about 16 percent of total loans. Nearly $19 billion was in CEE banks, and less than $7 billion in CIS banks. The latter may be understated, as some CIS countries reported suspiciously low levels of bad loans. For example, Uzbekistan reported zero percent non-performing loans, and Turkmenistan only 0.5 percent. There is also the larger issue of the degree of bad credit in the economy as a whole, which is often captured outside the banking system in the form of arrears to power and utility companies, fiscal accounts, employee wage accounts, and obligations to other enterprises.119 As noted in arrears figures cited below (see Table 6.3 and Annex 5), these issues are particularly problematic in CIS countries and, to some degree, in many of the Balkan states of southeast Europe. All together, bad loan figures were

118 Bad loan percentages are based on an EBRD survey of central banks, and include sub-standard, doubtful and loss loans as a percentage of total loans. 119 For more on this topic, see Siegelbaum, Sherif, Borish and Clarke, “Structural Adjustment in the Transition: Case Studies from Albania, Azerbaijan, the Kyrgyz Republic and Moldova,” World Bank Discussion Paper 429, 2002.

76

equivalent to only 12.5 percent of total deposits, but more than half of total capital for banks in transition countries.

The high bad loan-to-capital ratio points to the risk of undercapitalization, as noted above. In particular, the problem appears to be most acute in Poland, the Czech Republic and Russia, where an estimated $18.2 billion in bad loans are housed with banks. While specific links cannot be made to specific institutions, 2000 figures for the major state banks reporting in these countries showed loan loss reserves to be in the 5-15 percent range.120 If these loans were to deteriorate further, this would likely trigger the need for additional injections of capital. In fact, several countries are in the range of at least 75 percent bad loan-to-capital ratios for 2000, including Azerbaijan, Bosnia-Herzegovina, the Czech Republic, the Kyrgyz Republic, the Slovak Republic, and Ukraine. In some cases, corrective measures have already been taken or are being planned. Bosnia-Herzegovina is in the process of liquidating and privatizing many of its remaining state banks. The Czech Republic has privatized Komercni. The Kyrgyz Republic has essentially placed Kairat Bank under administration. The Slovak Republic has successfully privatized VUB. However, in Ukraine, there are still reported to be problems associated with several banks, including possible risks associated with the recently intensified lending activities of wholly state-owned Oschadny Bank.121

BOX 5.2 UKRAINE’S OSCHADNY BANK: BECOMING COMPETITIVE, OR RISKING FAILURE AND CRISIS?Oschadny is one of the largest banks in Ukraine due to its asset size and retail network. As of late 2000, Oschadny had about 35,000 employees, and controlled 25-30 percent of household deposits in the banking system. However, the amount of household deposits was only about $330 million in late 2000, small for a country of 50 . Thus, while perceived to be “large” in Ukraine, Oschadny is not really a significant bank in terms of aggregate intermediation, reflecting the comparative insignificance of banking and financial intermediation in the economy.

Similar to other state and quasi-state banks, Oschadny’s financial situation deteriorated in the late 1990s. The bank is characterized by (i) high operating costs associated with excessive branch structures and employment, and other operating inefficiencies; (ii) government-directed lending that has resulted in substantially impaired portfolios, reduced ratios of earning assets, and after-tax losses; (iii) ongoing governance problems associated with central and regional government involvement in decision-taking, and with the wholly non-transparent interventions of key business groups connected to the bank’s management and related authorities; (iv) the inability or unwillingness of the government to honor financial obligations to the bank in the form of payments on guarantees and capital contributions; and (v) lagging performance in the upgrading of management systems and informational technologies to lower costs, and to enforce centralized policies on the bank’s regional offices.

Oschadny experienced after-tax losses of $22 million in 2000, and there are concerns that its losses may have been more severe since, or that earnings may be artificially inflated due to questionable loan classification practices.

120 For example, Komercni (Czech Republic) reported loan loss reserves at end 2000 of 14.03 percent of gross loans, little changed from 14.56 percent at end 1999. Sberbank’s loan loss reserves were 11.96 percent of end 2000 gross loans, down from 18.41 percent in 1999. Meanwhile, in Poland, PKO BP reported loan loss reserves of less than 5 percent, and BGZ had no figures available, but reported a reversal of provisions on the income statement in 2000. This would suggest that the bad loan problem in Poland could be more of an issue for private banks than with state banks. 121 Oschadny Bank is the traditional savings bank. Its lending has increased significantly in recent years. While Ukraine’s economy has shown real growth since 2000, there is a general risk that more aggressive lending to generate increased earnings could backfire and require further capital and/or liquidity support should there be a downturn in the economy. This would be particularly serious as significant funding is derived from a major share of the overall household deposit market.

77

Meanwhile, despite its traditional savings bank orientation and portfolio problems, Oschadny increased lending rather than pursuing a more prudent program of corrective actions. Oschadny was also appointed by the government in mid-2000 to act as the authorized bank to service clearing accounts of electricity utility companies and their branches. Oschadny performed this responsibility until October 2001, and the potential losses from these operations and incremental lending remain to be seen.

Solving the problem of Oschadny depends largely on the willingness and capacity of the Ukrainian authorities to take prompt action in terms of governance and strategy. Although strict market logic argues for the bank’s closure, its crucial place in Ukraine’s social fabric makes this solution unlikely in the foreseeable future. To reverse current losses, the bank’s management has pursued a cost-reduction strategy by closing some non-viable offices and releasing staff. The bank closed 148 branches and 2,933 operational offices (agencies) from January 1, 1998 to December 31, 2000. However, the financial and institutional gap is still so large that it is questionable whether Oschadny can become competitive without a major acceleration of restructuring and cost reduction coupled with major assistance from the government. Preliminary estimates done in early 2001 indicate that Oschadny would need to reduce its costs by about 40 percent to achieve breakeven.

The key condition for stabilizing the situation requires that Oschadny not be used as a quasi-fiscal institution, or as a vehicle for directed lending as it so often has been in the past. Such practices subject the institution and government (as its sole shareholder) to a high level of financial risk. At a minimum, firewalls and safeguards need to be put in place to ensure that Oschadny decision-making is grounded in commercial principles. "Social" or "governmental" activity should be off-balance sheet and subject to commercial pricing. Any lending or investment should be explicitly guaranteed (in documented form) so that Oschadny is utilized strictly as an agent that assumes no risk.

Due to the depth of financial distress, there is a risk that Oschadny will seek to grow out of its problems, attempting to leapfrog from its current status as a specialized savings bank to a full-service "universal" bank. This is premature due to the weak financial condition and limited institutional capacity of the bank. Under such circumstances, there is a risk that Oschadny would assume excess risk to generate high earnings and reverse the losses that have accumulated for years. There is a serious risk of adverse selection under such circumstances, especially since the bank does not accurately measure risk and return.

Bad loan figures suggest that the problem of bad loans is not just an issue for state banks, but applies as well to private banks. This raises several issues with regard to banking sector modernization, including the drag on earnings that bad loans present for banks, and the degree to which such weakness in earnings undercuts the ability of banks to modernize (see below). This is not just a basic systems issue, but one in which modern banks require major resources for investment in product diversification, information technologies and management, market research, and internal requirements (e.g., improved standards of governance, management and general staff (re-) training, etc.). As such, considerable capital investment is still needed in CEE (where bad loans accounted for 64 percent of end 2000 capital) to assist with modernization, although this effort is already well under way in many countries, as it is in the Baltic states.

78

The lower bad loan to capital figure in CIS countries may very well be a mistake. First, it likely represents an understatement of bad loans,122 as noted above with regard to figures reported by Turkmenistan and Uzbekistan. Beyond that, it understates the problems of barter, netting and arrears that have functioned almost in parallel to the formal banking and fiscal systems for nearly a decade. As such, while bad loans may only be about 14 percent of deposits and 37 percent of capital in CIS countries, deposits and capital are lower in these countries partly because of the problems households, enterprises, and formal institutions face. This makes it difficult for well functioning systems to emerge, let alone to operate and go through the “growing pains” needed for eventual competitiveness and modernization. Such a bifurcation in development patterns between the CIS on the one hand, and much of Central Europe and the Baltics on the other represents one of the major structural challenges among transition countries today. For the foreseeable future, until these issues are addressed and confidence is gradually restored, the CIS will continue to lag other regions, notwithstanding better bad loan to deposit or capital ratios. The following table highlights key asset quality figures.

122 This problem is not restricted to CIS. While the bad loan figure for Yugoslavia increased in 2000 from earlier years (implying more overt recognition of bad loan problems), the 27.8 percent figure is understated relative to the billions of dollars in loans that may eventually be written off with the liquidation of the major banks. Other countries appear to have faced up to their bad loan problems, at least in terms of making a stronger effort at realization of the magnitude of these potentially bad loans (for monetary and financial sector stability). However, many countries still have weak loan classification practices. Other weaknesses, such as the absence of consolidated accounting or limited capacity to assess political/market risk, may also contribute (unwittingly) to an understatement of bad loans. Many supervisory authorities are aware of these weaknesses, and are in the process of addressing them. Likewise, banks themselves are often in the process of strengthening their internal reporting processes to capture risks. However, continued weaknesses add to the risk that portfolio quality may be overstated.

79

Table 5.10 Banks' Comparative Bad Loans and Impact on Banks’ Financial Condition at end 2000(million US dollars)

Total State Private Bad Loans Bad Loans as a % of:Loans Loans Loans % $ Deposits Capital

Albania 182 10 172 42.6 77 4.8 26.5Armenia 0 0 0 6.2 0 0.0 0.0Azerbaijan 446 184 262 37.2 166 30.4 90.3Belarus 1,254 1,075 179 15.2 191 16.5 50.5Bosnia 2,052 89 1,963 15.7 322 38.6 108.1Bulgaria 1,970 343 1,627 10.9 215 7.7 20.9Croatia 7,258 647 6,611 19.7 1,430 17.7 72.4Czech Republic 26,483 3,752 22,731 19.3 5,111 15.6 95.8Estonia 1,332 0 1,332 1.5 20 1.3 4.9FYR Macedonia 643 8 635 26.9 173 32.6 42.8Georgia 226 0 226 5.6 13 8.1 12.7Hungary 4,469 599 3,870 3.1 139 0.8 5.8Kazakhstan 2,338 385 1,953 5.8 136 6.8 17.1Kyrgyz 56 2 54 16.4 9 13.5 83.5Latvia 1,399 149 1,250 6.3 88 6.1 28.9Lithuania 1,377 201 1,176 10.8 149 7.6 48.5Moldova 182 13 169 20.6 38 22.1 41.9Poland 50,328 10,179 40,149 15.9 8,002 12.7 62.7Romania 2,641 1,015 1,626 3.8 100 1.6 16.9Russia 33,420 10,234 23,186 15.3 5,113 12.8 33.4Slovak Republic 10,105 2,723 7,382 26.2 2,647 23.5 98.1Slovenia 6,767 3,499 3,268 8.5 575 6.9 34.0Tajikistan N/A N/A N/A 10.8 N/A N/A N/ATurkmenistan 1,884 1,803 81 0.5 9 2.2 44.8Ukraine 3,815 322 3,493 32.5 1,240 36.6 101.2Uzbekistan 2,525 2,525 0 0.0 0 0.0 0.0Yugoslavia N/A N/A N/A 27.8 N/A N/A N/ATOTAL 163,152 39,757 123,395 15.9 25,963 12.5 52.5CEE total 112,898 22,864 90,034 16.6 18,792 12.3 63.7Baltic total 4,107 350 3,757 6.3 257 5.2 25.2CIS total 46,147 16,543 29,603 15.0 6,914 13.7 36.4Notes: Bad loan figures are for 2000 unless not available (1999 figures used or Azerbaijan, Kazakhstan and Turkmenistan; 1998 figures are used for Latvia). Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); EBRD; authors’ calculations

Earnings Performance and Solvency Issues

Earnings performance among state banks has varied. After-tax earnings were generally poor in CEE and the Baltic markets among state banks in 2000, while CIS countries reported more favorable earnings influenced mainly by the results of Sberbank of Russia. The latter was partly supported by a rebound in commodity prices and improvements in several CIS economies (e.g.,

80

Russia, Ukraine, Kazakhstan). However, there are also questions about the veracity of the CIS earnings data due to loan classification standards, accounting and audit practices, and the general investor view that risk-related information is insufficiently disclosed. Thus, it is hard to ascertain the degree of real earnings among CIS state banks, as well as the return figures (i.e., assets, equity) against which they are measured.

Taking these caveats into account, CIS state banks reportedly generated $710 million in after-tax earnings, most of it in Russia. These figures represented a 2.1 percent return on assets, and 19.3 percent return on equity. By contrast, CEE state banks only generated $118 million in after-tax earnings (net of Yugoslavia), about $2 million per state bank. ROA was only 0.2 percent, and ROE was 2.0 percent. If Yugoslavia is included in these measures, they all turn negative,123 as shown in the table below. The Baltic state banks showed better return measures in terms of assets and equity, but only had $5 million in after-tax earnings.

The table below shows about $152 million in after-tax earnings in 2000. This is based on reports from about 65 banks, and is about $170 million less than other estimates from varied sources. Irrespective of the figures, there was a high level of concentration of major profits, with only three banks generating more than $100 million in after-tax earnings. Several countries’ state banks showed losses in 2000. This included state banks in Croatia, Hungary, Kazakhstan, the Kyrgyz Republic and Romania. Since then, some of the loss-makers have been privatized (e.g., Bank Agricola in Romania, Dubrovacka in Croatia) or put under administration (e.g., Kairat in the Kyrgyz Republic). Several other state banks that showed low earnings have also been privatized (e.g., Komercni in the Czech Republic).

123 Beobanka Belgrade reported an estimated $500 million in after-tax losses in 2000, and Invest Banka reported $181 million in losses. These losses alone would turn the CEE figures into net losses for the region.

81

Table 5.11 State Banks' After-tax Earnings and Return Measures at end 2000 (million US dollars)After-tax

Earnings ($) RoA (%) RoE (%) State Banks Included in the FiguresAlbania 26 2.1 -173.3 Savings Bank

Armenia 0 0.0 0.0 Armenia Savings Bank

Azerbaijan 10 2.1 55.6 IBA, United Universal

Belarus 10 0.6 2.8

Belpromstroibank, Belagroprombank, Belbusinessbank, Belgazprombank, Belarusbank, Belvnesheconombank

Bosnia 2 0.6 1.5Investment Bank, Central Profit Bank, Gospodarska, Privredna

Bulgaria 13 1.5 12.4 DSK, Biochim, Central Cooperative

Croatia -33 -2.2 -6.9

Dubrovacka, Croatia, Croatian Bank for Reconstruction and Development, Hrvatska Postanska

Czech Republic 16 0.1 1.4Komercni, Ceskomoravska Zarucni, Ceska Exportni

Estonia 0 0.0 0.0 No state banks

FYR Macedonia 0 0.0 0.0 Macedonian Development Bank (estimated)

Georgia 0 0.0 0.0 No state banks

Hungary -17 -0.8 -3.9 Magyar Fejlesztesi, Postbank

Kazakhstan -5 -0.8 -6.7 Export-Import Bank, Halyk

Kyrgyz -1 -11.1 -82.5 Kairat, Energo Bank

Latvia 3 0.9 14.6Latvian Mortgage and Land Bank, Latvian Savings Bank

Lithuania 2 0.5 6.5 Agricultural Bank

Moldova 3 10.7 136.4 Banca de Economii

Poland 19 0.1 1.0 PKO BP, BGZ, National Economy Bank, Bank Ochrony Srodowiska

Romania -90 -1.8 -17.2 Banca Agricola (1999), BCR, CEC, EXIMBank

Russia 577 2.4 24.2

Sberbank, Medium & Long Term Credit Bank, Vnesheconombank, Russian Bank for Development

Slovak Republic 101 2.0 17.5

VUB, Investicna a Rozvojova, First Building Savings, Slovenska Zarucna a rojvojova, Banka Slovakia, Exportno-Importna

Slovenia 81 1.1 12.0 Nova Llubljanska, Nova Kreditna Maribor, Postna Banka, Slovene Export Corporation, Slovenska

82

Table 5.11 State Banks' After-tax Earnings and Return Measures at end 2000 (million US dollars)After-tax

Earnings ($) RoA (%) RoE (%) State Banks Included in the FiguresInvesticijska

Tajikistan N/A N/A N/A Insufficient data available

Turkmenistan 3 0.2 15.4 Bank for Foreign Economic Affairs

Ukraine 9 1.3 17.0 Ukreximbank, Oschadny

Uzbekistan 104 2.3 13.5

State Housing Savings Bank, Asaka, Uzpromstroybank, National Bank for Foreign Economic Activity

Yugoslavia -681 N/A N/A

TOTAL 152 0.2 1.6

CEE total -563 -1.0 -9.7

Baltic total 5 0.7 9.7

CIS total 710 2.1 19.3

Notes: Figures are for 2000 unless not available (1999 then used as alternative); banks highlighted in bold have been privatized or liquidated since end 2000; reliable data for Tajikistan not available; several banks’ profits derived from taking RoA and applying to average assets for 1999-2000. Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); authors’ calculations

The $152 million in after-tax income is particularly modest given the number of major state banks (108), and the size of some of the state banks in 2000.124 On average, such after-tax earnings translate into only $1.4 million in earnings per state bank, not enough for meaningful investment to modernize and become competitive by global standards. Netting out the $843 million in after-tax earnings for the four banks with greater than $100 million in 2000 earnings and offsetting the $681 million in losses from Yugoslavia, the other state banks were basically flat in their net earnings (at a $10 million loss across 100 banks). In general, as indicated by the aggregate statistics, return on (average) assets was only 0.2 percent for state banks, well below the 2-3 percent norms established in OECD and EU markets.

Net capital for state banks increased from $8.7 billion to $10.4 billion in total capital, well above the reported $152 million in after-tax earnings. Netting out Yugoslavia, this suggests that state banks increased capital from retained earnings as well as other sources. However, this increase in capital is still in stark contrast to the private banks in transition countries that showed a net increase in capital of $5.8 billion, about 3.4 times state bank incremental capital. The latter trend is likely a combination of retained earnings and direct investment, the latter of which is a more difficult prospect for state banks given the precarious position of many countries’ monetary and fiscal positions where state banks play a prominent role in the economy. However, the capital increase for private banks also should not be overstated. Given the 2,181 private banks operating in transition countries in 2000, this only amounted to a net capital increase per private bank of

124 Komercni, PKO BP and Sberbank combined for only $568 million in after-tax earnings on $45 billion in average assets in 2000. This is only a 1.3 percent ROA, suggesting high costs, weak operations, inefficient use of assets, and insufficient earnings from other sources/activities.

83

$2.6 million, far less than the $15.7 million increase for average state banks.125 Netting out Sberbank’s approximately $315 million capital increase from 1999 to 2000,126 state banks still showed increases of about $12.9 million on average in 2000.127

While the after-tax earnings of state banks were positive, this does not account for various tax breaks, forbearance, and other forms of “corrective action” needed to improve their solvency and liquidity positions. This suggests there is either little or no investment capital interested in these banks short of privatization, and the cost of keeping them afloat as “going concerns” requires continued tax breaks, forbearance and related benefits to be commercially viable to the extent that results barely above breakeven are viewed as commercially viable. There is greater risk due to many of these banks having higher personnel loads, and the relative lack of investment in new technologies. This perpetuates a manual cycle of inefficiency that makes it difficult for these banks to compete without preferential treatment or protection.

One of the main problems these banks face in reducing costs and boosting earnings is their heavy head count, which is symptomatic of their traditional manual processing and symbolic of labor protection. Rather than operating as commercial banks, they often continue to operate as “social” organizations in which stakeholder rights (e.g., of employees) are equal or superior to shareholder rights. This has weakened earnings, and perhaps more importantly, undermined the ability of state banks to invest in the systems and technologies needed to modernize. Of course, competitiveness and efficiency require more than new technologies and systems. Also required are better incentives for performance, professional management, autonomous and critical internal audit, and qualified boards for sound oversight. (Annexes 1-2 include operational measures for state banks, and head count where available.)

In general, the issue of undercapitalization appears to have been recognized in many countries as the basis for reversing financial weakness and helping to build sound financial systems. The small average capital of most banks has been noted above. Absent sufficient capital, it is questionable if many of the existing banks will have the financial wherewithal to invest as needed to upgrade systems, train personnel, and be competitive in a dynamic marketplace. As noted in the 2000 earnings, most state banks generated losses or were barely above breakeven, raising issues of long-term viability. However, private banks appear to recognize the need for greater capital, and this was evident in trends in 2000. Overall, transition countries’ bank capital increased 45.6 percent in the aggregate relative to asset increases in 2000.128 This is significant, and was particularly apparent among Russia’s private banks.

125 Private: $5,768 million/2,181 = $2.6 million. State: $1,693 million/108 = $15.7 million.126 2000: Equity-to-Assets ratio was 7.55 percent on $20 billion in total assets = $1,510 million in equity. 1999: Equity-to-Assets ratio was 7.47 percent on $16 billion in total assets = $1,195 million in equity. Therefore, Sberbank’s equity increased about $315 million.127 $1,378 million/107 = $12.9 million. Meanwhile, PKO BP ($139 million) and Komercni ($43 million) accounted for $182 million of the remaining $1,378 million in state banks’ capital increases. Thus, net of Sberbank, PKO BP and Komercni, the other state banks averaged about $11.4 million in net capital increases in 2000. 128 $7,461 million in incremental capital/$16,367 in incremental assets = 45.6 percent.

84

Table 5.12 Bank Capital Increases: 1999-2000 (million US dollars)Private Banks State Banks TOTAL

1999 Capital

2000 Capital

Net Increase

1999 Capital

2000 Capital

Net Increase

Net Increase

Albania 326 259 -67 -63 33 96 29Armenia 36 27 -10 0 0 0 -10Azerbaijan 15 0 -15 12 19 7 -8Belarus 89 57 -33 404 321 -83 -116Bosnia 152 202 50 102 96 -6 44Bulgaria 916 925 9 94 102 8 17Croatia 1,451 1,522 71 504 454 -50 21Czech Republic 5,957 4,191 -1,766 1,130 1,144 14 -1,752Estonia 374 408 34 0 0 0 34FYR Macedonia 392 400 8 4 4 0 8Georgia 95 100 5 0 0 0 5Hungary 2,042 1,980 -62 362 504 142 80Kazakhstan 524 718 194 74 76 2 196Kyrgyz 9 10 1 1 1 0 2Latvia 169 284 115 20 21 1 116Lithuania 252 275 23 30 32 2 25Moldova 71 86 15 0 4 4 19Poland 8,332 10,901 2,569 1,771 1,857 86 2,655Romania 175 -96 -271 356 690 334 63Russia 8,265 12,468 4,203 1,914 2,849 935 5,138Slovak Republic 1,679 2,111 432 573 588 15 447Slovenia 1,045 1,012 -33 668 681 13 -20Tajikistan N/A N/A N/A N/A N/A N/A N/ATurkmenistan N/A N/A N/A 18 21 3 3Ukraine 870 1,165 295 46 60 14 309Uzbekistan N/A N/A N/A 695 851 156 156Yugoslavia N/A N/A N/A N/A N/A N/A N/ATOTAL 33,235 39,003 5,768 8,715 10,408 1,693 7,461CEE total 22,467 23,407 940 5,501 6,153 652 1,592Baltic total 795 967 172 50 53 3 175CIS total 9,973 14,629 4,656 3,164 4,202 1,038 5,694Notes: Figures are for 2000 unless not available (1999 then used as alternative); reliable data for banks in Tajikistan and Yugoslavia not available, nor is reliable information available for private banks in Turkmenistan or Uzbekistan. Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); authors’ calculations

The table below highlights bank asset increases in 2000 (vs. 1999), which also shows that banks in Poland and Russia were responsible for most of the asset growth in transition economies in 2000. The Czech Republic and Hungary experienced major declines in total bank assets. Noteworthy with regard to 1999-2000 trends is that Russia’s private banks declined in assets, whereas state banks grew. In Poland, private banks were the primary drivers of growth, although state banks also showed increases in total assets. In general, state banks grew more than private banks in terms of assets, the inverse of what happened in terms of capital.

Table 5.13 Bank Asset Increases: 1999-2000 (million US dollars)

85

Private Banks State Banks TOTAL1999

Assets2000

AssetsNet

Increase1999

Assets2000

AssetsNet

IncreaseNet

IncreaseAlbania 684 763 79 1,197 1,230 33 112Armenia 279 339 60 9 9 0 60Azerbaijan 423 357 -66 307 653 346 280Belarus 369 505 136 1,836 1,702 -134 2Bosnia 2,775 2,490 -285 259 284 25 -260Bulgaria 3,712 3,691 -21 795 931 136 115Croatia 10,679 12,046 1,367 1,564 1,475 -89 1,278Czech Republic 40,400 36,508 -3,892 11,975 12,757 782 -3,110Estonia 2,725 3,162 437 0 0 0 437FYR Macedonia 1,202 1,265 63 14 14 0 63Georgia 255 322 67 0 0 0 67Hungary 24,607 22,783 -1,824 2,076 1,931 -145 -1,969Kazakhstan 1,712 2,533 821 477 769 292 1,113Kyrgyz 84 87 3 9 9 0 3Latvia 2,797 3,648 851 265 369 104 955Lithuania 2,322 2,608 286 384 417 33 319Moldova 231 288 57 21 35 14 71Poland 55,494 64,429 8,935 20,751 23,315 2,564 11,499Romania 2,656 3,043 387 5,296 4,564 -732 -345Russia 36,039 34,392 -1,647 21,367 27,181 5,814 4,167Slovak Republic 10,144 10,341 197 5,458 4,911 -547 -350Slovenia 5,685 5,659 -26 7,022 7,188 166 140Tajikistan N/A N/A N/A N/A N/A N/A N/ATurkmenistan N/A N/A N/A 1,827 2,075 248 248Ukraine 3,470 5,009 1,539 586 790 204 1,743Uzbekistan N/A N/A N/A 4,703 4,432 -271 -271Yugoslavia N/A N/A N/A N/A N/A N/A N/A

TOTAL208,74

4216,26

8 7,524 88,198 97,041 8,843 16,367

CEE total158,03

8163,01

8 4,980 56,407 58,600 2,193 7,173Baltic total 7,844 9,418 1,574 649 786 137 1,711CIS total 42,862 43,832 970 31,142 37,655 6,513 7,483Notes: Figures are for 2000 unless not available (1999 then used as alternative); reliable data for banks in Tajikistan and Yugoslavia not available, nor is reliable information available for private banks in Turkmenistan or Uzbekistan. Sources: IMF (International Financial Statistics); Bank Scope (Fitch IBCA); authors’ calculations

86

CHAPTER SIX: APPROACHES: WHAT HAS AND HAS NOT BEEN DONE?

THE COSTS OF MAINTAINING THE STATE BANK SYSTEM

State banks have long served as vehicles for government spending, directed lending, tax collection and processing, and other quasi-fiscal functions. Their most prominent roles have traditionally been to lend to the state enterprise and farm sector, to serve as savings institutions (sometimes with a state guarantee for deposits), to serve as a lender and agent for export-import financing on behalf of state ministries and trading companies, and more recently, to operate in the non-bank sector by managing securities brokerages, investment funds, and private pension funds. State banks have sometimes held sizeable shares of state insurance companies as well.

In terms of their main activities, the state banks have served as overdraft providers for troubled enterprises and farms. While this role has steadily diminished over the years, the marginalization of this role has evolved in different ways. In most CEE and Baltic countries, lending decreased once new prudential norms were introduced and enforced, and as banks recognized the need to recapitalize.129 Continued lending to troubled enterprises only jeopardized their ability to comply, while investment in Government securities was a far easier way to generate the profits needed to rebuild their balance sheets.130 This has coincided in these countries with an improvement in the legal and institutional environment for creditors, resolution of problem assets, and enhanced financial discipline on the part of private borrowers. The result has been an improvement in the returns and capital positions of banks, a general increase in competition, a wide array of financial products, and improved service levels.

In contrast, CIS countries experienced erosion in lending to all sectors. This has not been accompanied by an improvement in the environment (for creditors or debtors), nor has confidence fully returned after hyperinflation and numerous banking crises. As such, the deposit and capital base of the banking system in the CIS countries is weaker, financial intermediation is much lower, financial products are limited, and service levels are often considered poor.

BOX 6.1 TWO DIFFICULT LIQUIDATION DECISIONS: ROMANIA’S BANCOREX AND UKRAINE’S BANK UKRAINA

129 The latter occurred in different ways, sometimes directly from the budget, and in other cases from other approaches involving monetary mechanisms, forbearance, and related tools.130 In the mid-1990s, net spreads on Government securities investments to deposits were often in the 10 percent range. As these were perceived to be risk-free, they involved little credit risk analysis. Further, it was far easier for banks to comply with capital adequacy and regulatory liquidity requirements. By contrast, enterprise loans in a risky environment required greater effort and made it more difficult to comply with prudential norms. Only when the environment stabilized, competition increased, net spreads on Government securities declined and incentives for lending improved did banks begin to restore lending to the real sector.

87

Bancorex: Bancorex (BX), the former foreign trade bank, was the largest and most troubled of the four fully state-owned banks in Romania prior to its closure in 1999. Accounting for about one-fourth of total banking sector assets and about half of foreign currency loans in the early to mid-1990s, BX financed a significant portion of Romania's energy import requirements, as well as imports of capital goods. The authorities also used BX as a major vehicle to subsidize the state-owned energy sector and energy-intensive industrial sector, also primarily controlled by the state. At the end of 1997, BX received an equivalent of $600 million in government bonds (2 percent of GDP) to restructure nonperforming loans in the portfolio. However, the restructuring of BX, which was to accompany the recapitalization, never took . Although a new management team was appointed in April 1998 and other steps were taken, a comprehensive restructuring plan was never implemented and the bank's situation deteriorated further. When BX was again in crisis in late 1998, the authorities considered restructuring measures with a view to privatizing the bank, although international experience would have favored liquidation. The authorities were concerned about the systemic risk potentially associated with liquidation, and contemplated an up-front recapitalization, followed by restructuring and privatization. However, as the depth of the bank's problems became more fully known, it became clear that BX was in much worse shape than expected, and that privatization with recapitalization would be prohibitively costly. A recapitalization would have required up to $2 billion from the budget, or almost 6 percent of GDP. Finally, in April 1999, BX collapsed as depositors lined up to withdraw their money. It became clear that a rapid liquidation was the only solution that would avoid further runs on the bank and a systemic crisis amid a fragile external economic period. Realizing the magnitude of BX's problem, in April 1999, the authorities finalized a liquidation plan aimed at the orderly removal of BX from the banking system.

Bank Ukraina: Bank Ukraina (BU) was the traditional specialized agricultural bank that became a “privatized” institution, but without the restructuring and changes in governance to become viable. In the case of BU, the government exerted direct influence on the decision-making process by keeping a residual state stake of at least 13 percent until 1998. Specifically, the government manage the institution through the Ministry of Agriculture, Ministry of Finance, and the National Bank of Ukraine (NBU). Regional authorities appeared to exercise some control over regional offices. This complicated governance structure effectively turned the bank into a series of regional banks operating under the same name. The practice of government-directed loans to mostly non-viable state-owned enterprises that were approved under some form of “instruction” by local governments, ministerial resolutions, or presidential decrees proved to be particularly harmful to BU’s financial health. This was because most of the directed loans were never intended to be repaid. The situation of the bank deteriorated dramatically from 1998. share of bad loans became unsustainable even by Ukrainian standards, and BU had to rely on central bank credits be maintain its liquidity position. IMF-led diagnostic review of the bank in the same year confirmed its deep insolvency. The state-protected bank came to be seen as having wasted the country's financial resources by subsidizing loss-making industries and the largely unreformed agricultural sector. In November 1998, the authorities finally put it into a rehabilitation program, and in June 1999 the NBU instructed BU to sign a Commitment Letter aimed at bank recovery. However, the bank failed to the financial targets of the program. Given the high level of deterioration in its loan portfolio (approximately 75 percent of which was non-performing by the end of 2000, although estimated as high as 90 percent if reclassified under IAS), negative liquidity, and capital erosion, the bank's financial condition became dangerous enough to be a potential threat to the entire system. This resulted in the introduction of Provisional Administration in the bank as of September 25, 2000.

Continued state ownership in the banking system of transition countries has often shown itself to be harmful to the economy, and particularly problematic in the year or so running up to elections. Moreover, the continued presence of these banks has often deterred prime-rated foreign investment from the market, or deterred these and other banks from taking on more risk due to potential distortions resulting from such patronage or preferential treatment. Consequently, in countries where state banks continue to play a prominent role, lending has tended to be less available and more costly to the enterprise sector. This, in turn, has undercut financial intermediation as a whole, as enterprises have less incentive to place funds with banks since they

88

find it difficult or uneconomical to borrow from banks. The following table provides a summary of the net spreads on loans (to deposits), levels of bad loans, and the relative degree of state ownership in the banking system from 1996-2000. The figures indicate that in virtually every country where state ownership in the banking system has been high, non-performing loans have been a problem, and net spreads between loans and deposits have been costly for enterprises.

The figures also show that portfolio quality and intermediation costs are not just a function of ownership, and that they are highly interdependent with the macroeconomic framework and standards of governance. Thus, there are several countries (e.g., Albania, Croatia, FYR Macedonia, Georgia, Kyrgyz Republic, Moldova, Romania, Ukraine) in which state ownership of bank assets has diminished in recent years, yet non-performing loans and net spreads have stayed high or increased. Part of this has to do with banks’ need to increase earnings (e.g., from net spreads) to recapitalize, particularly when tougher prudential norms are in place (as manifested in clearer recognition of non-performing loans) and the government eliminates or reduces refinancing options for poorly performing banks. The table below highlights some of the trends in state ownership of the banks, and non-performing loans by country and region from 1996-2000.

89

Table 6.1 State Ownership, Non-performing Loans and Net Spreads: 1996-2000 (in percent)1996 1998 2000

State% NPLs Spreads State% NPLs Spreads State% NPLs SpreadsAlbania 93.7 40.1 7.2 85.6 35.4 8.5 64.8 42.6 13.8Armenia 3.2 22.6 34.2 3.7 10.4 23.6 2.6 6.2 13.5Azerbaijan 77.6 20.2 20.0 65.5 19.6 16.8 60.4 N/A 15.0Belarus 54.1 14.2 29.9 59.4 16.5 12.7 66.0 15.2 30.1Bosnia N/A N/A N/A N/A N/A 21.6 55.4 15.7 15.8Bulgaria 82.2 15.2 48.8 56.4 11.8 10.3 19.8 10.9 8.4Croatia 36.2 11.2 16.9 37.5 12.6 11.1 5.7 19.7 8.3Czech Rep. 16.6 21.8 5.8 18.6 20.3 4.7 28.2 19.3 3.7Estonia 6.6 2.0 7.6 7.8 4.0 8.6 0.0 1.5 3.9FYR Macedonia 0.0 21.7 8.0 1.4 7.8 9.4 1.1 26.9 7.7Georgia 0.0 6.3 27.3 0.0 6.5 30.0 0.0 5.6 28.6Hungary 16.3 9.0 5.1 11.8 6.8 3.1 8.6 3.1 3.0Kazakhstan 28.4 19.9 24.1 23.0 4.7 3.9 1.9 2.1 4.3Kyrgyz 5.0 26.1 28.3 7.1 0.2 37.6 7.1 16.4 33.5Latvia 6.9 20.0 14.1 8.5 6.8 9.0 2.9 5.0 7.5Lithuania 54.0 32.2 7.6 44.4 12.5 6.2 38.9 10.8 8.3Moldova 0.3 46.0 11.3 0.3 32.0 10.6 9.8 20.6 10.5Poland 69.8 14.7 6.1 48.0 11.8 6.3 24.0 15.9 5.8Romania 80.9 48.0 14.7 75.3 58.5 16.6 50.0 3.8 20.3Russia 37.0 13.4 91.7 41.9 30.9 24.7 41.9 15.3 17.9Slovak Rep. 54.2 31.8 4.6 50.0 44.3 4.9 49.1 26.2 6.4Slovenia 40.7 10.1 7.5 41.3 9.5 5.5 42.2 8.5 5.7Tajikistan 5.3 2.9 13.0 29.2 3.2 34.0 6.8 10.8 -8.4Turkmenistan 64.1 11.4 70.0 77.8 2.2 34.4 N/A N/A N/AUkraine N/A N/A 46.3 13.7 34.6 32.3 11.9 32.5 27.8Uzbekistan 75.5 0.0 22.0 67.3 0.1 21.0 77.5 0.0 N/AYugoslavia 92.0 12.3 162.4 90.0 13.1 44.1 90.9 27.8 43.3CEE Avg. 53.0 23.6 23.9 43.0 22.2 12.2 36.7 22.6 11.9Baltic Avg. 22.5 18.1 9.8 20.2 7.8 7.9 13.9 5.8 6.6CIS Avg. 31.9 16.6 34.8 32.4 13.4 23.5 26.0 11.3 17.3Notes: For regional averages, countries are excluded from the denominator if no data are available. Sources: IFS (IMF); EBRD; authors’ calculations.

While there is clearly a link between bad policy and poor credit management, another look at the data shows that the link between NPLs and collapse is not fully predictable in terms of timing and impact. Given the prominence of state banks in much of the directed lending to loss-making state enterprises, there is a correlation between state control of banks, the link between peak levels of NPLs and state ownership in the banking system, and eventual collapse in the economy. In this regard, problems of loan quality, ownership and, by extension, governance and management, are all linked. However, the data below show that in only three countries,131 NPLs peaked shortly (within two years) in advance of the country’s poorest output performance. Bulgaria probably represents the best case of causality, with peak NPLs in 1996, the year in which the economy collapsed. Figures in 1997 reached their nadir as a result. In eight countries,

131 Bulgaria, Kazakhstan and the Kyrgyz Republic.

90

peak NPLs appeared to be an after-effect (within two years) of the output collapse.132 This occurred at different times, with most of these countries recognizing the problem early in the 1990s or by the mid-1990s. For example, in Armenia, a high-level strategy committee on bank restructuring was established in late 1996 (when NPLs were nearly 23 percent of total) to develop a restructuring strategy for the remaining three state banks based on individual rehabilitation agreements signed with each of the three banks.133 By 1998, NPLs had dropped to 10 percent of total, and net spreads had declined substantially. In three countries,134 these figures converged in the same year, suggesting there was fairly simultaneous recognition of the link. Estonia and Hungary recognized these problems relatively early on, and both took resolute action on their portfolios and banking systems thereafter.135

In about half the transition countries, the relationship was less connected. For example, in Albania, output reached its lowest level in 1992, yet 1997 was the worst year for recorded non-performing loans. This is more related to the collapse of the pyramid schemes than anything else, but reflects poor institutional capacity and distorted incentives that made these schemes possible. In some countries, the ruble collapse proved decisive in terms of peak NPLs (Belarus) and output (Russia). Azerbaijan’s peak NPLs in 1999 were likely an accumulation and a reflection of delayed recognition. However, this might have also have been affected by the ruble collapse at the traditional agricultural, industrial and savings banks that are now reconstituted. The peak NPL problem in Latvia represents problems that culminated in the 1995 collapse of Bank Baltija. Moldova’s output collapse in 1999 is more linked to the ruble collapse. That this occurred three years after peak NPLs and most privatization suggests the banks (and the National Bank) were addressing loan quality problems years before general output collapse. In Romania, peak NPLs in 1997 reflect delayed recognition and slow reform. It should be noted that recognition of NPLs often tightened up after macroeconomic and structural collapse as a necessity, and that peak NPLs may have reached higher proportions in earlier years with better accounting information and classification standards.

The following table seeks to provide some figures and relationships between NPLs and output collapse, and the role of state banks in these dynamics. In most countries, the state share of bank assets in the peak year for NPLs was fairly high. Considering the late recognition of many problem loans, bank assets in general would have been lower had there been earlier recognition (as warranted) in most transition countries. Likewise, the use of “private” banks for directed and

132 Armenia, Czech Republic, FYR Macedonia, Georgia, Lithuania, Poland, Slovenia and Tajikistan.133 The strategy required each of the banks to (i) raise specified amounts of additional capital over a short period; (ii) meet key prudential requirements on an accelerated basis (relative to other banks in the system); and (iii) suspend dividend payments (in two of the banks). In addition, a timetable was set for the removal of the main government representatives on bank boards throughout 1997. The agreements also envisaged that banks, if in compliance with prudential norms, would be compensated for the bad loans they made before 1996 under government pressure. With these loans estimated at about 2 billion drams, such compensation would have been equivalent to about 15.4 percent of total system bank capital at end 1997. (According to IFS, banking system capital was about 13 billion dram when netting out “other items” from bank capital.)134 Estonia, Hungary and Turkmenistan.135 Estonia liquidated problem banks and moved towards consolidation and full privatization of the banking system. Hungary introduced its revised Privatization Law in 1995 and moved aggressively with strategic privatization thereafter.

91

patronage lending understates the role of the “state” banks in the overall relationship, as many of these banks were effectively owned by state-owned enterprises, but considered “private.”

Table 6.2 Comparative Loan Portfolio Problems Among Transition CountriesYear of Worst

Output Performance

Peak NPLs as % of

Total Loans

Peak Year for NPLs

State % of Bank Assets in Peak NPL Year

NPL-State Share

CorrelationAlbania 1992 91 1997 90 HighArmenia 1993 36 1995 <3 LowAzerbaijan 1995 >60 1999 >70 HighBelarus 1995 16.5 1998 59.5 HighBulgaria 1997 15 1996 82 HighCroatia 1993 15 1998 38 MediumCzech Republic 1992 36 1994 N/A HighEstonia 1994 <4 1994 28 MediumFYR Macedonia 1995 42 1996 N/A HighGeorgia 1994 41 1995 46 HighHungary 1993 26 1993 75 HighKazakhstan 1998 20 1996 28 MediumKyrgyz Republic 1995 92 1994 77 HighLatvia 1993 20 1996 7 LowLithuania 1994 32 1996 55 HighMoldova 1999 17 1996 0 LowPoland 1991 36 1993 86 HighRomania 1992 57 1997 80 HighRussia 1998 6 1995 N/A HighSlovak Republic 1993 44 1998 50 HighSlovenia 1992 22 1994 40 HighTajikistan 1996 3 1996-98 30 HighTurkmenistan 1997 14 1997 68 HighUkraine 1999 N/A N/A N/A HighUzbekistan 1995 N/A N/A N/A HighNotes: Data not available for Bosnia-Herzegovina and YugoslaviaSource: IMF; EBRD; Finance & Development

On a more positive note, several countries show that private ownership has generally led to an improvement in loan quality and reduced intermediation spreads. Examples of this trend include Armenia, Bulgaria, Estonia, Hungary, Kazakhstan, Latvia, and Lithuania (in terms of loan quality).

Overall by region, NPLs have remained high in CEE, although they appear to have diminished in the Baltic states in particular, as well as in the CIS countries. However, as noted earlier, this is not fully verified due to differing accounting and audit standards, levels of disclosure, and varied approaches to loan classification.

Meanwhile, nominal net spreads have been cut in half in the CEE and CIS countries, while coming down a third in the Baltic states. While this does not address the supply of credit, it does suggest that more disciplined monetary policy, better enforcement of tougher prudential

92

requirements, and increased competition are bringing down net spreads. However, net spreads remain very high in CIS countries at more than 17 percent in 2000. They are also high in the CEE region at about 12 percent, and increasingly reasonable in the Baltic states at less than 7 percent. As these are unweighted averages, the distribution also shows that countries that have been the most persistent with macroeconomic stabilization, structural reform and strategic privatization in the banking sector have the lowest net spreads, as shown in the cases of the Czech Republic, Estonia, Hungary, Poland, the Slovak Republic and Slovenia.136 Kazakhstan has also shown low net spreads in recent years.

Notwithstanding these comparatively favorable examples, in most CIS countries and several non-CIS countries, needed banking reforms have been delayed, governance remains weak, and state banks (and some “private” banks) continue to be used as vehicles for non-commercial purposes. In many of the late reforming countries, state banks continue to account for major portions of bank assets, loans and deposits. However, due to poorly performing portfolios, loan and asset values would shrink considerably if proper provisioning and write-down practices were followed. Meanwhile, poor asset quality undermines earnings performance, slows capital formation, and props up high real intermediation costs.

Delayed reforms in these countries have generally correlated with more sluggish economic performance overall, except where a strategic commodity has been used as a source of cash to insulate the economy from the negative consequences of slow reform.137 Even where laws are adequate, institutional capacity has been slow to emerge. In many cases, issues of financial discipline, loan default, collateral, veracity of financial information, and other staples of market-based banking have not been adequately addressed or developed. The failure or disappearance of poorly performing banks has undermined public confidence in banks as a whole, particularly where there is no deposit insurance and people have lost their savings. Under commercial conditions, this has weakened incentives for lending. When commercial disputes occur, the judicial process has frequently proved itself to be debtor-friendly, further reducing incentives for banks to lend.

Meanwhile, because banks are now under pressure to comply with prudential regulations, they often make decisions that are consistent with the prevailing incentive structure. This has prompted a wholesale shift away from formal financial institutions in many CIS countries, where central banks have often tried to instill some discipline by bringing down high inflation rates and by seeking to achieve exchange rate stability (net of 1998-99 when exposure to the GKO market hit many fragile CIS economies). As an example, the stock of barter and arrears throughout the enterprise sector is many times higher than balance sheet figures of the banks. For example, in Ukraine, net enterprise arrears (payables) are estimated to be four to five times total credit from

136 There are some variations on this point. For example, Poland and the Slovak Republic delayed strategic privatization in the banking sector until the late 1990s, and Slovenia has been even slower in privatizing its major banks. Likewise, the Czech Republic only privatized Komercni in 2001. However, in general, these countries have pursued disciplined monetary policies throughout the years, often (e.g., Estonia, Slovenia) fairly disciplined fiscal policies, and been attractive to prime-rated investors due to favorable business environments and comparatively high levels of purchasing power.137 Several CIS countries have been able to leverage the effects of favorable oil and gas prices to present better economic results. However, should these prices decline, this would be expected to weaken economic indicators. To the extent the banking system is exposed to these trends, this would then adversely affect banking sector indicators.

93

the banks to the enterprise sector. By 1999, overdue payments were estimated to be 92 percent of GDP.138 In Russia, total arrears were estimated to be nearly 24 percent of 2000 GDP, with most of it to non-bank accounts (i.e., arrears to suppliers, tax accounts, and off-budget funds).139

Belarus has likewise experienced arrears of 19-23 percent of GDP since 1998.140

Nor has this problem been restricted to CIS countries. In Bulgaria, state enterprise arrears approximated 20 percent of GDP in 1999, although these have come down significantly since 1997. Arrears in 1999 to banks were only 1 percent of GDP, with the remaining 19 percent (of GDP) of arrears to suppliers, government, the state pension fund, employees and other accounts. In Croatia, arrears approximated 11.6 percent of 2001 GDP, and were as high as 20.1 percent of GDP in 1999.141 More seriously, Romania’s arrears climbed steadily from 1994, reaching 42 percent of 1999 GDP, most of it to suppliers, Government and other accounts, and 6.44 percent (of GDP) to banks.142 The following table provides a comprehensive breakdown of arrears among selected transition countries. Annex 5 provides specific detail by types of arrears.

Table 6.3 Synopsis of Arrears in Selected Transition Countries as a percent of GDP (in percent)1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

Azerbaijan 60.6 100.2 68.3 96.8 148.2 166.1 200.0 215.6 N/A N/ABelarus N/A N/A 30.4 13.5 18.2 13.3 23.1 19.2 22.4 19.1Bulgaria 68.8 60.6 47.5 41.7 66.3 27.9 24.1 19.9 N/A N/ACroatia N/A N/A 3.4 6.2 7.4 8.1 11.4 20.1 14.4 11.6Kyrgyz N/A N/A N/A 7.3 N/A N/A N/A 6.3 N/A N/ALithuania N/A N/A N/A N/A 9.3 9.0 N/A N/A N/A N/ARomania 34.6 23.9 26.1 25.2 36.1 33.7 36.2 42.2 N/A N/ARussia N/A N/A 14.8 13.3 23.4 29.1 47.8 30.3 23.7 N/AUkraine 8.0 6.0 13.0 20.0 24.0 85.0 98.0 N/A N/A N/ANotes: Enterprise arrears to government may not equal tax arrears since tax arrears include household arrears and enterprise arrears to government can include other forms of arrears.Sources: TACIS (2000) for Azerbaijan; 94-95 ECSPF (1998), IMF (2000, 2002) for Belarus; IMF (2001) for Bulgaria; World Bank (2001) for Croatia; IMF (2000) for the Kyrgyz Republic; ECSPF (1998) for Lithuania; IMF (1998, 2001) for Romania; 93-94 Bagratian and Gurgen (1997), Russian European Center for Economic Policy (2002) for Russia; and IMF (1999) for Ukraine. (All data for this table compiled by George Clarke.)

While payables alone are not the problem, long-term overdue payables are, and these are symptomatic of major problems in the overall payment system of the economy. That these enterprises are broadly cash-constrained or continue to maximize tax-avoidance opportunities at the expense of long-term enterprise performance and competitiveness is also a reflection of the unproductive incentives that have prevailed in many transition economies over the years. What is

138 Estimate by IMF. 139 According to the Russian European Center for Economic Policy, total arrears were 23.7 percent of 2000 GDP. This was constituted by 10.1 percent arrears to suppliers, 4.9 percent to tax accounts, 4.6 percent to off-budget funds, and 4.1 percent to banks.140 Figures are from the IMF. These include arrears to government, workers and suppliers.141 Figures are from the National Bank of Croatia and the IMF.142 Figures are from the IMF. Of the 42.2 percent GDP figure in 1999, arrears were to suppliers (18.0 percent), Government (8.3 percent), banks (6.4 percent) and other accounts (9.5 percent).

94

striking about this is that most of these enterprises are state-owned, or are “privatized” but running according to earlier methods. This has depleted cash and capital, and reduced or eliminated credit worthiness according to traditional commercial banking measures. Moreover, the poor condition of these state enterprises has reduced tax payments to the budget, including for social insurance, exacerbating the poor state of public finances.

Where banks are still used for non-commercial purposes, they are often (although not always) state-owned. As noted above, state banks generally account for a large share of balance sheet values of banking systems. Their non-commercial approaches have often been reminiscent of old-style management and governance, the use of directed lending for political purposes, the traditional orientation of the bank catering to state farms and state enterprises, and the social orientation of the bank’s culture (e.g., to finance production to meet output targets, to provide branches everywhere to ensure nobody is without a place to retrieve savings or to initiate a cash transfer, to ensure employment). Where there has been recognition of the unsustainability of such an approach, hard decisions to restructure, streamline or foreclose on the bank (if technically insolvent) have been put off because of the political difficulties associated with such a move. Often, this approach has led to a financial crisis in the banking system, high levels of corruption, and a major cost to the budget (or central bank resources).143 In cases where traditional savings banks have been used for other purposes (e.g., the Czech Republic in the mid-to-late 1990s to absorb smaller banks; several countries to channel loans that later became non-performing), this has also raised the issue of deposit safety and general levels of public confidence.

These problems have also pointed to weaknesses in corporate governance structures, although several countries have improved the accounting framework, introduced tougher requirements for internal operations, strengthened the internal audit function, restricted the issuance of licenses when “fit and proper” standards are not met, and called for more professional standards and qualifications for board members. Much of this has been consistent with general monetary policy strengthening, enhancement of banking supervision, implementation of stricter prudential norms, and requirements for increased integration into the global marketplace. While corporate governance performance (and the degree of preferential provisions for state banks) has varied from country to country, and still remains weak in many cases, there has been a gradual recognition of the need to address this weakness as a precondition for stable banking systems, and as a preventive measure in advance of potentially damaging bank collapses (and associated costs).

PROBLEM ASSETS, BANK RESTRUCTURING COSTS, AND APPROACHES IN TRANSITION COUNTRIES

Among the 27 transition countries, numerous approaches have been taken to address the problem of bad loans, and to build viable banking systems. These developments are clearly influenced by a multitude of factors, not the least of which are stable business environments, functioning institutions, stable macroeconomic frameworks, and credit worthy and “equity worthy”

143 To the extent that leakage in the system has increased outside the banking system, closing down troubled banks has been less of a problem as the issue of patronage has migrated to other parts of the economy (e.g., power companies).

95

enterprises and households. Solving financial sector issues in the absence of real sector progress has frequently led to frustrating results, stifling the ability of banks and other financial institutions to prosper even when inflation rates have been stabilized, banks have reformed, and funds are available for lending and investment.

In general, apart from some of the Balkan countries that have been torn by war and related conflagration, there is a general view that the northern CEE countries and Baltic states have broadly attended to their fundamental financial sector weaknesses, and that underlying stability is more likely to be sustained in these countries. There are clearly potential shocks and disturbances that could challenge this point of view, particularly in small open markets (e.g., Baltic states) where their economies can fluctuate significantly on a year-to-year basis, depending on developments in their major trading partners’ economies.144 Likewise, in countries such as Romania, there is a risk of future instability due to delayed progress on structural reform, notwithstanding their potentially large market and current improvement in economic performance. However, in general, there is a view that CEE and Baltic countries have pushed ahead with many of the needed reforms they have had to pursue to become competitive. The lure of accession to the European Union, to which all three Baltic states and seven CEE countries have been invited to negotiate, provides an incentive to these countries that is largely lacking in the CIS.

Evidence of the improvement in the overall environment of many CEE and Baltic markets has been found in the approach they have taken to the resolution of problem assets. This has taken many forms and approaches, yet NPLs are now much less of a problem for their banks than they are for CIS countries. In the case of the latter, most NPLs have been to either state enterprises and farms, or to “privatized” firms that have operated along traditional lines rather than competitive commercial criteria. The impact of the NPLs on CIS banks’ balance sheets has been devastating, just as they were costly for CEE and Baltic banks. Part of the legacy of these problems is the doubt surrounding after-tax earnings of CIS state banks relative to their counterparts in several CEE countries. This relates back to questionable loan classification standards, accounting practices, and associated issues of information management, reporting and disclosure that are widely viewed to be weaker in CIS than in many other transition countries.

Estimates of cost have varied in terms of how these bad loans and delays in resolving structural banking problems impacted CEE, Baltic and CIS banks. For example, one source notes that the costs of bank restructuring and deposit compensation ranged from 13-30 percent of GDP from 1991-98145 in FYR Macedonia (29.6 percent), Bulgaria (26.0 percent), the Czech Republic (20.9 percent), Kazakhstan (18.4 percent), and Hungary (12.9 percent).146 More recently, the National Bank of Croatia estimated the cost of bank rehabilitation to have approximated 26 percent of

144 This was demonstrated during the 1998 ruble crisis. Reforms in these countries, particularly in Estonia, also vindicated the position that they are able to maintain underlying stability on the condition that they continue to attend to structural reforms and remain an attractive destination for direct investment from their Scandinavian and Eurozone friends. 145 Costs are cumulative from 1991-98, and then measured against 1998 GDP.146 See E. Zoli, “Cost and Effectiveness of Banking Sector Restructuring in Transition Economies,” IMF Working Paper 01/157, October 2001.

96

GDP from 1991-2000.147 However, not all countries endured high explicit costs, with some managing to work out problems over time (e.g., Poland, Estonia, Lithuania), while others have experienced limited growth and development in their banking systems (e.g., Georgia, Kyrgyz Republic, Latvia).148 For example, the costs of bank restructuring and deposit compensation in these countries ranged from as low as 0.1 percent of GDP from 1991-98 in Georgia to 4.9 percent in the Kyrgyz Republic, with Estonia (1.4 percent), Poland (2.4 percent), Latvia (2.5 percent) and Lithuania (3.0 percent) showing lower explicit costs. However, these figures do not necessarily take implicit costs into account in the form of higher net spreads, foregone lending to sound companies for sound projects, foregone GDP growth, diversion of deposits out of the banking system, etc.

BOX 6.2 THE CHALLENGING EXPERIENCE OF REFORM IN UZBEKISTAN

[This box will be filled out in the next version after revisions are made.]

What is broadly agreed is that long-term plans to provide ongoing financing so companies could grow out of their problems have rarely been successful without major financial, managerial and operational restructuring. Many countries deferred major restructuring at the bank level, only to experience severe collapse late in the transition (e.g., Moldova, Russia, Ukraine all experienced their worst output performance in 1998-99). In other cases, countries engaged in costly bank restructuring programs, only to experience major collapse even afterwards (e.g., Bulgaria, Kazakhstan).149 Likewise, several countries (e.g., Bulgaria, Croatia, the Czech Republic, Hungary, Kazakhstan) have experienced multiple recapitalizations due to the insufficiency of original measures. The question in the last two cases is the degree to which banking systems adequately restructured, particularly as other countries engaged in less expensive restructuring exercises (e.g., Poland, Estonia), did so only once, and ultimately emerged as more competitive. However, in the case of the first approach that essentially assumed little restructuring at the bank level, there is fairly clear recognition that banks and economies could not grow out of their problems without some form of structural adjustment, and that delays on this front only perpetuated the problem and possibly added to the overall cost.150

In the case of CIS countries, enterprises, banks and depositors were all broadly left exposed and not bailed out through formal recapitalization mechanisms—although there are some exceptions, such as in Azerbaijan with the consolidation and recapitalization of four state-owned banks .151. However, CIS enterprises (usually state-owned, or formerly state-owned) were bailed out through bank rollovers of de facto non-performing loans, and the run-up of arrears. These

147 Cedo Maletic, “Overview of Croatian Banking Sector: the Causes and Cost of the Two Recent Banking Crises,” National Bank of Croatia, 2002.148 See E. Zoli, “Cost and Effectiveness of Banking Sector Restructuring in Transition Economies,” IMF Working Paper 01/157, October 2001.149 Output performance figures from S. Fischer and R. Sahay, “Taking Stock,” Finance & Development, September 2000.150 To be fair, some countries are simply at a disadvantage. Moldova is an example, with only three million people and limited resources, as opposed to Russia and Ukraine with their populations and opulent natural resources.

97

practices are still in evidence in many state-owned banks,152 and their continued practice has undermined efforts to achieve competitiveness in the economy and banking sector.

THE GENERAL IMPACT OF BAD LOAN QUALITY AND THE PACE OF REFORM

The practice of rollovers is rooted in imprudent loan classification and provisioning practices. In the process, this has deferred recognition of problem assets that ultimately have decapitalized and marginalized banks. As noted before, many CEE and Baltic banking systems introduced stricter prudential frameworks once banking system problems were uncovered. This often resulted from external audits of major banks according to IAS, and the major variances that were consequently uncovered relative to regulatory reports and norms. Once addressed and resolved, these banking systems generally moved on to restore confidence, to improve credit management skills, to boost resources and capital, and to become more effective intermediaries. However, where these problems have not been resolutely addressed, NPLs have continued to be a drag on bank liquidity, capital and general financial sector stability. This has applied to the banking system as a whole, and not been specific to state vs. private ownership.

Much of the problem for banks in lagging economies has related to distortions in the credit markets. Above all, given the reliance on secured transactions, there has been a major disconnect between the legal environment and financial values assigned to collateral. In most transition countries, bankruptcy legislation and enforcement have been weak. There has been progress in some countries, but judicial capacity building and a restructuring of incentives towards creditors (to increase their willingness to assume credit risk) has taken time. In most cases, the legal environment has been weak, and the prospects for seizing and selling collateral to recover part of bad loan values have been poor. Nonetheless, most banks have assigned higher values or inadequate risk weights relative to real market values of what has been pledged in support of the loans. This, in turn, has led to an overstatement of asset and capital values. By extension, this has been predicated on an understatement of risk, which can also lead to distorted (lower) pricing due to passive reliance on collateral and guarantees. Once problems emerge, this also provides an opportunity for corruption, resulting in additional financing for some troubled enterprises on clearly non-commercial lines. The prevalence of these relationships in the state sector has exacerbated problems in lagging economies.

The persistence of state banking problems has been dangerous to overall financial sector stability. State banks in particular are known for high cost-income ratios, excess overhead, much higher head count and manual processing, and a lack of modernization. This has made it more difficult for them to achieve profitability, or to do so sustainably and adequately relative to competitive needs (e.g., investment in needed technologies). Due to traditional earnings problems associated with state banks, this has often made them more willing to assume risk to compensate for losses. By extension, this has sometimes led to imprudent cross-ownership, or

151 There are now two major state banks in Azerbaijan. The International Bank of Azerbaijan is the former foreign trade and export-import bank. United Universal represents the new bank that consolidates the former industrial, agricultural and savings bank following a carve-out of bad assets from the earlier banks’ balance sheets. 152 See M. Builov, “Who Owns Russia: Russia’s Banking Sector,” The Russia Journal, January 25, 2002, with regard to the financing of enterprises in Russia.

98

traditional exposure to affiliates and related parties. In the absence of consolidated supervision, this has meant that state banks are vulnerable to large losses from exposures to leasing companies, insurance firms, pension funds, investment funds, and other financial services companies through ownership or exposure, as well as to enterprises. In particular, state savings banks have been a target for these ventures, given their household deposits and retail networks. In the end, the inability to manage these risks according to strict commercial guidelines has led to adverse effects. When they occur, they have undermined public confidence, and often required costly intervention by the state to contain losses.

The risk associated with imprudent cross-ownership and exposures applies to all institutions in the financial sector. The absence of consolidated supervision has contributed to these problems. However, this is only part of the problem. Imprudent cross-ownership and the move to universal operations in the absence of good governance and sound management have been harmful across the board. Many of the lingering problems faced by the Czech Republic relate back to weaknesses associated with cross-ownership. The relatively recent ownership by banks in investment funds in Romania has likewise raised questions about underlying financial sector stability. In CIS countries, the prevalence of loss-making “pocket” banks has pointed to the weak state of bank governance in most of these countries, partly due to the overall weakness of the economy and incentive structures in place. While these banks are “private,” they generally operate on non-commercial criteria until their losses mount, their liquidity dries up, and they are not able to benefit from regulatory forbearance. The failure of most CIS banks to meet even $5 million in capital shows how poorly capitalized the great majority of CIS banks are. Under such circumstances, particularly since many if not most remaining state banks are barely solvent or are technically insolvent, the continued existence of state banks delays competitiveness and market-based restructuring.

There is a regulatory and supervisory dimension to the problem of state banks. While banking supervision has been tightening for years in transition countries as an extension of monetary discipline, every system exercises a measure of forbearance. State banks have long benefited disproportionately from such forbearance to permit them time to restructure and recapitalize, even with privatization as an objective. In the case of privatization, this is legitimate, as private investors will not buy an insolvent institution. However, in many cases, the process of rehabilitation has been dragged out for non-commercial reasons, and this has distorted the playing field. Several countries have protected their markets from foreign competition while their state banks restructure. Meanwhile, because of the widespread and inconsistent use of regulatory forbearance, state banks are often able to operate with some protection, even when foreign investment is materializing in the banking sector. In the end, from an institutional perspective, preference shown for state banks undercuts the ability of bank supervisors to enforce their mandates in support of banking, financial sector, and monetary stability. This is a problem that is widespread in transition countries.

Ultimately, if state banks are protected and losses mount, there is a macroeconomic cost, as has been experienced in most transition economies. As the sole or major shareholders, the state has a financial obligation to recapitalize state banks as going concerns. Failure leads to a loss of confidence, so recapitalization is fairly continuous, as is liquidity support. Ultimately, the resources for this support are monetary or fiscal. More recently, as central banks have been fairly disciplined and focused on pricing stability, most of the leakage has come from the fiscal side.

99

These practices are political by definition, and often lacking in transparency. Financial support for state banks is often extra-budgetary, or where cash resources are constrained, through arrears (on taxes, social funds, electricity payments, and other obligations). Given the precarious fiscal and macroeconomic state of the weakest transition countries, this has generally represented a costly effort based on the use of scarce resources. As noted above, even where the explicit costs are not so high, the implicit costs can cause major economic damage. This can be in the form of forbearance and protection, thereby distorting the competitive environment. This can also be in the form of opportunity costs, such as higher net spreads that serve as a brake on lending flows and economic growth. Likewise, as noted above, these problems are not restricted to transition country banks.

REGIONAL FUNDING AND INTERMEDIATION TRENDS

In general, deposit trends, access to syndicated borrowings, and general levels of capital suggest that many of the CEE and Baltic states have managed to put in place structures that have helped to restore public, creditor and investor confidence in banks. This shows up in terms of deposit increases, broad money to GDP trends, and increasing capital. In general, CEE countries have been responsible for the greatest improvement in deposit mobilization and capital formation. The Baltic states’ banks have shown positive trends. In contrast, the CIS countries have shown more limited deposit mobilization (given the number of countries, people, banks, etc.), and significant decapitalization of their banks since 1995. The following table highlights some of these trends by country and region.

100

Table 6.4 Basic Funding Indicators by Country: 2000

Deposits +/-: 2000 vs. 1995

2000 Broad Money/GDP (%)

+/- Broad Money/GDP: 2000 vs. 1995

(%)Capital+/-:

2000 vs. 1995Albania 909 60.1 1.3 174Armenia 130 14.7 7.0 -28Azerbaijan 387 17.5 5.2 26Belarus -63 17.7 2.7 184Bosnia 247 27.6 12.8 3Bulgaria -4,420 34.8 -30.1 263Croatia 4,111 46.1 21.2 -570Czech Republic -1,517 75.7 -2.9 373Estonia 978 39.3 12.6 319FYR Macedonia 163 21.0 8.8 -468Georgia 108 10.3 2.8 28Hungary 1,623 46.8 4.9 611Kazakhstan 981 15.4 4.0 537Kyrgyz -7 11.9 -5.3 -53Latvia 815 30.4 7.0 184Lithuania 1,033 23.3 0.0 299Moldova 36 22.4 3.2 28Poland 28,506 42.7 8.8 1,500Romania 581 23.2 -1.9 -627Russia -1,871 22.1 4.2 -3,836Slovak Republic 561 67.8 3.1 1,992Slovenia 2,309 49.5 13.0 -188Tajikistan N/A 8.8 -12.0 0Turkmenistan 298 14.9 -3.8 21Ukraine 1,007 17.9 5.2 368Uzbekistan 2,384 11.9 -6.3 851Yugoslavia N/A 20.3 N/A N/ATOTAL 39,289 1,991CEE total 33,073 3,063Baltic total 2,826 802CIS total 3,390 -1,873Notes: Turkmenistan broad money figures for 1999Sources: IMF; World Bank

Reviewing the figures, Bulgaria, the Czech Republic and Russia have all experienced net outflows of deposits in their banks since 1995. In all three countries, a major shock or development has occurred.153 However, this has also been the case in several other countries, even though their deposit figures were positive in 2000 compared with 1995. In this regard, there appears to be little consistency. However, overall, CEE countries have increased their deposits significantly since 1995, about 10 times the increment generated in the CIS countries.

153 The Czech Republic has battled back against structural problems in the economy. Bulgaria faced economic collapse in 1996-97. Russia faced a collapse of the currency in 1998.

101

Meanwhile, broad money figures continue to show that CEE and Baltic state countries have higher intermediation levels than CIS. Overall broad money was greater than 30 percent of GDP in eight of 12 CEE countries and two of three Baltic states. However, in the CIS, there were no countries in this range, and only Russia and Moldova had above 20 percent ratios among the 12 CIS countries. Thus, in terms of a general funding base, the CIS countries remain especially weak.

This is made all the more apparent by the net $1.9 billion decrease in bank capital in CIS countries, compared with the nearly $3.9 billion increase in bank capital in CEE and Baltic banks. In terms of the former, it should be noted that Russia has experienced a significant drain on capital, whereas the rest of the CIS countries in the aggregate have shown improvement. If the capital figures are accurate, resource-rich Uzbekistan and Kazakhstan have shown significant capital improvement since 1995. Meanwhile, among all countries, the Slovak Republic showed a major increase in capital during the period of nearly $2 billion, partly the result of a determined effort to strengthen its two major banks (VUB and Slovenska Sporitelna) prior to privatization.154

Part of these trends also reflect bank operations in the regions. CEE and Baltic banks have attracted greater foreign direct investment in their markets, which has also helped domestic banks (directly when bought, indirectly through competition) to strengthen systems and diversify their product offerings. These improvements have helped to create a more positive earnings stream for CEE and Baltic banks. Their more recent diversification of earnings has been reflected in their growing array of products and services for businesses and households. By contrast, aside from some large banks, most banks in CIS countries tend to be small and relatively inconsequential as intermediaries. (This has also been true in some of the Balkan countries.)

As noted earlier, most banks in transition countries are small, with capital of only $22 million on average per bank. This is particularly true in the CIS region, where the average bank had $11 million in capital. Baltic banks had about $25 million, and CEE banks had $61 million. Breaking out the Balkan countries of CEE, average capital is also very small (under $20 million on average) in Bosnia-Herzegovina, FYR Macedonia and Romania, while being closer to the CEE average in Croatia ($45 million in 2000) and Slovenia ($60 million). Albania ($22 million) and Bulgaria ($29 million) were more like the Baltic states. What is noteworthy is that most banks are unable to access the international syndicated borrowing market if they stay at low levels of capital. In some cases, they are also unable to access the domestic inter-bank market, and only then at high rates. All of this combines to make it difficult to mobilize deposits, as it undercuts the ability of the smaller banks to make the necessary investment in products and services that would encourage depositors to place their funds with the banks.

In general, CEE and Baltic countries appear to have adopted a prudent enough framework for banks to operate in a stable fashion. While NPLs are high in many of these countries, there is better recognition of this as a problem than in earlier years. Moreover, as banking supervision capacity increases and private banks are required to report on prospective non-compliance and associated risks, the system is often moving forward in a way that contains potentially damaging risks from becoming dangerous to the system as a whole. This has been tested in several countries in recent years, without panic. The case of Hungary’s Postbank in early 1997 was an

154 These two banks have been privatized.

102

example of this, as the underlying condition of the bank led to deposit withdrawals. However, the system as a whole did not experience a major net outflow. Rather, deposits were simply re-deposited in stronger and better managed banks, essentially strengthening the system as a whole at the expense of some vested interests.

However, not all countries are free and clear from these destabilizing effects. Weak regulation and supervision in Romania culminated in a challenge to banks in 2000 when rumors circulated about the financial condition of major institutions. The National Bank signaled its willingness to provide needed liquidity support, and Romania’s major banks were able to handle increasing withdrawals. However, such rumors might not have had as strong an impact had a better supervisory regime been established earlier. This is rooted in the larger issue of efforts throughout the 1990s in Romania to defer needed reforms. This has translated in a more fragile economy, less public confidence in the banks, more difficult access to international capital markets (and at higher cost when accessible), and lower levels of capital.

The transition from the monobank system to a stable, well-funded two-tier banking system and diversified financial sector has been far more difficult in the CIS than in the other transition countries. In the CIS, the trend has been towards large state banks in most of the key resource-rich countries, and very small private banks that have served as pocket banks for insiders and controlling interests. The latter has undermined economic growth, compounding the after-effects of hyperinflation and the loss of confidence in domestic currencies. While some of the state banks have made efforts to professionalize and modernize, their existence has made it difficult to create an open competitive environment for banking. Meanwhile, banks such as Sberbank in Russia remain “strategic” and coveted due to the deposit share they have.155 Large CIS banks such as Sberbank, Vneshtorgbank and Vneschekonombank in Russia, the International Bank of Azerbaijan, Halyk Savings in Kazakhstan, Vneshekonombank in Turkmenistan, the National Bank for Foreign Economic Activity in Uzbekistan and Oschadny in Ukraine remain essential players in these countries. In some cases, this is because they are still vehicles of directed lending. In other cases, they are the caretakers of hard currency accounts and responsible for payment and settlement for the major enterprises of the economy. Meanwhile, much of the rest of the economy has fled the banking system, running up arrears to other accounts. All of this points to problems of funding in the real sector, governance and management in the economy at large, and the greater difficulties of CIS countries in dealing with the significant challenges of the transition.

155 Sberbank has about three quarters of the local currency deposit market and half of the foreign currency deposit market.

103

CHAPTER SEVEN: WHAT NEEDS TO BE DONE AND HOW TO GET THERE

GENERAL FINDINGS AND CONCLUSIONS

In most CIS countries and several non-CIS countries, governance remains weak, boards are not represented by people with sufficiently specialized banking skills, management is entrenched, accounting and financial information is incomplete or inaccurate, and state banks (and some “private” banks) continue to be used as vehicles for non-commercial purposes. Most transition countries still have a few state banks, on average about seven, and frequently savings banks and agricultural banks where household savings in local currency remain. In many countries, state banks continue to account for major portions of bank assets, loans, and deposits. However, due to poorly performing portfolios, loan and asset values would shrink considerably if proper provisioning and write-down practices were followed. Meanwhile, poor asset quality undermines earnings performance, slows capital formation, props up high real intermediation costs, and makes it difficult for these troubled banks to make the needed investment in systems and modernization to be competitive. Their continued quasi-fiscal function in many countries continues to pose a risk to general macroeconomic and financial sector stability, making it difficult to develop a fully competitive market in which the public has confidence.

Delayed reforms in several countries have generally correlated with more sluggish economic performance overall (except where a strategic commodity has been used as a source of cash to insulate the economy from the negative consequences of slow reform). Even where laws are adequate, institutional capacity has been slow to emerge. In many cases, issues of financial discipline, loan default, collateral, veracity of financial information, and other foundations of market-based banking have not been adequately addressed. The failure of poorly performing banks has undermined public confidence in banks as a whole, particularly where there is no deposit insurance and people have lost their savings. Under commercial conditions, this has weakened incentives for lending. When commercial disputes occur, the judicial process has proved itself to be debtor-friendly, further reducing incentives by banks to lend.

Meanwhile, as there has been growing recognition of the cost of directed lending and the use of state banks for quasi-fiscal purposes, banks are increasingly under pressure to comply with prudential regulations. Consequently, lending requires that enterprises and other prospective borrowers meet stricter credit worthiness criteria to be able to obtain loans. This requires that these borrowers have debt capacity, which is often based on higher levels of reported capital as well as sound cash flow, a strong management team, modern operations, and a vibrant market. To the extent that these basic requirements cannot be met, prospects for obtaining credit are often difficult and costly. Even then, the tenor of the credit may not be sufficient, particularly if the borrower has investment requirements for long-term development (as usually found in the industrial, mining, power and transport sectors). Bankers will often shy away from such decisions if the legal framework is inadequate for secured transactions, so often a problem in transition countries.

104

While the trends in CEE and Baltic banks have been more favorable recently, there has been a wholesale shift away from formal financial institutions in many CIS countries. While central banks have played a disciplined role in bringing down high inflation rates and achieving reasonable measures of exchange rate stability in this region (net of 1998-99 when exposure to the GKO market hit many fragile CIS economies), there has been only modest progress in the aggregate figures for credit and general intermediation. As noted earlier, many CIS countries have operated increasingly on barter, arrears and netting arrangements through the real and state-owned sector than through the banks. In places like Ukraine where net enterprise arrears (payables) are high, “veksels” (e.g., promissory notes) have been somewhat monetized through the system for years. Other countries like Moldova have significant financial flows running on this basis, bypassing and marginalizing the banking system as a whole.

These enterprises are broadly cash-constrained and continue to maximize tax-avoidance opportunities at the expense of long-term enterprise performance. This reflects the survival mentality of many troubled enterprises and regions. The poor condition of these enterprises has reduced tax payments to the budget, including for social insurance, exacerbating the poor state of public finances. All of this has accumulated over time in a loss of confidence in banks and other public institutions, making it difficult for many if not most transition countries to achieve sustainable and continuous growth. In this regard, many CIS countries have been hindered in their efforts, largely due to poor structures, incentives, and governance. However, the problem is not just restricted to CIS countries. The same fate has hindered growth in Romania, triggered collapse in Bulgaria, and hamstrung efforts in many other cases in southeastern Europe. Even where cross-border conflicts have not surfaced, internal problems of weak laws, inadequate regulation and supervision, connected lending and political patronage have eventually culminated in problems that have adversely affected banking trends and general economic development.

Where banks are still used for non-commercial purposes, they are often state-owned. As noted above, these banks generally account for a large share of balance sheet values of banking systems. Their non-commercial approaches often relate back to old-style management and governance, the use of directed lending for political purposes, the traditional orientation of the bank catering to state farms or state enterprises, and the social orientation of the bank’s culture. More often than not, this has led to a severe financial crisis in the banking system, high levels of corruption, and a major cost to the budget (or central bank resources). Where state banks are more modern than this, they are still constrained in their ability to achieve increased productivity, efficiency and competitiveness due to excessive head count, labor protection, weak skills, manual orientation, lagging technologies and information systems, old organizational patterns, and a lack of service culture.

In the end, troubled banks need to be either restructured and recapitalized, or liquidated. The latter is important as a means of consolidating systems and creating open conditions for market competition. Absent this, banks will not assume risk, and intermediation will be limited and distorted. Estonia is a good example of a country that pursued this approach. Along with the introduction of other reforms, that country’s banking system is now more stable and poised for growth than it would have been had it tried to nurse along troubled banks.

105

Meanwhile, most remaining state banks have poor prospects for privatization. In fact, in many of the most troubled countries where state banks continue to play a key role, privatization prospects are often less promising than they have been at earlier times, or by contrast with countries that have entered formal negotiations for entry into the European Union. The latter has served as a major incentive for many countries to stay focused on politically sensitive reforms over a long period that might otherwise not have occurred without such an incentive. That this option is not available to CIS countries represents a major disadvantage to the reform process of half the transition countries of our study.

Given poor privatization prospects as a starting point, probably the fastest way to modernize the banking system is to engage in a purchase and assumption exercise, allowing the market to determine what is salvageable among the state banks. This would send a signal that markets are open and transparent, and that the state is getting out of activities better left to the market. However, in some cases, restructuring is likely warranted due to the potential for strategic privatization. Whether this should precede privatization, or be carried out as part of the post-privatization process (with the government accepting defined costs for a certain period to attract strategic investors and allow them to restructure banks that would otherwise run up losses or distort the environment) is a case-by-case question that is subject to specific conditions of economies and the marketplace. For example, the timing for privatization in resource-rich countries would likely be better when commodity prices are high. By and large, the sooner these countries can bring in investment grade ownership, the sooner the benefits should accrue to the economy and marketplace.

Restructuring and privatization efforts need to be bolstered by effective supervision in support of consistent stability in the financial sector. This also depends on accurate, timely and complete information subject to sound standards of preparation and audit, for both internal (managerial) purposes as well as external (regulatory compliance, investor and creditor information) purposes. A stable macroeconomic and legal framework reinforced by clean government and stable politics would create most of the conditions needed for well-managed banks to compete effectively in the marketplace. The alternative is continued patronage, waste and corruption. The balance of the chapter looks at some of these preconditions for a competitive financial sector, with an effort to customize some of the recommendations based on regional and country-specific characteristics.

STATE BANK PRIVATIZATION AND RESOLUTION

Prospects for State Bank Privatization

In light of the weaknesses noted above, most state banks are generally uncompetitive, insolvent, loss making, and lacking in franchise value. In some cases, such as some of the specialized banks engaged in export-import financing, they have adequate systems and professional personnel who have been exposed to international norms of banking. In some CEE countries, the former state banks that were the easiest to privatize and that generated reasonable privatization proceeds were corporate banks with earlier traditions of international exposure. However, in other cases, these were often among the most troubled institutions. Thus, there is no consistent pattern.

Savings banks are often treated differently due to their prominent role in mobilizing household deposits. In many countries, they are the only banks with a retail network. In countries that have

106

poor infrastructure, such a retail network could potentially have value. However, more often than not, countries with poor infrastructure have poor economies. Thus, sorting through the problems of a state bank with high personnel levels, poor systems, and traditions of “social” banking are often not worth the price of privatizing, even if the price is zero. Meanwhile, in countries where infrastructure is adequate or prospects are improving, most modern banks tend to go for “brick and click” operations rather than traditional “brick and mortar.” The difficulties many banks have had in cross-selling (e.g., banking and insurance) have undercut the earlier perceived franchise value of extensive retail branch networks in most transition countries.

In the end, troubled banks either need to be restructured and recapitalized, or they need to be liquidated. The latter is important as a means of consolidating systems and creating open conditions for market competition. Absent this, banks will not assume risk, intermediation will be limited, and rates will be high. The costs of high net spreads serve as a punitive measure for both depositors and borrowers, while being a defensive necessity for banks against risk. In light of this, most state banks have poor prospects for privatization at this point. There are exceptions, particularly in countries where strategic resources exist, and purchasing power is increasing (particularly in the more highly populated markets). However, their definitive privatization or closure combined with a stable macroeconomic and legal framework (with adequate incentives for creditors) is a precondition for establishing a competitive environment for banking. Until then, it is unlikely that adequate capital will be in place for meaningful levels of intermediation. Absent capital and a conducive environment, it will take even more time to restore confidence.

If this effort is done via state banks, this again represents a very costly endeavor based on the use of scarce resources. This also runs the risk of state banks reverting back to traditional practices due to the perception of “market failure.” This would invite restored financing of old conglomerates and financial-industrial groups around job-creating, patronage and non-commercial criteria. This would then lead to a return to banks that could conceivably run up NPLs to levels that far exceed capital, adding to the costs of intermediation and raising future risks to system soundness and stability. All of this has occurred in the past, and some of these practices continue to exist in several countries. The high cost of these practices points to why transition economies need to move on to the final chapter of privatization in the banking sector based on principles of sound governance and management, accurate and timely disclosure of meaningful information, effective banking supervision, and developed legal systems reinforced by stable macroeconomic policies. However, because the banks and countries are often among the least attractive to investors and the most impaired as institutions, closing out this chapter represents a difficult challenge.

Preconditions for State Bank Privatization

State bank privatization (as with any other privatization) benefits from stable macroeconomic conditions, a reliable legal framework, credible financial information that is adequately disclosed, signals from government of firm commitment to competitive banking environments, financial discipline and competitiveness in the real sector, and adequate systems and personnel to build on acquired franchise value. Not all of these are preconditions for the privatization of state banks. In particular, outside investors generally want a free hand to reorganize, restructure, (re)train, and introduce their own operating systems based on lessons and experience elsewhere. Thus, systems and personnel in an existing institution are not considered indispensable on the

107

condition they can be replaced. Likewise, foreign investors have entered markets where the real economy is uncompetitive and distorted. However, in general, where these conditions prevail, there has been far less foreign investment in financial services than would otherwise have been achieved, and an exodus of such investment has attested to negative views of economic and market prospects. More often than not, countries that have weak legal systems have performed poorly. At a minimum for lenders and investors, legal framework weaknesses reduce the willingness of these institutions to take risk. At a minimum, this keeps them focused on “cherry-picking” while overall intermediation statistics remain low.

Reversing these weaknesses and building on whatever existing strengths have been achieved seem to be the minimum needed to bring the problem of state banks to a close. Maintaining a stable macroeconomic framework is one of the key preconditions, and this is an area where most transition countries have made major progress. As noted in the text, even countries that have shown weak economic performance have often shown improvements from even weaker positions early in the transition period. Most countries have had to battle hyperinflation at some point. Today, inflation rates rarely exceed 10 percent in CEE and Baltic countries,156 and about half of CIS countries have single-digit inflation rates.157 While the average of 15.6 percent for transition countries is still high compared with many strong economies around the globe, it is a vast improvement from triple digits as recently as 1995.158 Likewise, fiscal deficits are now generally less than 4-5 percent in all transition countries. Again, this is high compared with strong economies, yet a major improvement relative to earlier fiscal balances, particularly in the CIS and some Balkan countries.159 At a minimum, the concept of macroeconomic stability is a precondition for successful resolution of state bank issues. Weakness only reduces their prospects for privatization, and increases the prospects for losses at these banks.

A second precondition is a sound legal framework. Laws are often in place, but court capacity, precedent, experience, commercial training, and alternative dispute resolution mechanisms are often lacking. Specifically for financial firms, the unpredictability of the judicial process reduces the incentive to assume risk. While there are numerous out-of-court options, there is a general view that a sound legal framework provides a more conducive environment for risk assumption. This provides an incentive to borrowers to be disciplined, and for shareholders to more properly monitor their investments. It also provides a framework for disputes to be settled out of court in an orderly and professional manner. Courts can then be used as a final arbiter should out-of-court methods not succeed. Courts can also be used for minimum thresholds of disputes, while smaller cases are automatically sent to specialized arenas. All of this is in contrast to the “informal”

156 Exceptions are FRY Yugoslavia and Romania. Both of these countries are currently engaged in programs that will get these rates under control and eventually into single digits. 157 Armenia, Azerbaijan, Georgia, Kazakhstan and the Kyrgyz Republic all had inflations rates of less than 9 percent in 2001. Moldova, Turkmenistan and Ukraine were in the 11-13 percent range. Only Belarus, Russia, Tajikistan and Uzbekistan had rates above 20 percent, with Belarus having the highest at 60 percent. 158 According to the EBRD, the average for all transition countries was 176 percent in 1995. The peak was in 1994 at 1,262 percent. The results are somewhat skewed by the poor performance of the CIS, Romania and Yugoslavia. Most other CEE and Baltic states have had inflation rates in single digits since 1997-98. 159 On this front, CIS countries have shown steady progress over the years since fiscal deficits peaked in 1992-93. CEE countries have shown increases in recent years, while the Baltic states have maintained strict fiscal discipline and low deficits.

108

methods frequently employed in countries where the legal framework and judicial systems are less developed, and where decision-making is more arbitrary, less transparent, and counter-productive in broadening market participation.

A third precondition is sound financial information. This cannot be taken for granted anywhere, and is particularly problematic in markets where information disclosure is not a tradition. Moreover, as the market test for many asset values is not in place in less developed economies (e.g., real estate values, valuing collateral), valuation is a challenge along with the normal issues of loan classification, soundness of government finances (as underwriter of securities held by state banks), and general audit standards. It is not uncommon for many investors to uncover additional bad loans in acquired banks after privatization deals have been closed. This may be due to the acquiring bank’s own internal deficiencies, or market developments that reduce the quality of some assets after the transaction that were not identified as risks beforehand. However, in other cases, it is also due to poor internal records, lack of consolidated accounting, etc. All of this adds to the risk of investing in that market, ultimately reducing investor interest, reducing the amount investors are willing to pay for a state bank, and adding to the cost that is incurred by the state to make the privatization transaction happen.

A fourth precondition is independent supervision based on a sound prudential framework. While there is always an interim economic and financial cost to dealing with structural weaknesses and losses, not addressing these problems only perpetuates the myth that results are better than they are. Overstatement of the performance of loans and the quality of assets (through rollovers and capitalization of unpaid interest) only serves to perpetuate the notion that many banks have high earnings, strong capital, and robust operations. In the end, as they run short on cash, they have to approach the government for refinancing. Eventually, such liquidity shortfalls ultimately point back to these banks’ insolvency and the need for corrective action. This, in turn, is rooted in the excessively high cost structures and general lack of competitiveness of their real sector borrowers. Tightening up on bank prudential norms and providing the supervisory authorities with a mandate to enforce is a way for banks to become more disciplined and properly managed. Ultimately, this should be transferred in the process to the real sector as a condition for obtaining loans. As an extension of this precondition, this includes a supervisory mandate to enforce sanctions as needed on state banks to ensure their compliance. In many (if not most) countries, the supervisory mandate is extended to most private banks, but not equally applied to state banks. This distorts the market, and provides undue forbearance that is costly in the long run. To the extent that some private banks are also able to escape the supervisory mandate is an added distortion.

A fifth precondition is for governments to signal and seriously enforce their commitment to an open and competitive banking environment. By extension, this should apply to the economy in general as an effort to encourage responsible direct investment. In most cases, this is an absolute precondition to maximize investor interest. It is also necessary (along with the other conditions noted above) to encourage intermediation. Where distortions continue to exist, many of the banks that are most able to lend choose not to do so to avoid the risk associated with the uneven environment. The withdrawal of state institutions from the market provides opportunity for private banks. It has been demonstrated that net intermediation spreads are lower in markets that are the most competitive and backed by a stable macroeconomic framework. Continuing to float state banks in the market only delays movement towards the level of competitiveness needed to

109

increase the supply of products and services, and to bring pricing to levels that are attractive to depositors, borrowers, and shareholders. There are many small and weak economies that would see little initial benefit to a wholly privatized banking system, as has already been the case in some of the CIS countries (e.g., Georgia) or in the Balkans (e.g., FYR Macedonia, where there is only one state bank that is relatively inactive). However, it is more than arguable that achieving a competitive banking environment and economy is a building process that takes time. Moreover, there are many other examples of where movement towards privatization has generated benefits that would otherwise not have occurred (e.g., Bulgaria today as opposed to 1996-97).

A sixth precondition is telecommunications capacity, including adequate facilities for electronic commerce, ATMs, and an infrastructure of support to ensure operations run without interruption. As countries move increasingly to closer integration of systems (e.g., banks and central banks) and real time gross settlement, this prospect improves. This is important for low cost entry into consumer and retail banking, as well as for countries with large areas and dispersed resources away from major population centers (e.g., Russia, Kazakhstan).

STATE BANK PRIVATIZATION AND RESOLUTION RECOMMENDATIONS

General Approach

Transition countries should, as a general matter, approach their remaining state banks as resolution cases. While there are a few exceptions, most carry relatively high levels of bad loans, and investments in government securities as a high proportion of earning assets and reflecting earlier bad loans. In cases where this helps to increase earnings enough to recapitalize, it is due to relatively high yields on bonds that are costly to the budget. To the extent that earnings from these securities are not sufficient to recapitalize, they call into question the ability of these banks to generate sufficient earnings from operations. This, in turn, raises questions of what behavior they will assume to generate such earnings. One option is to seek political favors, which is already a significant problem in most transition countries (and throughout much of the world) and usually culminates in major losses that are costly to the economy (and often depositors). A second option is to engage in adverse selection in the hopes of generating “extraordinary” or “abnormal” profits, which also usually culminates in a major charge to these institutions, as well as often to government, depositors, and sometimes, shareholders. In either case, these are undesirable scenarios that can more broadly undermine confidence in banks and weaken efforts to maintain and sustain stable financial markets.

To avoid these risks, it is recommended that governments design strategies to effectively eliminate state banking within a prescribed time period. This can be done fairly systematically, in a manner that reinforces some of the efforts already under way to create a stable environment, and as a strategic framework for accelerated institutional capacity building where institutions continue to lag what is needed for effective market performance. Many countries have already begun this effort by introducing BIS-recommended prudential norms, and requiring banks to design their own corrective actions to be in compliance with liquidity, solvency and other requirements. In cases where new systems need to be developed (e.g., internal audit, MIS, new charts of accounts, electronic compliance with UBPRs), there has been an understanding by the authorities that time, training and resources are needed for these systems to be introduced and effectively implemented. Often, this has involved not only development within individual banks,

110

but harmonizing these systems with those of the regulatory authorities. Thus, in terms of drawing up programs of corrective action for non-compliant banks, this is not a new development. However, at this juncture, given the limited after-tax earnings (or losses) of most state banks relative to their assets, more needs to be done than to simply comply with regulatory norms. New injections of capital, better management systems, closer links to regional and global markets, innovative systems, and modernized personnel based on international professional standards are all needed to achieve commercial viability.

There are a few broad options available. These include restructuring or rehabilitation prior to or after privatization, consolidation of banks prior to and after privatization, liquidation or reorientation of savings banks within well defined mandates and risk parameters, and incentives for the establishment of non-bank lenders. These are discussed below.

Restructuring or Rehabilitation Prior To or After Privatization

Many countries have already engaged in significant restructuring and rehabilitation efforts, particularly in CEE countries. While some have kept explicit costs down (e.g., Poland), many others have experienced high costs as a percentage of GDP (e.g., Bulgaria, Croatia, FYR Macedonia, Kazakhstan, Czech Republic, Hungary). As results in these countries have varied, there is no hard and fast rule about whether to proceed with this approach. Likewise, countries like Georgia and Moldova have kept their restructuring costs low, yet have little banking intermediation to show for their efforts (largely due to larger issues involving political and macroeconomic instability, and weaknesses in the real sector). Other countries with relatively weak economies that have fallen in between have also faced recurrent problems, such as in the Kyrgyz Republic, where restructuring costs approximated 5 percent of 1998 GDP.

Should countries opt for this approach, it should be structured to include a particular strategic objective in mind, and to establish explicit performance indicators based on a series of operational reforms, managerial changes, and financial requirements. Given the small capital of many of these banks once properly adjusted for risk, one approach would be to seek a minimum threshold of absolute capital combined with specific capital adequacy targets that point to future soundness. While these figures would vary bank-by-bank, average capital figures in markets where banks appear to be functioning relatively well (as measured by after-tax earnings, asset quality, revenue growth, sustainable net interest margins, competitive cost-to-income ratios, ROA, ROE, etc.) suggest that capital should strive to be at least in the $50-$100 million range, and preferably higher. This is based on most banks not being able to generate after-tax earnings of much more than 15-20 percent of average equity. Among the state banks for which figures were available, only 22 state banks exceeded 15 percent ROE in 2000, and many of these were small banks that generated only a few million dollars. The importance of these measures is that even with high ROE and ROA, very low aggregate earnings prevent the opportunity for such a bank to acquire needed systems and technologies to provide adequate risk management for modern banking, and adequate support and variety in terms of products and services for the public.

Thus, in approaching a state bank, the question is whether the investment put into restructuring and rehabilitation is going to generate an adequate return relative to other uses of those same resources. Irrespective of size, banks might be more efficiently privatized by having the

111

Government insist on certain activities being undertaken to ensure the private sector has access to financing, with the Government benefiting from the fiscal revenue and employment generation that results from such intermediation,160 rather than going through the time-consuming process of restructuring and rehabilitation. If the bank is large, this approach would likely involve larger sums, and a longer time horizon for new owners and management to achieve such economic objectives. If the bank is small, it might not be worth the effort or time to restructure and rehabilitate. With most state banks having less than $300 million in assets and less than $50 million in capital, it appears that making concessions on sales prices and developing a five-to-10-year time horizon for desirable financial and economic objectives from outside investment would more fully benefit banking sector modernization than working through a time-consuming rehabilitation and restructuring approach prior to privatization.

This would appear to apply to most countries, with specific bank exceptions at this point being PKO BP of Poland (already undergoing restructuring with the intention to eventually privatize), BCR of Romania (the country’s largest bank that is already slated for privatization), Sberbank and Vneshtorgbank of Russia, and Nova Ljubljanska and Nova Kreditna Maribor of Slovenia. Otherwise, banks that are not supervisory concerns and are not large should ordinarily be sold off quickly, with the long-term financial and economic objectives serving as the main focus of negotiations for sale.

The only other possible exception should be savings banks, but only where a significant portion of household deposits is held in these banks. The analysis of the deposit issue should, however, take into account foreign currency deposits, not just local currency deposits. In some cases, savings banks have a substantial portion of local currency household deposits, but most of the banking system’s assets (and deposits) are held in foreign currency, while substantial sums are also held outside the banking system. In such a scenario, the privatization (or holding) strategy should not be driven by local currency household deposits as a reason to reject privatization. Economies that have become increasingly dollarized and show much broader dispersion of bank market shares in hard currency deposits should accelerate measures to reduce the comparatively high levels of concentration of local currency deposits in savings banks. The latter gives them undue influence in inter-bank markets, makes it more difficult for more modern banks to access local funding and compete, perpetuates political patronage through the banking system, and rewards what is often complacent and sluggish management in a period when such institutions need an infusion of capital, know-how, and modern management systems. Examples of savings banks that should accelerate internal restructuring and be privatized promptly are DSK in Bulgaria, CEC in Romania, Postbank in Hungary, the Latvian Savings Bank, Sberbank in Tajikistan (if it can find a buyer), the Uzbek Housing Savings Bank in Uzbekistan, and Oschadny in Ukraine.

Consolidation Prior To and After Privatization

As an extension of the post-privatization approach described above, lack of investor interest for banks even under the “deferred benefit” scenario might trigger the need for closure of these

160 This was the negotiating position taken by the Czech government when privatizing Ceska Sporitelna, requiring the purchasing bank to commit to a specified level of housing finance and venture capital as a condition for the guarantee provided by the Government for half of Ceska Sporitelna’s loan portfolio. Other governments have taken similar positions when privatizing their banks.

112

banks, or the unsalvageable portions of banks that have been privatized. This should be recognized as a contingency for which to plan and should be considered a normal part of the negotiating process, as investors may not want the full “franchise” and associated costs of spinning off or restructuring unwanted parts. This is not just a financial issue, but also a managerial and operational issue that drains time and energy.

One alternative, in the appropriate circumstances, would be a “purchase and assumption” approach which hives off unwanted parts of one franchise to another bank that might be able to give them some value. For example, a small bank that might want to develop a branch network might value the very same branches that a potential owner of another bank does not want as part of the privatization process. While the ideal approach would be through the market in a normal acquisition, given the limited market in many transition countries, this might be done through the regulatory authorities. They should be in a position to know whether the potential acquiring or absorbing bank meets the financial and related conditions for such an acquisition. Should the units being hived off be of questionable existing value, the regulators would be in a position to know whether the acquiring bank had the capacity to turn them around. Should such an approach present risks to deposit safety or systemic stability, the same assets could be offered to financial institutions that are not deposit taking, or are not covered under the deposit insurance fund. In any case, post-privatization consolidation should be considered at a minimum as a contingency by the countries with remaining state banks. Should there be no interest in the banks, or in the unsalvageable parts of the banks, this would then lead to closure.

It is important to note that several countries have engaged in consolidation prior to privatization, and it has often turned out to be a difficult and costly exercise. In the Czech Republic, an effort was made to help Ceska Sporitelna to diversify from a fairly narrow savings bank to a more diversified commercial bank by absorbing smaller regional banks. The end result was negative for Ceska Sporitelna, which was only privatized in 2000. Such a privatization could have occurred five years before had the consolidation strategy not been pursued. In Poland, Pekao SA (now owned by Unicredito of Italy) had franchise value in the mid-1990s. However, the Government of Poland opted to consolidate three regional banks with Pekao SA. The exercise turned out to be costly time-wise and complex due to the regional specifics of the three smaller banks. In the end, there are doubts about whether such a consolidation added value, and if it enhanced the level of banking services in the Polish economy. This is currently a challenge in Azerbaijan, as United Universal is in the process of amalgamating the traditional savings, industrial and agricultural banks after balance sheet restructuring and the carve-out of bad loans from the earlier banks to a collection agency.

In summary, consolidation of banks prior to privatization may be useful in reducing the transaction costs of negotiating privatization transactions, particularly when investment banks are used as the major agent for those transactions. However, in general, they are complex, difficult to carry out successfully, and often negative or sub-optimal in their results. It is more likely that most of the remaining state banks could be consolidated more efficiently by simply offering them for sale to banks, domestic and foreign. This would achieve consolidation through market mechanisms.

To the extent that market mechanisms are not sufficiently available to consolidate the banking system, the regulatory approach is an option. As with the restructuring and rehabilitation

113

approach, a separate administration could be responsible for working with specialists to establish a consolidation plan that would be the counterpart to the corrective action plan described above. This would include a strategic objective with performance indicators. Such an approach would require the two or more institutions to establish an action plan and time line for the achievement of objectives, with financial results serving as the key barometer for success. This could include performance incentives, although reasonable indicators would need to be set early on. From the state’s point of view, the benefit would presumably come from cost savings, and enhanced intermediation over time. Should consolidation involve at least partial privatization from the outside, this would include the added benefit of enhanced solvency in the system. However, in general, consolidation would likely be far less complex if individual state banks were acquired by other banks already well established in the marketplace, and with the capital and managerial capacity to integrate what is salvageable from these banks as expeditiously as possible. Again, with cost savings as a benefit to the government, it should be possible to structure financial and economic targets for the consolidation on a long-term basis.

BOX 7.1 CESKA SPORITELNA AND BANKING SECTOR CONSOLIDATION IN THE CZECH REPUBLIC

In 2000, the Czech government sold Ceska Sporitelna, the state savings bank and the country’s second largest bank, to Erste Bank of Austria. The sale brought to a close one of the government’s most lengthy and difficult bank privatizations. The sale of Ceska Sporitelna came at great expense to the government as it followed several years of consolidation and restructuring that culminated in a massive government bailout during the final year before the sale. The government strategy proved far more costly than originally expected and, most likely, than if it had pursued privatization more vigorously at an earlier stage.

Ceska Sporitelna was the state savings bank during the socialist era and remains the largest retail bank in the Czech Republic. As of 2000, it had $12 billion in assets, a 34 percent market share in retail savings, and a network of 934 branches. As part of the Czech government’s financial sector restructuring during the mid-1990s, Ceska Sporitelna was included in the first wave of Czech privatization programs.

Recognizing that the commercial banks created from the monobank system inherited significant stocks of non-performing loans from the central planning era, the government developed a two-staged program to financially restructure and then privatize the new banks. A total of 37 percent of the Ceska Sporitelna’s shares were offered for vouchers and 20 percent was sold to towns and municipalities, while 40 percent was retained by the state. The government pursued this policy as well for its other three major state-owned banks, including Komercni, Investicni, and Obchodni. By the end 1995, these four banks in total accounted for 62 percent of banking system assets, and were 47-63 percent divested by vouchers, with the state-owned National Property Fund retaining the largest block. While not fully privatized, these banks were effectively “corporatized” with the expectation that full privatization would occur after additional restructuring and as accession to the EU neared.

Ceska Sporitelna languished in its quasi-privatized status until mid-1999, when the government began to speed up its planned sale of a majority stake in the bank. In the meantime, the bank’s prospects had been hurt by poor lending decisions that resulted in an increase in non-performing loans, and by the operational complications imposed on it by having to absorb several smaller banks as part of an exercise to consolidate the banking sector while providing the bank with lending “expertise.” By mid-1999, the bank was expected to lose $389 million—the equivalent of more than half of the bank’s capital. According to reports at the time, the bank needed to cover the loss by writing off part of its capital, which would then result in its capital adequacy ratio falling below the legal minimum set by the Czech National Bank. This forced the state to intervene to recapitalize the bank, and increased the incentive for the government to privatize the bank.

Although several foreign banks expressed an interest in Ceska Sporitelna, the government felt that some of their offers insufficiently valued the bank’s franchise. Rather than launching a formal tender, the government entered exclusive negotiations with Erste. In the end, although Erste raised its bid, the bank paid only approximately 1.55

114

times book value for the Ceska Sporitelna. The state also was forced to make significant concessions. In particular, it gave Erste Bank five-year guarantees on about half of Ceska Sporitelna’s loans. Prior to the sale, approximately $1.1 billion of non-performing loans were transferred to the state, which also increased the bank’s share capital by $201 million to bolster attractiveness to foreign buyers.

Liquidation

The liquidation of banks has been mentioned above through purchase and assumption techniques that could be woven into a post-privatization restructuring or consolidation approach. The reason to mention it separately is because many governments have simply resisted this option. The difficulty has often come with savings banks or agricultural banks, with their branch networks and traditional catering to households and under-served rural regions.

In general, it has been difficult to liquidate these banks in the past, irrespective of their financial condition for political reasons, and because there has been little alternative for people in rural areas. Savings banks have also catered to many older people who have small accounts, and are not comfortable changing banks or using electronic methods of banking.

While these are understandable reasons for deferring closure, the reality is that these banks can often be liquidated fairly easily, with few of the feared social consequences emerging. First, there are usually alternative institutions, such as savings houses, credit unions and micro-finance institutions that are fully capable of serving households and rural communities. Their development has often been stunted by politically-inspired protection provided to savings and agricultural banks. Moreover, under the cover of protection of small households and rural communities, these banks have typically been used for patronage purposes. In the end, many of the banks have run up losses because of the political patronage that has come from directed lending to large blocks of the population (e.g., through agricultural lending, loans for housing through savings banks).

State-owned savings banks showed losses or zero earnings in Armenia, Belarus, Kazakhstan, Tajikistan, and Ukraine in 2000. Meanwhile, after-tax earnings at state savings banks exceeded $10 million in only Albania, Poland, Romania and Russia. Likewise, state-owned agricultural banks showed losses or zero earnings in Bulgaria and Romania (now private), while showing after-tax earnings in excess of $10 million only in Poland. In general, the meager earnings of state banks raise questions about long-term sustainability. The large number of loss makers raises the question of potential costs to the budget, and the added costs to the system of carrying loss-makers. In the long run, if a bank cannot be privatized or absorbed by a sound and stable bank, it should be a candidate for liquidation. If it is determined that the bank shows no potential for commercial viability, liquidation should occur promptly. Branches can be spun off to other interested parties, including non-bank financial institutions such as credit unions and micro-finance groups, that focus on rural communities, households, and other market segments often neglected by mainstream banks. These banks can also be spun off to specialized finance companies, such as commercial finance and leasing firms. Under such circumstances, the sale would simply involve a revocation of the right of such an institution to market itself as a “bank”, explicit and well publicized removal of such an institution from deposit insurance, and other conditions that would transform the troubled bank into a non-bank financial institution. In this

115

sense, consolidation in the banking sector would proceed while the economy benefits from an expansion of non-bank financial services. If through corrective action a bank can move from being a supervisory concern to one that meets certain needs in the market, that approach should be taken on the condition that the forbearance costs are reduced rapidly so as to eliminate distortions to a competitive marketplace.

Narrow Banking Licenses for Savings Banks

Given the weak financial condition of most transition economy savings banks, yet the unwillingness of many governments to part with these banks, one alternative is to re-define the license and range of activities permissible for savings banks. DSK of Bulgaria is an example of a traditional local currency savings bank that has pursued basic reform within well-defined risk parameters, with a gradual increase over time in its relatively low credit limits. The Savings Bank of Albania, currently in the process of being privatized, is another example of a savings bank that was originally limited to a narrow range of activities, took on a more diverse role, effectively failed, and now has been sufficiently recapitalized and reorganized to be offered for sale to strategic investors.161 Other savings banks are likely to follow suit in the coming years.

However, there are risks to the delay inherent in this approach. For example,, as appears to be occurring in Ukraine, these banks can be misused for high risk connected lending. This has long been the main problem with troubled banks in transition countries. While savings banks have generally not been closely linked to banking sector crises in transition countries, there is always the potential for misuse, particularly as other state banks disappear as patronage vehicles. If these banks cannot be privatized, there should be an effort to provide some limitations on the kinds of lending and investment activities they can pursue. CEC (Romania) had exposures to an investment fund that collapsed in 2000, jeopardizing its financial position. Several years ago (as noted), Ceska Sporitelna was used as an anchor for the consolidation of several smaller banks. An earlier government also used Slovenska Sporitlena (Slovak Republic) to serve as a channel for directed lending to preferred SOEs. In short, limits should be placed on the risk-oriented activities of savings banks if they hold a major share of the deposit market (e.g., more than 25 percent of the local and/or foreign currency deposit market).

There are strengths and weaknesses of such an approach. By placing strict limits on what savings banks can do, there is a good chance of keeping them solvent. On the other hand, their high costs and low level of service translate into limited earnings, or losses. Meanwhile, because of limitations on what they can do on the asset side (or otherwise) to generate income, this almost certainly ensures that the banks will never be able to grow or modernize unless they retain significant shares of the deposit market, as is the situation in Russia with Sberbank. But this is also risky for systemic stability. Banks generally lack an inexpensive source of funding if they do not have a viable deposit base. This reduces their resource base for lending, and drives up lending rates. It also makes the inter-bank market highly dependent on a particular bank for liquidity, and it can create a circuitous dynamic leading to collapse if the savings bank is closely intertwined with government finances. If the government is dependent on the savings bank for funding and experiences a major downturn in its own finances, this could lead to a major erosion

161 As of 1Q 2002, there was no guarantee of success. However, the preliminary stages of the process had been carried out, and there was optimism at the time that Savings Bank would be effectively privatized by end 2002.

116

of the savings bank’s liquidity position, driving up rates in the inter-bank market or causing panic should there be excess concentration. In general, countries should make an effort to break up excessive concentration in the deposit market for greater funding balance. Beyond that, the savings banks themselves should be privatized, merged or consolidated in some form.

Non-Bank Lenders and Equity, Turnaround, and Workout Institutions

One of the reasons why troubled economies have been slow to finalize privatization of banks, particularly in “strategic” sectors or roles, is because of the underdevelopment of other financial services firms. In general, capital markets are underdeveloped, insurance penetration is limited, and second and third pillar pension funds are new or have not yet been introduced. All of this has translated into limited institutional investment, and limited market development. This is important not only in terms of financial flows and investment, but also because of the foregone benefits that could potentially occur as a result of enhanced governance.

From the lending side, there has also been a shortage of non-bank creditors active in many markets. The benefit of non-bank creditors is that they can provide loans without putting depositors at risk. This is typically found in specialized commercial finance, leasing and factoring companies. Often, banks themselves are involved in these businesses. In general, it would be helpful in many transition countries if legal, tax, accounting and related issues could be resolved to encourage entry of these institutions into the marketplace. In particular, favorable tax and accounting rules for equipment leasing could serve as a spur for manufacturing and capital-intensive businesses that might otherwise have problems getting loans from banks. Likewise, an improved environment for secured transactions would create incentives for such lending from banks to the risk-takers.

In addition, problems affecting the banking sectors of many transition countries are more often rooted in real sector weaknesses. In some cases, it is poor management, weak financial management, lack of market research or information, outmoded technology, etc. In other cases, it is because of inadequate organization of information, under-utilization of the company’s competitive strengths, etc. The entry of turnaround funds into the market would help to create the potential for streamlined operations and better financial results. Above all, with the introduction of management teams that have performance-based contracts, there is a better chance that these companies can be salvaged in part as going concerns, or that they can be merged to be competitive. While this is another case-by-case situation, most transition countries have resisted serious turnaround operations that could then securitize debt or equity packages within a period of years for their own payout. While such an approach presumes deep discounts for companies to be turned around, such a discount would need to be weighed against the ongoing costs of keeping the enterprises afloat via subsidies, transfers, and arrears. In general, and in the CIS countries in particular, this approach has been costly to the economy as a whole. From a financial sector perspective, the prospect of companies being turned around provides more of a potential market for banks, as well as for broader market development through venture funds and eventual securitization.

In terms of general scale of enterprise, most transition countries have SMEs and micro-enterprises. In addition, as noted above, because many of the remaining state banks are savings and agricultural banks that cater to households, alternative savings and credit institutions might

117

be more useful to the economy than banks that are dependent on costly branch networks. Credit unions and savings houses can certainly play this role, subject to sound prudential norms, adequate supervisory capacity to examine operations to ensure the willingness and ability of these institutions to comply with prudential norms, and careful consideration of permissible activities under the license. This would imply low thresholds of lending exposure, limitations on open foreign currency positions, etc. Likewise, there are several examples of micro-finance groups building their loan portfolios based on sound quality and performance standards. In many cases, these groups can lend in the $10,000-$20,000 range, and sometimes higher. While this is not sufficient for medium-sized or larger enterprises, it does help to finance small businesses and little household operations. In general, these kinds of institutions are often far more appropriate for the vast majority of households. However, there has been periodic resistance to their formation due to the added cost to banking supervision at a time when effective supervision is just getting rooted in transition countries. This is an understandable view on the part of supervisors. However, with proper guidelines and staffing, such non-bank institutions should be encouraged to operate in the interest of increasing intermediation services throughout the economy.

As an extension of this last option, there should be incentives in place for credit histories to be fully documented so that as businesses grow and increase their debt capacity, they are able to present their documentation to banks attesting to credit worthiness. Such incentives would be in the form of commercial credit information services having access to information, and being able to disseminate such information as requested. However, at the moment, this is not an available service in most transition countries. Over time, businesses should be more willing to disclose information when their performance has been positive and such disclosure would improve prospects for obtaining loans. As banking systems become more competitive and expand into retail and consumer banking, these kinds of services will be increasingly in demand.

WHAT SHOULD DONORS DO?

General Framework

There are a number of initiatives that donors can undertake to accelerate the commercialization of the transition countries’ banking systems. As a general notion, policy and institutional support for financial sector stability is key. This is macroeconomic and monetary as well as structural. A part of this effort includes clean-up of NPLs on banks’ balance sheet prior to or simultaneous with any major movement forward on banking reform. However, a focus on speed should be paramount to avoid the potential risk of higher costs to the economy mounting through state banks or pocket banks that can serve a similar purpose. At this juncture, governments should be less concerned about generating revenues from bank sales, and more concerned about having a well-designed series of post-privatization benefits that would be expected to accrue to the institution and the economy.

Effective and independent supervision is also a key link in the chain between the design and implementation of monetary (and, by extension, macroeconomic) policy and a stable environment for market players. Given general movement (in a phased manner) towards risk-based precepts of regulation and supervision, high standards of governance and management are required for stable financial intermediation. For boards to play their proper role, they need to be

118

qualified, and have access to timely and accurate information. The role of an autonomous internal audit function with access to fully functioning information systems is essential for such information to flow accordingly. Meanwhile, professional standards of disclosure are required when risks emerge that could have a material impact on the solvency of the institution, or the general stability of the financial markets. Adequate legal and supervisory mandates need to be in place to ensure that board members, managers and others are held accountable. Criminal penalties need to be invoked should major violations occur.

As for the ingredients for specific country strategies for bank privatization, these should include establishment of more comprehensive and useful data bases for management and strategic planning purposes, performance-based incentives and the adoption of modern human resource management tools, improvements in governance, technical assistance and training needs to build human capital, and better communication and interaction between market players and regulatory officials. This last suggestion includes the utilization of supervisory tools to assist management with improved financial monitoring, better operating systems and increased coordination in the identification of risks to avert the need for more formalized corrective actions. These could all be framed within the context of larger global efforts to comply with standards that reinforce financial sector stability, cross-border cooperation, and transparency for market-based decision-making.

With regard to actual privatization-related assistance, a general program could be drawn up to assist different countries and banks at different stages of development, or with differing prospects for strategic investment from firms that conform to international norms and best practices. General recommendations are broken out into two broad sections, involving pre-privatization and post-privatization assistance.

Pre-Privatization Assistance

Pre-privatization assistance would apply to countries and state banks where attracting investment-grade investors is not a likely prospect. For example, the Kyrgyz Republic has successfully pursued significant banking sector reform, yet has experienced only limited direct investment in the real economy (mainly in the gold sector, which is declining) and has been unsuccessful at attracting major international banks. With a small population and relatively remote location, it has been difficult for the Kyrgyz Republic to achieve its investment objectives. On the other hand, there are many countries with potentially more favorable prospects that could benefit from pre-privatization assistance. These countries include most of the CIS where state banks remain, as well as some of the Balkan countries. In many cases, elements of pre-privatization assistance could be useful to most transition countries, even where there are no longer any state banks (e.g., Armenia, Georgia), where there will soon be no state banks (e.g., Albania, Moldova), or where the reform process is well under way. In these cases, improved management information and public sector performance monitoring would benefit these countries even if state banks are limited or non-existent in number and influence.

Pre-privatization assistance could effectively involve a unified (or customized) program that is based on the following fundamental building blocks: (i) development of effective management information systems for enhanced performance at state banks (and, by extension and with customization, enterprises); (ii) strengthened standards of governance and accountability as part

119

of a merit-based human resource management strategy; (iii) building more effective capacity of governments to oversee the performance of state-owned banks, leading up to their privatization (and to provide them with more informed negotiating positions); and (iv) providing assistance to governments for them to more ably manage investment banks, professional management firms and other agents active in the pre-privatization and privatization process. These are discussed below.

Management Information Systems. Many transition countries’ banking sectors are still undermined by weak and incomplete financial data, the absence of sound accounting and audit standards and capacity, and the inability to use existing information for meaningful strategic planning, management and operational purposes. There have been improvements in recent years as new reporting requirements have been introduced for banks, prudential norms have been toughened, and banking supervision has become more effective. However, weaknesses still permeate many systems. These weaknesses result in owners failing to have the information needed to provide proper oversight of management, which impairs their ability to hold management accountable for poor performance and losses. One way to remedy this weakness would be to establish a performance tracking mechanism and database for state-owned banks based on commercial considerations. This would also reconcile with regulatory requirements, and with international accounting standards.162

The database would contain appropriate financial accounting (e.g., balance sheets, income statements, flow-of-funds), sensitivity analyses and scenarios to anticipate endogenous and exogenous risks, and e-based peer information by activity, sector, and size. Much of this may already exist in early warning systems, off-site supervision databases, etc. However, even if they exist, they are sometimes not available in centralized form. Thus, their use for contingency planning, performance monitoring and risk management is often undermined by fragmented control and a lack of administrative and policy coordination. The database would seek to construct timely financial statements in a manner that is as consistent with IAS as is feasible. This would provide an accurate representation of the value, returns and profits or losses of the individual bank to the overall state-owned portfolio, and the impact this has on government finances.

In addition, a more comprehensive flow-of-funds database for state banks detailing lending flows, arrears and cross-ownership positions would be of great use to policy makers. Information on arrears and netting arrangements is inconsistent in quality across countries, yet constitutes a major economic and structural challenge in most transition economies. Such a database would provide the government with a tool that would be helpful for budgetary purposes, as well as for the tracking of state banks’ portfolio risks.

Once an inventory has been constructed, governments could put contractual arrangements into place that are more commercial in their orientation.163 This has been done in several countries already as part of the pre-privatization restructuring effort. Contractual agreements would serve as a catalyst to reduce chronic losses of state banks (and enterprises), and to unwind some of the

162 This approach would be equally applicable to state-owned enterprises and utilities, for similar reasons.163 For guidance on bank governance contracts for enhanced performance, see R. Roulier, “Bank Governance Contracts: Establishing Goals and Accountability in Bank Restructuring,” World Bank Discussion Paper 308, 1995.

120

arrears of SOEs to banks. “Stock” and “flow” measures would be essential parts of this database, as well as key components to restructuring plans and contractual agreements on which compensation and promotions would be partially awarded.

More broadly, development of this database would provide clear and quantifiable information to assess management, financial and operational performance based on commercial indicators. In its most comprehensive format, the database would include macroeconomic indicators honing in on reductions in fiscal and quasi-fiscal losses, as well as microeconomic or firm-specific indicators. Bank data would draw on existing data reported to the monetary and supervisory authorities and focus on liquidity and capital ratios, earnings sustainability measures, asset quality indicators, cost and income measures, productivity measures (e.g., revenue/head count, profitability/employee), and systemic risk issues. In the last case, a key issue would be the financial condition of savings banks, their role in the inter-bank market and government financing, the implications of reduced deposit market share, the costs-benefits of privatization, and any specific contingencies that would need to be in place to enhance the competitive environment once privatized. Likewise, for remaining agricultural banks, the database would also need to take into account primary sector indicators and risks, and the role of these banks in agriculture sector development.

The database would primarily serve senior government officials who are managers of the state banks and enterprises. This should ultimately lead to improved financial and management information for more efficient resource allocation, better overall operating performance, and enhanced prospects for commercial viability and privatization.

Strengthened Governance and Merit-based Human Resource Management. Traditionally in the state sector, human resource management practices have operated on the basis of political patronage rather than commercial capacity, professional skills, or objective standards of merit and performance. This practice has often detracted from state bank performance, and has intensified losses at the firm level as well as at the government (budgetary and extra-budgetary) level.

One method for correcting this problem is to establish a performance evaluation mechanism for banks based on new standards of transparency and accountability at management and board levels. This would be based on more clearly defined standards of hiring and dismissal, and reporting on performance based on commercial criteria. The purpose of this mechanism would be to develop a culture of management accountability, to reward good performance, and to change management teams and structures when performance at firms can be markedly improved. This might require a change in many countries’ overly protective labor codes, but this is long overdue.

Overall human resource management would be enhanced by developing guidelines for professional standards of management and governance for state banks. Broad guidelines and standards would provide a framework for individual banks to establish suitable job descriptions for senior managers and board members, to introduce objective standards for hiring, and to implement performance-based standards for dismissal and compensation. These principles would be integral parts of overall strategic plans developed by management and approved by bank boards, with specific targets and objectives mapped out to serve as a basis for performance

121

evaluation. Individualized job descriptions and performance evaluations could be introduced, at a minimum, for senior managers, department heads, and others responsible for resource management and implementation of strategic plans. The focus of board members in reviewing management performance would be how well they have performed in reducing losses or increasing gains to shareholders (in this case, the state), and how well their individual or group performs when compared with targets set in job descriptions and contractual agreements.

To the extent that boards, management and employees are able to outperform peer competitors without any additional preferences or forbearance from the state, this performance would also be subject to recognition and reward. However, given the objective of privatization, there would be clear limits on the financial rewards offered, as one of the key objectives would be to retain maximum earnings to boost capital, not to pay these out in benefits and dividends prior to privatization. Capitalization of these benefits into shares would be an option, recognizing that excessive compensation would dilute ownership and undermine privatization prospects when the bank is up for sale.164

Implementation of improved human resource management would be predicated on practices of transparency and accountability, and strengthened governance standards and requirements. All senior management positions could be advertised, with objective criteria established for selection. Efforts should be made to contain patronage as a basis for hiring and retention, shifting the focus to commercial results. Best international practices should be applied to promote professionalization of management ranks, fairness and objectivity in the hiring process. Likewise, dismissal criteria should be based on unacceptable performance relative to goals and objectives based on realistic strategic plans. These plans, managers hired, and oversight criteria applied would all be focused on achieving commercial viability as a forerunner to privatization. At a minimum, the focus would be on material reductions in losses at the bank level to contain fiscal and quasi-fiscal losses at the government finance level.

To reinforce the importance of transparency and accountability, regular disclosure of performance in the form of quarterly financial statements, annual assessments of management performance relative to targets, and accomplishments and failures of banks in meeting their contractually agreed targets would be recommended. These reports would be public, given that the banks are state-owned, and their financing is based on public resources.

Government Oversight Capacity for State Bank Performance: Just as many transition country human resource management practices have traditionally been decided on the basis of political patronage at the bank level, standards set by government have encouraged and reinforced these practices. Thus, poor financial performance at the state bank level has often resulted from state-oriented patronage rather than commercial viability as the primary goal. This has been most explicitly manifested in directed lending practices, which have ultimately culminated in portfolio erosion and loss of capital. (These practices have also been problematic for many small private banks that lack sound governance and management.)

164 Such rewards would need to be sufficiently large to provide meaningful incentives for improved performance, yet sufficiently small as to contain leverage or dilution. Leverage would come from the issuance of debt instruments to finance compensation. Dilution would come from the issuance of preference or common shares to employees.

122

Developing capacity for government oversight of financial sector performance would be critical to general economic management, as well as to ongoing financial sector development and stability. It is recommended that this be done in a comprehensive manner in the form of a Management Performance Enhancement Commission at the most senior levels of government. This would clearly involve more than state banks, and would effectively constitute an effort to revamp the entire public administration framework towards professionalization, compression, and rationalization. The role of state bank performance would be essential not only for privatization purposes, but also as a source of financial discipline on SOEs. While this might create some tension, this would at least create a clear framework for the explicit identification of losses, the budgetary impact, and a program of adjustment to operate within agreed parameters. Meanwhile, more positively, the reduction of losses and potential for profitability in the state banks as a result of such efforts would automatically improve the prospects for banking privatization due to better performance at both micro and macro levels.

In effect, the starting point would be the professionalization of management and governance standards in the banking sector to establish well capitalized and liquid banks with improving prospects for financial sustainability, effective intermediation, and sound risk management practices. The commission would be responsible for establishing indicators and systems required to properly oversee management performance of state banks, to ascertain the impact of performance on government finances and progress in restructuring, and to formulate compensation packages and awards. The commission would also be empowered (with assistance as needed) to develop guidelines and procedures for disclosure of bank results and management performance, including timing requirements, media channels, and information requirements.

It is assumed that the commission would work with banking supervision to specify information that would help the regulatory authorities assess management capacity within the context of financial sector stability and the 25 Core Principles (Basle Committee on Banking Supervision). This would also apply to other financial sector regulatory standards (e.g., IAIS for insurance, IOSCO for capital markets) to the extent that state banks are permitted to operate in these fields. Hence, as found in market economies and at the core of many transition countries’ existing legal and prudential frameworks for banking, this effort would be designed to reinforce financial sector stability, and to reduce or eliminate the “connectedness” of ownership structures and exposure patterns that are often still found between banks and SOEs in many transition countries. Notwithstanding some of the efforts of recent years to prevent such damaging practices, these problems continue to weaken economic performance in many transition countries. The value added of the Management Performance Enhancement Commission would be in the form of applying effective governance and strict scrutiny of management plans and performance.

The goal would be to train analysts to be able to monitor the performance of state banks and make recommendations to decision-makers. These analysts would also be trained to implement performance contracts with managers of state banks. To accomplish these goals, the Management Performance Enhancement Commission would complement banking supervision by overseeing performance in the largest state banks based on clearly specified commercial criteria that would also be completely consistent with prudential regulations.

As noted above in the management information component, the financial sector database would be designed to provide clear and quantifiable information to assess state bank management,

123

financial and operational performance based on commercial incentives. This would include macroeconomic indicators honing in on reductions in fiscal and quasi-fiscal losses (or potential revenue increases), as well as bank-specific indicators and prospects for privatization. The database would be complementary to existing data with banking supervision, would serve senior government officials who are managers of the state banks, and would help the commission hold bank managers to better performance standards.

The establishment of the database should be combined with needed training and technical assistance to increase administrative capacity and efficiency. The focus would be to build institutional capacity to ensure bank management was performing according to contractually agreed targets and requirements. As with state enterprises under scrutiny, this component would help to develop the government’s technical capacity to manage large state banks and to facilitate restructuring. These efforts would also help to reinforce efforts by banking supervision to monitor and enforce prudential regulations as they apply to “fit and proper” standards of management in the operation of the major state banks.

Parallel to efforts in the enterprise sector, extensive training would be provided to develop the appropriate skills needed for government officials to analyze the financial statements of large state banks, to approve and monitor the implementation of restructuring and commercialization plans for these banks, to develop performance contracts based on appropriate criteria, and to ensure that state banks contribute to economic stability and underlying safety and soundness in the banking and financial sectors. Thus, technical assistance and training would be provided for operational requirements, such as the development of contractual necessities for qualified experts (domestic and international), as well as for broader governance and oversight responsibilities. A central reporting unit focused on banks would be trained and established to analyze the financial statements and make recommendations to decision-makers. This group would coordinate with banking supervision as part of a broader effort to stabilize the banking system.

The commission’s mandate to monitor state bank performance would need to be clear, as would its ability to make decisions on performance-based compensation awards, and the hiring and dismissal of managers. The commission’s mandate would also be subject to public scrutiny, particularly as public confidence in the banking system is a necessary condition for sustainable economic development.

Monitoring Contractor Performance. Recognizing that some of the expertise required for restructuring, privatization and ongoing performance evaluation would be outsourced, it would be helpful for governments to be able to more actively and professionally administer these arrangements and monitor performance. In many cases, governments have been too passive on the assumption that investment bankers, management advisory services, and other contractors were first-rate professionals who would get the job done efficiently, within budget, and according to schedule. In some cases, the situation was too complex for such objectives to be met. In other cases, incompetence or unsuitability of specific contractors undermined performance. At the other end of the spectrum, governments have sometimes meddled and interfered, making it impossible for paid professionals to do the jobs they were contracted to do.

Given these ongoing challenges, efforts should be made to create and enhance the capacity of government to implement privatizations of state banks with help from investment banks and

124

professional management companies (i.e., turnkey operations). This would require that funds be available for transaction-related preparation requiring the expertise of investment banks and others for due diligence, preparation of the prospectus, marketing efforts to attract interest/, handling inquiries, etc. For this to proceed properly, the Management Performance Enhancement Commission would be responsible for developing appropriate TORs and SOWs, as well as handling, managing and overseeing contracting and performance. These efforts should provide incentives for meaningful privatization accomplishments based on strategic investment, with the objective that technical assistance, training and other assistance would lead to tangible and measurable results at the end of a specific implementation period. In the end, the commission would be responsible for (i) identifying the most feasible methods of privatization based on advice supplied by international investment banks or professional management firms, including mergers and consolidation options; and (ii) establishing financial and economic parameters to justify the optimal transaction approach focused on long-term economic development objectives (e.g., macroeconomic stability, increased competition, direct investment flows, market links, increased productivity and efficiency). These would ultimately guide the government’s privatization strategy and specific negotiating position on remaining state banks.

Post-Privatization Assistance

While CIS and Balkan countries are often faced with the challenges of restoring confidence in banks, strengthening governance standards and public administration, and resolving macroeconomic and structural fundamentals to increase investment flows, several CEE and all three Baltic states are faced with the challenges of economic integration with Europe. In several countries that are considered more advanced in the reform process, state banks remain. These are often comparatively small at this point (by domestic and European measures), although some exceptions remain. For example, both PKO BP and BGZ in Poland are fairly sizeable banks165

that have taken more than a decade to restructure. In Slovenia, its largest banks remain state-owned.166 Hungary and the Czech Republic likewise have a handful of state banks. Among the first wave of EU accession candidates from the transition countries, only Estonia has eliminated state banks. Thus, as a catalyst to definitively close out state banking in these and other countries, a series of post-privatization measures might be useful to consider. These would be particularly helpful to second wave EU candidate countries (i.e., Bulgaria, Latvia, Lithuania, Slovak Republic, Romania), as well as other countries (e.g.., Croatia) that are now performing well, but have not received invitations for accession negotiations. This post-privatization assistance might also provide some guidance on future activities in markets where “pre-privatization” assistance might currently be useful. There is also the possibility of blending components in places like Russia, where there would already appear to be a ready market.

More specifically, the fundamental components of post-privatization assistance would involve fuller development of professional advisory services for banks (and mid-market firms) as well as efforts to promote increased listings on local and regional exchanges. These forms of assistance would be most appropriate in the more advanced transition country markets. However, they would also be useful as a framework for large state banks in CIS and other countries that could

165 Combined, these banks had more than $21 billion in assets at end 2000.166 Nova Ljubljanska, Nova Kreditna Maribor and SKB combined had about $8 billion in assets at end 2000.

125

potentially envision privatization in the coming years (e.g., Azerbaijan, Belarus, Russia, Kazakhstan, Ukraine, Uzbekistan).167

Professional Advisory Services. EU “first-tier” candidate countries (and some others) have made substantial progress in recent years in restructuring and privatizing the enterprise and banking sectors. In the real sector, legal and regulatory frameworks have been progressively harmonized with EU directives and international standards. Corporate governance standards have improved. Trade and direct investment have increased in total and proportion with EU markets. Access to financing has increased. Meanwhile, for banking sectors, most major institutions have been recapitalized and privatized, intermediation rates have increased, and systemic financial risk has diminished as new systems and management have been put into place, prudential regulatory frameworks have matured, and more timely and accurate financial information have been made available to the public and to supervisory authorities.

Notwithstanding these favorable developments, a large and potentially significant number of mid-sized banks that have been privatized have often come up short in achieving needed managerial, operational and financial conditions for the next threshold of growth. In the banking sector, this has often been due to their lack of strategic investment, which has kept capital levels low and prevented the development of a broader array of fee-generating services to offset risks associated with traditional lending and investment patterns.

There are several examples of insufficient development of professional services for medium-scale firms and banks. In the accounting and auditing realm, the Big Five168 provide services to the largest firms able to pay their fees, while local accounting firms primarily trained in domestic tax accounting attend to the needs of smaller enterprises (and sometimes banks) at far lower rates. There has been limited development of a middle-market for accounting, audit and, by extension, needed business advisory and consulting services that are trained in international standards. A cursory review of the professional services market in EU leading candidate countries shows limited investment on the part of “second-tier” accounting and audit firms from Western markets, and limited capacity of local firms versed in IAS, ISA and other international standards. The same has been true with law firms, as local firms are rarely versed in the skills and contacts for modern corporate practices, while internationally renowned firms are generally too expensive for many of the medium-sized firms and banks.

A dearth of business management and advisory firms specialized in banking and financial services and able to prepare needed business or financial plans and market or feasibility studies has also undercut growth. While some of these firms exist, they often lack the databases, international contacts and experience in best practices and standards to provide mid-sized banks with the plans needed for growth.

167 Some of these countries’ state banks have received restructuring assistance. One example is Oschadny in Ukraine, which has benefited from considerable technical assistance funded by the EU. This is consistent with the first component—professional advisory services. However, there has been no movement towards listing of shares. Latvia’s Unibanka likewise received assistance prior to its privatization. In other cases, they have received investment from external sources.168 At the time of writing this document, Arthur Andersen was still considered part of the “Big 5”, despite significant fracturing in the post-Enron period. By the time the process winds down, the accounting profession may be more commonly referred to as the “Big 4”.

126

Some of this gap could be closed by providing technical assistance seed funds for mid-market service providers in conjunction with criteria set by securities commissions for listing requirements on primary stock exchanges. At a minimum, this would include a focus on the preparation of (i) financial statements that conform to IAS and comply with securities market requirements; (ii) business plans for banks based on merger or acquisition strategies in local markets, and consolidation potential with larger banks whose operations are usually headquartered in G-24169 countries; (iii) performance-based contracts to improve returns and justify compensation patterns; (iv) restructuring plans to meet financial goals, management performance targets, and operational efficiency measures to qualify for institutional investment through publicly-traded exchanges; (v) market studies for increased knowledge of competitive risks and peer comparisons; (vi) feasibility studies to clarify investment strategies; and (vii) corporate (re)organization and tax strategies.

Bank Listings on Local and Regional Exchanges: For existing state banks, and for privatized or private banks that remain small and are unlikely to play a useful intermediary role in the coming years, a strategic plan focused on public listing as an objective would serve to trigger the kinds of internal changes needed to continue as a going concern. Thus, assistance could focus on working closely with securities commissions and institutional investors for increased listings on exchanges. Specific technical assistance would rely on services specified above and provided by specialized advisory services to banks as “input” for bank listings, with increased capitalization of banks under sound ownership serving as the output goal.

Specific requirements to make this happen would be assistance for (i) due diligence of banks when ready to be offered for sale, merger or consolidation; (ii) prospectus development; (iii) valuation of IPO and shares; and (iv) the preparation of legal documents for distribution prior to offering. For newly privatized banks, access to additional financing could help banks achieve adequate levels of capital to increase liquidity channels, allow for investment in new product offerings and services, enhance risk management capacity and systems, and build up reserves for unanticipated losses that might otherwise lead to insolvency, forbearance, or the need for corrective action.

Assistance in this regard would also be expected to improve standards of corporate governance. Stimulating movement towards listings on publicly traded markets should also encourage movement towards professional standards of governance, and active oversight of management performance based on approved strategic plans. This would help to counter the often diluted and non-strategic shareholdings of recently privatized firms that tend to be weak in their oversight of management performance. Compensation awards for professional management and strategic investment (including venture firms) would also more likely be based on financial incentives made possible with IPOs on major exchanges than for banks that do not have prospects for public trading.

Key implementation partners would be securities commissions and exchanges, mutual funds and pension funds, insurance companies, accounting firms, and law firms. This assistance would also serve as an outreach vehicle for securities commissions and markets to increase listings consistent with criteria shaped by growth and stability objectives. The latter would comply with

169 This is a reference to the 24 OECD members from a few years back. Today, the OECD has 30 members.

127

guidance from BIS, IOSCO and the Joint Forum for Financial Stability. Meanwhile, the former would comply with EU directives on securities markets and “collective undertakings.” In the process, increased listings would also benefit the banking system by reducing dependence on commercial bank credit for financing. For countries that are not candidates for EU accession, international best practice and standards of the above mentioned groups would be considered more than satisfactory as guidelines, while making it possible to conform to EU requirements if and as needed.

128

ANNEX 1: Financial Profile of State Banks: 2000 (million US dollars)

Names of Public Banks Assets Loans Deposits CapitalNet Income

Actual No. Employees

Albania Savings Bank 1,230 10 1,176 33 26 N/AArmenia Armenian Savings Bank 9 3 9 0 0 N/AAzerbaijan International Bank of Azerbaijan 615 184 474 19 9 N/A

United Universal Bank 38 0 11 5 1 2,590Belarus Belpromstroibank 299 149 239 44 5 5,192

Belagroprombank 283 220 153 113 5 7,267Belbusinessbank 150 79 117 11 2 N/ABelgazprombank 35 5 21 11 0 419Belarusbank Savings Bank 779 579 622 108 -2 N/ABelvnesheconombank 201 73 179 19 0 2,126

Bosnia-Herzegovina Federation Investment Bank 62 25 N/A 59 N/A N/A

Banjalucka Banka 42 19 27 10 N/A N/APrivredna Banka (PBS) Sarajevo 34 9 24 5 N/A N/ACentral Profit 162 47 115 29 1 N/AGospodarska Mostar 24 8 19 3 0 82Data unavailable for PBS Srpska Sarajevo, PBS Doboj, PBS Prijedor, PBS Gradiska, PBS Brcko, UNA Bihac, Sipad, Postanska, Ljubljanska, Semberska, Postanska sed., Kristal, Rosvojna and Agroprom

Bulgaria DSK Bank 588 271 492 79 8 5,697Commercial Bank Biochim 248 72 223 23 5 N/ACentral Cooperative 95 86 67 13 0 N/A

Croatia Dubrovacka Banka 401 153 231 14 -12 567Croatia Banka 159 82 102 15 -11 N/ASplitska Banka 997 462 884 67 6 1,070Croatian Bank for Reconstruction & Development 693 298 223 397 11 N/AHrvatska Postanska Bank 222 114 164 28 -21 172Riadria Banka 172 65 128 26 -5 N/A

Czech Republic Komercni Banka 11,000 3,000 9,000 1,000 12 N/ACeska Exportni Banka 631 17 153 51 2 N/ACeskomoravka Zarucni a Rozvojavo Banka 1,126 681 762 93 2 N/A

Estonia No state banks leftGeorgia No state banks leftHungary Postbank and Savings Bank Corp 1,193 400 986 148 3 N/A

Hungarian Development Bank 738 199 328 356 -20 N/AKazakhstan Halyk Savings Bank 707 342 616 48 -2 N/A

Eximbank 71 43 32 28 -4 N/AKyrgyz Republic Kairat Bank 7 0 5 1 1 N/A

Savings and Settlement Company 5 0 3 1 0 N/AEnergo Bank 6 2 5 1 0 N/A

Latvia Mortgage and Land Bank of Latvia 123 87 59 13 1 N/ALatvian Savings Bank 246 62 222 8 1 1,234

Lithuania Lithuanian Savnigs Bank 830 238 711 55 -8 3,586Agricultural Bank of Lithuania 417 201 322 32 2 1,769

FYR Macedonian Development Bank N/A 14 N/A 14 0 N/A

129

ANNEX 1: Financial Profile of State Banks: 2000 (million US dollars)

Names of Public Banks Assets Loans Deposits CapitalNet Income

Actual No. Employees

MacedoniaMoldova Banca de Economii 35 13 25 4 3 N/APoland Powszechna Kasa Oszczednosci BP 16,627 6,872 15,129 594 153 N/A

Bank Gospodarki Zywnosciowej 4,408 2,033 3,776 214 24 N/ARomania Banca Comerciala Romana 2,769 754 1,874 539 117 N/A

Savings Bank (CEC) 880 56 750 110 26 12,832EXIM Bank 203 29 35 28 -2 N/ABanca Agricola 448 313 471 26 -85 5,837

Russian Federation Sberbank 20,000 9,000 18,000 1,000 403 197,122

Vneshtorgbank 4,414 962 2,704 1,599 170 3,669Vneschekonombank 2,599 272 2,415 119 1 1,465Russian Development Bank 168 N/A 37 131 1 105Rosselkhozbank N/A N/A N/A N/A N/A N/AMoscow Municipal Bank 1,525 848 1,348 85 1 N/ABahkir Republic Investment Bank 610 330 485 112 1 N/A

Slovak Republic Vseobecna Uverova Banka 3,887 1,873 2,722 278 67 N/A

Investicna a Rozvojva Banka 509 365 472 23 8 1,073First Building Savings Bank-Prva Stavebna Sporitelna 834 463 643 51 16 449Slovak Guarantee and Development Bank 140 2 21 103 6 93Banka Slovakia 81 19 64 15 0 N/A

Slovenia Nova Ljubljanska Banka 5,051 2,633 4,289 429 52 4,271Nova Kreditna Banka Maribor 1,563 674 1,344 158 28 N/APosta Banka Slovenija 259 108 223 11 0 209SKB Banka DD 1,353 742 1,142 121 2 N/ASlovene Export Corporation 182 2 31 73 1 N/ASlovenska Investijska Banka 133 82 101 10 0 N/A

Tajikistan Savings Bank (Sberbank) 9 3 8 0 0 N/ATurkmenistan Vneshekonombank 2,075 1,803 434 21 3 341

Data unavailable for Sberbank, Turkmenistanbank, Turkmenbashibank, and Daykhanbank

UzbekistanUzbek State Joint Stock Housing Savings Bank 142 68 101 36 2 N/AAsaka Bank 247 149 108 138 2 N/AUzbek Joint Stock-Commercial Industrial Construction Bank 435 245 338 55 2 N/ANational Bank for Foreign Economy Activity of Republic of Uzbekistan 3,913 2,227 2,071 662 2 N/A

Ukraine Savings Bank 390 98 333 28 -22 N/AExport-Import Bank 400 224 245 32 10 2,239

Yugoslavia Jugobanka Bor 241 133 103 7 0 N/ABeobanka Belgrade 489 137 668 -263 -500 4,124Invest Banka 1,541 919 416 68 -181 N/AJugobanka Beograd 1,826 1,392 171 95 N/A N/ABeogradska Banka 2,018 1,804 322 209 N/A N/AVojvodjanska Banka 560 324 176 93 0 3,215

Sources: IMF; Bank Scope; Bulgarian National Bank; authors’ calculations

130

ANNEX 2: Financial Ratios of State Banks: 1999-2000 (in percent)

Names of Public Banks

Loan Loss Reserve/Gross Loans

Equity/Total Assets

Net Interest Margin ROA ROE

Net Loans/ Total Assets

Liquid Assets/ST Funding

Albania Savings Bank 88.3 2.7 3.6 2.2 N/A 0.8 100.7Armenia Armenian Savings Bank N/A 0.0 0.3 -0.0 -1.2 0.3 0.1

AzerbaijanInternational Bank of Azerbaijan 12.5 3.0 3.7 2.1 62.1 29.9 80.2United Universal Bank

Belarus Belpromstroibank 11.4 14.8 18.8 2.2 18.5 50.0 39.3Belagroprombank N/A 39.8 22.3 2.5 6.0 77.8 11.4Belbusinessbank N/A 24.9 24.1 2.9 14.4 46.9 43.8Belgazprombank N/A 31.5 13.7 0.8 2.4 14.2 102.6Belarusbank Savings Bank N/A 13.9 11.0 -0.5 -3.1 74.4 19.1Belvnesheconombank 17.6 9.3 11.0 0.1 1.4 36.4 48.4

Bosnia-Herzegovina Federation Investment Bank 45.3 94.5 4.5 2.5 2.6 39.6 N/A

Banjalucka Banka 22.0 24.1 2.0 0.5 7.1 44.2 38.7Privredna Banka (PBS) Sarajevo 45.8 14.8 9.9 0.5 2.9 25.7 45.7Central Profit 5.7 17.6 5.8 0.1 0.3 29.2 66.4Gospodarska Mostar 15.1 13.1 6.1 1.7 11.9 31.7 81.7Data unavailable for PBS Srpska Sarajevo, PBS Doboj, PBS Prijedor, PBS Gradiska, PBS Brcko, UNA Bihac, Sipad, Postanska, Ljubljanska, Semberska, Postanska sed., Kristal, Rosvojna and Agroprom

Bulgaria DSK Bank 5.5 13.4 8.4 1.4 10.2 46.2 26.3Commercial Bank Biochim 32.7 9.2 7.5 2.2 29.1 29.0 68.9Central Cooperative N/A N/A N/A N/A N/A N/A N/A

Croatia Dubrovacka Banka 20.6 3.6 1.5 -2.8 -59.5 38.2 25.1Croatia Banka 43.4 9.6 3.3 -6.0 -53.7 51.4 39.9Splitska Banka 10.4 6.8 4.1 0.6 9.7 46.4 26.8Croatian Bank for Reconstruction & Development 16.3 57.3 5.0 1.7 2.8 43.0 63.4Hrvatska Postanska Bank 11.7 12.6 2.6 -9.2 -55.0 51.4 40.5Riadria Banka N/A 15.3 6.0 -3.1 -17.7 37.6 16.5

Czech Republic Komercni Banka 14.0 5.2 3.7 -0.1 -1.1 31.1 44.9Ceska Exportni Banka N/A 8.0 5.2 0.3 3.0 11.2 8.7Ceskomoravka Zarucni a Rozvojavo Banka 5.7 8.2 0.5 1.8 17.3 60.4 9.6

Estonia No state banks leftGeorgia No state banks left

HungaryPostbank and Savings Bank Corp 5.4 12.4 5.1 0.3 2.2 33.5 43.2Hungarian Development Bank 5.2 48.3 1.8 -3.0 -7.7 26.9 0.0

Kazakhstan Halyk Savings Bank 4.1 6.7 6.0 -0.3 -3.7 48.3 44.8Eximbank N/A 45.4 7.6 -6.9 -14.5 69.9 14.9

Kyrgyz Republic Kairat Bank 2.0 52.0 N/A -11.5 -86.0 0.0 25.0Savings and Settlement Company 0.0 19.0 N/A 0.9 4.4 0.2 115.0

131

ANNEX 2: Financial Ratios of State Banks: 1999-2000 (in percent)

Names of Public Banks

Loan Loss Reserve/Gross Loans

Equity/Total Assets

Net Interest Margin ROA ROE

Net Loans/ Total Assets

Liquid Assets/ST Funding

Energo Bank 11.1 12.0 1.6 0.2 1.5 36.0 64.0

LatviaMortgage and Land Bank of Latvia 2.6 10.9 7.6 1.1 8.9 70.7 12.5Latvian Savings Bank 4.5 3.3 5.6 0.6 17.8 25.4 8.3

Lithuania Lithuanian Savings Bank 2.1 6.7 5.2 -1.0 -13.6 28.6 28.7Agricultural Bank of Lithuania 2.8 7.7 5.3 0.5 5.9 48.2 22.6

FYR MacedoniaMacedonian Bank for Development Promotion N/A 100.0 6.7 3.2 3.2 N/A N/A

Moldova Banca de Economii 5.5 12.2 N/A 11.5 132.8 36.2 49.7

Poland

Powszechna Kasa Oszczednosci Bank Polski SA 4.8 3.6 5.7 1.0 29.6 41.3 5.4Bank Gospodarki Zywnosciowej n/a 4.8 5.2 0.6 12.0 46.1 6.9

Romania Banca Comerciala Romana N/A 19.5 10.3 3.6 20.1 27.2 47.4Savings Bank (CEC) 0.8 16.5 13.6 2.7 17.6 5.8 28.1EXIM Bank N/A 13.7 8.0 -0.9 -7.0 14.2 99.1Banca Agricola 13.7 -12.4 -9.3 22.2 N/A 12.6 41.9

Russian Federation Sberbank 12.0 7.6 9.5 2.2 29.9 44.0 48.3

Vneshtorgbank 18.2 36.2 6.0 4.7 15.0 21.8 115.1Vneschekonombank 10.2 4.6 6.3 0.4 8.8 10.5 87.2Russian Development Bank N/A 78.2 17.8 0.5 0.7 N/A 450.0Rosselkhozbank N/A N/A N/A N/A N/A N/A N/AMoscow Municipal Bank 2.3 5.6 0.3 1.0 16.1 55.6 40.6Bahkir Republic Investment Bank 12.3 18.3 13.5 10.3 78.7 54.0 37.0

Slovak Republic Vseobecna Uverova Banka 13.9 8.6 2.9 2.2 27.2 58.0 32.9Investicna a Rozvojva BankaN/A 4.4 0.9 1.6 43.5 71.7 18.7First Building Savings Bank-Prva Stavebna Sporitelna N/A 6.1 1.5 2.0 35.2 55.4 12.9Slovak Guarantee and Development Bank N/A 73.7 8.6 5.0 6.6 1.3 522.9Banka Slovakia 7.2 19.6 3.2 0.7 3.3 23.4 72.8

Slovenia Nova Ljubljanska Banka 6.8 8.5 4.7 1.1 13.1 52.1 13.1Nova Kreditna Banka Maribor N/A 10.1 6.8 1.9 19.4 43.1 14.0Posta Banka Slovenija N/A 4.4 4.6 0.1 2.0 41.5 18.1SKB Banka DD N/A 9.0 4.6 0.2 1.7 54.8 18.4Slovene Export Corporation 36.7 40.3 2.2 0.4 0.9 1.0 479.4Slovenska Investijska Banka N/A 7.6 2.1 0.3 4.0 61.4 14.4

Tajikistan Sberbank N/A 0.5 N/A 0.5 95.0 37.0 20.0Turkmenistan Vneshekonombank 2.3 1.0 0.6 0.2 15.5 86.9 48.8

Data unavailable for Sberbank, Turkmenistanbank, Turkmenbashibank, and DaykhanbankUkraine Savings Bank N/A 7.1 N/A N/A N/A 25.2 N/A

Export-Import Bank 27.3 8.1 7.9 2.6 34.6 55.9 N/AUzbekistan Uzbek State Joint Stock 6.0 25.6 19.5 2.1 8.8 47.7 42.3

132

ANNEX 2: Financial Ratios of State Banks: 1999-2000 (in percent)

Names of Public Banks

Loan Loss Reserve/Gross Loans

Equity/Total Assets

Net Interest Margin ROA ROE

Net Loans/ Total Assets

Liquid Assets/ST Funding

Housing Savings BankAsaka Bank 4.8 55.8 7.6 33.4 69.7 60.4 63.7Uzbek Joint Stock- Commercial Industrial Construction Bank 3.9 12.8 4.7 1.8 13.9 56.4 36.7National Bank for Foreign Economy Activity of Republic of Uzbekistan 1.8 16.9 2.8 0.7 4.3 56.9 62.9

Yugoslavia Jugobanka Bor 3.1 2.8 N/A 0.0 0.3 54.7 N/ABeobanka Belgrade 49.2 -80.9 N/A N/A N/A 28.0 N/AInvest Banka 15.2 -6.3 N/A N/A N/A 59.6 N/AJugobanka Beograd 1.6 3.9 N/A N/A N/A 76.2 N/ABeogradska Banka 5.4 5.3 N/A 0.1 0.9 89.4 N/AVojvodjanska Banka 0.0 16.6 N/A N/A N/A 57.9 N/A

Sources: Bank Scope; Bulgarian National Bank; authors’ calculations

133

ANNEX 3: Macro-Financial Ratios of State Banks: 1999-2000 (in %)Names of Public Banks Assets/GDP Loans/GDP Deposits/GDP Capital/GDP

Albania Savings Bank 33.2 0.3 31.8 0.9Armenia Armenian Savings Bank 0.5 0.2 0.5 0.0

AzerbaijanInternational Bank of Azerbaijan 11.7 3.5 9.0 0.4United Universal Bank 0.7 0.0 0.2 0.1

Belarus Belpromstroibank 1.0 0.5 0.8 0.1Belagroprombank 0.9 0.7 0.5 0.4Belbusinessbank 0.5 0.3 0.4 0.0Belgazprombank 0.1 0.0 0.1 0.0Belarusbank Savings Bank 2.6 1.9 2.1 0.4Belvnesheconombank 0.7 0.2 0.6 0.1

Bosnia-Herzegovina

Federation Investment Bank 1.4 0.6 N/A 1.4Banjalucka Banka 1.0 0.4 0.6 0.2Privredna Banka (PBS) Sarajevo 0.8 0.2 0.6 0.1Central Profit 3.8 1.1 2.7 0.7Gospodarska Mostar 0.6 0.2 0.4 0.1Data unavailable for PBS Srpska Sarajevo, PBS Doboj, PBS Prijedor, PBS Gradiska, PBS Brcko, UNA Bihac, Sipad, Postanska, Ljubljanska, Semberska, Postanska sed., Kristal, Rosvojna and Agroprom

Bulgaria DSK Bank 4.9 2.3 4.1 0.1Commercial Bank Biochim 2.1 0.6 1.9 0.0Central Cooperative 0.8 0.4 0.6 0.1

Croatia Dubrovacka Banka 1.8 0.7 1.0 0.1Croatia Banka 0.7 0.4 0.5 0.1Splitska Banka 4.4 2.1 3.9 0.3Croatian Bank for Reconstruction & Development 3.1 1.3 1.0 1.8Hrvatska Postanska Bank 1.0 0.5 0.7 0.1Riadria Banka 0.8 0.3 0.6 0.1

Czech Republic Komercni Banka 22.0 6.0 18.0 2.0Ceska Exportni Banka 1.3 0.0 0.3 0.1Ceskomoravka Zarucni a Rozvojavo Banka 2.3 1.4 1.5 0.2

Estonia No state banks leftGeorgia No state banks left

HungaryPostbank and Savings Bank Corp 2.6 0.9 2.2 0.3Hungarian Development Bank 1.6 0.4 0.7 0.8

Kazakhstan Halyk Savings Bank 3.9 1.9 3.4 0.3Eximbank 0.4 N/A N/A N/A

Kyrgyz Republic Kairat Bank 0.5 0.0 0.4 0.1

Savings and Settlement Company 0.4 0.0 0.2 0.1Energo Bank 0.4 0.2 0.4 0.1

LatviaMortgage and Land Bank of Latvia 1.7 1.2 0.8 0.2

134

ANNEX 3: Macro-Financial Ratios of State Banks: 1999-2000 (in %)Names of Public Banks Assets/GDP Loans/GDP Deposits/GDP Capital/GDP Latvian Savings Bank 3.5 0.9 3.1 0.1

Lithuania Lithuanian Savings Bank 7.4 2.1 6.3 0.5Agricultural Bank of Lithuania 3.7 1.8 2.9 0.3

FYR Macedonia

Macedonian Bank for Development Promotion N/A 0.4 N/A 0.0

Moldova Banca de Economii 2.7 1.0 1.9 0.3

Poland

Powszechna Kasa Oszczednosci Bank Polski SA 10.3 4.2 9.3 0.4Bank Gospodarki Zywnosciowej 2.7 1.3 2.3 0.1

RomaniaBanca Comerciala Romana 8.7 2.4 6.5 1.7Savings Bank (CEC) 2.6 0.2 2.1 0.4EXIM Bank 0.6 0.1 0.4 0.1Banca Agricola N/A N/A N/A 7.0

Russian Federation Sberbank 8.0 3.6 7.2 0.4

Vneshtorgbank 1.8 0.4 1.1 0.6Vneschekonombank 1.0 0.1 1.0 0.0Russian Development Bank 0.1 N/A 0.0 0.1Rosselkhozbank N/A N/A N/A N/AMoscow Municipal Bank 0.6 0.3 0.5 0.0Bahkir Republic Investment Bank 0.2 0.1 0.2 0.0

Slovak Republic

Vseobecna Uverova Banka 20.5 9.9 14.3 1.5Investicna a Rozvojva Banka 2.7 1.9 2.5 0.1First Building Savings Bank-Prva Stavebna Sporitelna 4.4 2.4 3.4 0.3Slovak Guarantee and Development Bank 0.7 0.0 0.1 0.5Banka Slovakia 0.4 0.1 0.3 0.1

Slovenia Nova Ljubljanska Banka 28.1 14.6 23.8 2.4Nova Kreditna Banka Maribor 8.7 3.7 7.5 0.9Posta Banka Slovenija 1.4 0.6 1.2 0.1SKB Banka DD 7.5 4.1 6.3 0.7Slovene Export Corporation 1.0 0.0 0.2 0.4Slovenska Investijska Banka 0.7 0.5 0.6 0.1

Tajikistan Sberbank 0.9 0.3 0.8 0.0Turkmenistan Vneshekonombank 47.2 41.0 9.9 0.5

Data unavailable for Sberbank, Turkmenistanbank, Turkmenbashibank, and DaykhanbankUkraine Savings Bank 1.2 0.3 1.0 1.2

Export-Import Bank 1.2 0.7 0.8 1.2

UzbekistanUzbek State Joint Stock Housing Savings Bank 1.1 0.5 0.7 0.3

135

ANNEX 3: Macro-Financial Ratios of State Banks: 1999-2000 (in %)Names of Public Banks Assets/GDP Loans/GDP Deposits/GDP Capital/GDP Asaka Bank 1.8 1.1 0.8 1.0Uzbek Joint Stock- Commercial Industrial Construction Bank 3.2 1.8 2.5 0.4National Bank for Foreign Economy Activity of Republic of Uzbekistan 29.0 16.5 15.3 4.9

Yugoslavia Jugobanka Bor 2.7 1.5 1.1 0.1Beobanka Belgrade 5.4 1.5 7.4 -2.9Invest Banka 17.1 10.2 4.6 0.8Jugobanka Beograd 20.3 15.5 1.9 1.1Beogradska Banka 22.4 20.0 3.6 2.3Vojvodjanska Banka 6.2 3.6 2.0 1.0

Sources: IMF; Bank Scope; Bulgarian National Bank; authors’ calculations

136

ANNEX 4: Market Share Ratios of State Banks: 1999-2000 (in %)Names of Public Banks Asset Share Loan Share Deposit Share Capital Share

Albania Savings Bank 63.6 5.8 71.8 13.6Armenia Armenian Savings Bank 4.1 1.4 5.1 0.4

AzerbaijanInternational Bank of Azerbaijan 60.9 33.8 87.0 11.9United Universal Bank 3.8 0.1 2.0 3.1

Belarus Belpromstroibank 16.0 10.6 20.7 12.0Belagroprombank 15.1 15.6 13.2 30.9Belbusinessbank 8.0 5.6 10.1 3.0Belgazprombank 1.9 0.4 1.9 3.0Belarusbank Savings Bank 41.7 41.2 53.8 29.5Belvnesheconombank 10.7 5.2 15.5 5.2

Bosnia-Herzegovina

Federation Investment Bank 2.2 1.2 N/A 11.3Banjalucka Banka 1.5 0.9 7.1 1.9Privredna Banka (PBS) Sarajevo 1.2 0.4 6.3 1.0Central Profit 5.8 2.3 30.3 5.6Gospodarska Mostar 0.9 0.4 5.0 0.6Data unavailable for PBS Srpska Sarajevo, PBS Doboj, PBS Prijedor, PBS Gradiska, PBS Brcko, UNA Bihac, Sipad, Postanska, Ljubljanska, Semberska, Postanska sed., Kristal, Rosvojna and Agroprom

Bulgaria DSK Bank 12.2 10.7 17.7 15.7Commercial Bank Biochim 5.2 2.9 8.0 4.6Central Cooperative 0.9 6.6 2.0 1.8

Croatia Dubrovacka Banka 3.0 1.6 2.9 0.5Croatia Banka 1.2 0.8 1.3 0.5Splitska Banka 7.4 4.7 10.9 2.2Croatian Bank for Reconstruction & Development 5.1 3.1 2.8 13.0Hrvatska Postanska Bank 1.6 1.2 2.0 0.9Riadria Banka 1.3 0.7 1.6 0.8

Czech Republic Komercni Banka 20.8 9.6 26.8 9.3Ceska Exportni Banka 1.2 0.1 0.5 0.5Ceskomoravka Zarucni a Rozvojavo Banka 2.1 2.2 2.3 0.9

Estonia No state banks leftGeorgia No state banks left

HungaryPostbank and Savings Bank Corp 4.4 1.8 5.5 5.1Hungarian Development Bank 2.7 0.9 1.8 12.4

Kazakhstan Halyk Savings Bank 19.3 17.2 29.0 7.0Eximbank 3.0 N/A N/A 3.9

Kyrgyz Republic Kairat Bank 7.1 0.0 7.4 3.6

Savings and Settlement Company 5.7 0.0 4.4 4.0Energo Bank 6.1 3.3 7.4 2.8

LatviaMortgage and Land Bank of Latvia 2.8 4.8 1.9 3.9

137

ANNEX 4: Market Share Ratios of State Banks: 1999-2000 (in %)Names of Public Banks Asset Share Loan Share Deposit Share Capital Share Latvian Savings Bank 5.0 3.0 7.0 2.0

Lithuania Lithuanian Savings Bank 27.4 11.8 36.5 7.5Agricultural Bank of Lithuania 13.8 10.0 16.5 4.4

FYR Macedonia

Macedonian Bank for Development Promotion N/A 1.9 N/A 2.9

Moldova Banca de Economii 9.3 7.0 18.9 3.8

Poland

Powszechna Kasa Oszczednosci Bank Polski SA 18.9 9.4 27.8 7.2Bank Gospodarki Zywnosciowej 5.0 2.8 6.9 2.6

RomaniaBanca Comerciala Romana 33.4 29.1 31.6 53.5Savings Bank (CEC) 10.6 2.2 12.7 10.9EXIM Bank 2.4 1.1 0.6 2.8Banca Agricola 3.6 2.6 3.2 2.5

Russian Federation Sberbank 24.7 16.9 45.0 6.4

Vneshtorgbank 5.4 1.8 6.8 10.3Vneschekonombank 3.2 0.5 6.0 0.8Russian Development Bank 0.2 N/A 0.2 0.3Rosselkhozbank N/A N/A N/A N/AMoscow Municipal Bank 0.0 0.0 0.0 0.0Bahkir Republic Investment Bank 1.9 1.6 8.7 0.2

Slovak Republic

Vseobecna Uverova Banka 23.6 14.1 24.2 12.7Investicna a Rozvojva Banka 3.1 2.7 4.2 1.0First Building Savings Bank-Prva Stavebna Sporitelna 5.1 3.5 5.7 2.3Slovak Guarantee and Development Bank 0.8 0.0 0.2 4.7Banka Slovakia 0.5 0.1 0.6 0.7

Slovenia Nova Ljubljanska Banka 38.1 24.2 52.3 26.8Nova Kreditna Banka Maribor 11.8 6.2 16.4 9.9Posta Banka Slovenija 2.0 1.0 2.7 0.7SKB Banka DD 10.2 6.8 13.9 7.6Slovene Export Corporation 1.4 0.0 0.4 4.6Slovenska Investijska Banka 1.0 0.8 1.2 0.6

Tajikistan Sberbank N/A N/A N/A N/ATurkmenistan Vneshekonombank N/A N/A N/A N/A

Data unavailable for Sberbank, Turkmenistanbank, Turkmenbashibank, and DaykhanbankUkraine Savings Bank 6.7 2.9 9.8 4.4

Export-Import Bank 6.9 6.7 7.2 5.0

UzbekistanUzbek State Joint Stock Housing Savings Bank

N/A N/A N/A N/A

138

ANNEX 4: Market Share Ratios of State Banks: 1999-2000 (in %)Names of Public Banks Asset Share Loan Share Deposit Share Capital Share Asaka Bank N/A N/A N/A N/AUzbek Joint Stock- Commercial Industrial Construction Bank

N/A N/A N/A N/A

National Bank for Foreign Economy Activity of Republic of Uzbekistan

N/A N/A N/A N/A

Yugoslavia Jugobanka Bor N/A N/A N/A N/ABeobanka Belgrade N/A N/A N/A N/AInvest Banka N/A N/A N/A N/AJugobanka Beograd N/A N/A N/A N/ABeogradska Banka N/A N/A N/A N/AVojvodjanska Banka N/A N/A N/A N/A

Sources: IMF; Bank Scope; Bulgarian National Bank; authors’ calculations

139

ANNEX 5: Different Arrears in Selected Transition Countries as a Percentage of GDP (in percent)1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

ARMENIA Wage Arrears 0.9 1.3 1.5 1.6 Industry 0.5 0.6 0.5 0.6 Agriculture 0.1 0.1 0.1 0.1 Transportation 0.0 0.0 0.0 0.0 Construction 0.1 0.2 0.2 0.2 Trade, Material, Supply, Procurement 0.0 0.0 0.0 0.0 Education and Science 0.1 0.1 0.1 0.1 Credit and Insurance 0.0 0.0 0.0 0.0 General Administration 0.0 0.1 0.1 0.1 Health 0.1 0.2 0.4 0.4 Other 0.0 0.0 0.0 0.0

AZERBAIJAN Total 60.6 100.2 68.30 96.80 148.20 166.10 200.00 215.60BELARUS Total 30.4 13.5 18.20 13.30 23.10 19.20 22.39 19.08BULGARIA Total 68.8 60.6 47.5 41.7 66.3 27.9 24.1 19.9

Arrears to Banks 11.0 12.2 4.2 4.6 7.2 1.7 2.3 1.0 Arrears to Suppliers 20.2 15.2 13.8 11.6 23.2 9.2 7.7 7.0 Arrears to Workers 2.9 3.1 2.4 1.6 2.3 1.1 0.9 0.9 Arrears to Government 7.8 7.5 9.5 8.5 10.4 5.5 6.1 4.3 Arrears to Pensions 2.5 2.6 2.1 2.1 2.1 0.7 1.0 1.2 Other Arrears 24.4 19.9 15.3 13.3 21.1 9.7 6.2 5.6

CROATIA Total 3.4 6.2 7.4 8.1 11.4 20.1 14.4 11.6KAZAKHSTAN Arrears to Workers 1.9 3.0KYRGYZ Total 7.3 6.3LITHUANIA Total 9.3 9.0

Tax Arrears 1.1 1.1 0.8 0.7 Energy Arrears 0.3 0.6 0.3 0.2 0.1 Banking Arrears 4.4 5.1 Inter-enterprise Arrears 3.9 3.1

MACEDONIA Arrears to Workers 5.9 5.7 5.5 Arrears to Government 16.0

MOLDOVA Wage Arrears 4.6 4.1 7.0 4.5 2.8 2.0 Agriculture 2.1 2.1 2.4 1.1 0.6 0.5 Manufacturing 0.5 0.5 0.7 0.4 0.4 0.3 Construction 0.2 0.2 0.3 0.2 0.2 0.1 Transport 0.2 0.2 0.3 0.3 0.3 0.2 Real Estate 0.1 0.1 0.2 0.2 0.1 0.1 State Administration 0.2 0.3 0.9 0.6 0.3 0.3 Education 0.7 0.3 1.0 0.7 0.3 0.2 Health Care and Social Assistance 0.4 0.2 0.6 0.5 0.2 0.2

ROMANIA Total Arrears 34.60 23.90 26.10 25.15 36.07 33.74 36.15 42.22 Arrears to Suppliers 22.03 15.05 13.88 13.35 16.05 11.92 15.22 18.02 Arrears to Others 4.51 3.64 3.25 3.57 6.90 6.21 6.78 9.46 Arrears to Bank 4.03 1.42 2.07 3.12 6.22 5.81 6.06 6.44 Arrears to Government 0.00 2.37 4.83 5.11 6.89 6.62 8.08 8.29

RUSSIA Total Arrears 9.5 14.8 13.3 23.4 29.1 47.8 30.3 23.7 Arrears to Suppliers 6.5 10.7 12.8 21.4 13.0 10.1 Tax Arrears 3.1 4.6 6.0 8.3 5.8 4.9 Arrears to off-budget funds 0.9 4.2 5.7 9.0 6.3 4.6

140

Overdue Bank Credit 2.8 3.9 4.6 9.1 5.2 4.1

Arrears to Workers 0.2 0.4 0.7 0.8 1.6 1.7 1.9 0.7 0.4 Industry 0.1 0.2 0.4 0.5 1.0 1.1 1.2 0.4 0.2 Agriculture 0.0 0.2 0.2 0.2 0.3 0.3 0.3 0.2 0.1 Construction 0.0 0.1 0.1 0.1 0.3 0.3 0.4 0.1 0.1

UKRAINE Total Arrears 7.95 6.00 13.00 20.00 24.00 85.00 98.00 Arrears to Workers 5.0 6.0 Other Overdue Payments 6.00 13.00 20.00 24.00 80.00 92.00

Wage Arrears 1.00 5.1 5.5 6.4 5.0 2.8 2.2Note: Enterprise arrears to government may not equal tax arrears since tax arrears include household arrears and enterprise arrears to government can include other forms of arrears.

141

ANNEX 6: ROMANIA: THE CLOSING OF BANCOREX170

The story of Bancorex highlights the failure of an institution that long served as the key financial intermediary to implement the government’s poor macroeconomic policies and sustained support for loss-making state enterprises. Such operations ultimately led to the point where its losses were unsustainable, leading to the bank’s liquidation in 1999.

Bancorex (BX), the former foreign trade bank, was the largest and most troubled of the four fully state-owned banks in Romania prior to its closure in 1999. Accounting for about one-fourth of total banking sector assets in the early to mid-1990s, BX financed a significant portion of Romania's energy import requirements, as well as imports of capital goods. BX was also used by the authorities as a major vehicle to subsidize the state-owned energy sector and energy-intensive industrial sector, also primarily controlled by the state. Most of BX’s clients had very poor prospects for repaying loans, not just because of their inefficiency and poor management practices, but also due to external constraints (particularly price controls) that they faced. The legacies of subsidized loans, years of mismanagement, and behind-the-scenes political dealings rendered BX the most troubled bank in the wake of exchange rate and price liberalization in early 1997. As BX was greatly exposed to debtors who traditionally relied on directed credit and the highly subsidized exchange rate, the termination of National Bank of Romania's (NBR) directed credit and exchange rate liberalization in 1997 made it unequivocally evident that the bank was insolvent.

At the end of 1997, BX received the equivalent of $600 million in government bonds (2 percent of GDP) to restructure its nonperforming loans in the portfolio. However, the restructuring of BX, which was to accompany the recapitalization, never took place. Although a new management team was appointed in April 1998 and other steps were taken, a comprehensive restructuring plan was never implemented and the bank's situation deteriorated further. When BX was again in crisis in late 1998, the authorities considered restructuring measures with a view to privatizing the bank, although international experience would have favored liquidation. The authorities were concerned about the systemic risk of liquidation, and proposed an up-front recapitalization, followed by restructuring and privatization. However, as the depth of the bank's problems became more fully known, it was clear by early 1999 that BX was in much worse shape than expected, and that privatization after recapitalization would be prohibitively costly. A recapitalization would have required up to $2 billion from the budget, or almost 6 percent of GDP.

An estimate at the end of February 1999 put the nonperforming loans of BX at about 85-90 percent of its loan portfolio, or $l.7 billion (5 percent of GDP; this number increased as more became known about BX during the process of closing the bank), with most of the portfolio being in foreign currency. At that time, Bancorex accounted for one fourth of total banking

170 Alex Pankov is the author of this annex. Sources used include (i) C. James, Banking and Financial Sectors in East and Central Europe. Financial Times Management Reports, 1993; (ii) BankScope, Bancorex Report, 1998; (iii) International Monetary Fund, Romania: Selected Issues and Statistical Appendix, 2001; and (iv) internal World Bank and IMF Reports.

142

system assets and 47 percent of all foreign currency loans. Finally, in April 1999, BX collapsed in a liquidity squeeze as depositors lined up to withdraw their money. It became clear that a rapid liquidation was the only solution which would avoid further runs on the bank and a systemic crisis during a fragile external economic period. Realizing the magnitude of BX's problem, in April 1999, the authorities finalized a liquidation plan aimed at the orderly removal of BX from the banking system

The final resolution of BX was completed in the following manner:

Following the appointment of a special administrator to replace BX's management (February 1999), all bad assets at the end of 1998 were transferred to the newly established Asset Recovery Agency (AVAB) for loan workout and debt recovery by July 31, 1999.

Some of the deposit liabilities and most foreign debt liabilities were transferred to another state-owned bank, Romanian Commercial Bank (BCR), while a large part of the deposits were withdrawn before July 31, 1999, owing to delays in transfers. The NBR provided special credit to staunch the financial hemorrhage of the bank. Both BCR and the NBR were compensated by government securities in corresponding currencies.

The remainder of BX was merged with BCR, which absorbed the balance sheet of BX, as the authorities considered an actual liquidation politically unacceptable and too lengthy to complete. BCR received government securities to compensate for the hole in BX's balance sheet, and was given the right of first refusal on any BX assets transferred (on and off the balance sheet).

The government approved the withdrawal of BX’s banking license on July 31, 1999 (effective August 2).

The final absorption of BX by BCR was completed in September of 1999, while BCR's exercise of its right of first refusal of the BX assets, which were transferred to AVAB in exchange for government securities, continued well into 2000. The Ministry of Finance also agreed to guarantee BX's more than $400 million in off-balance sheet items transferred to BCR.

The closure of BX removed a large destabilizing factor in the Romanian financial system, albeit at a very heavy cost to the taxpayers, and removed some $2 billion in non-performing assets from the banking system, which helped to improve the general soundness of the system. In this process, the government took on public debt amounting to $1.5 billion (net of provisions and other assets) or 4.5 percent of GDP in 1999. To this cost should be added the 1997 recapitalization of $500 million; the assumption by the government of off-balance sheet items; and legal liabilities (the exact amount of which is still unknown)

It should be noted that the heavy fiscal costs incurred in connection with the liquidation and closure of BX are mostly the realization of losses which were actually incurred before 1997, caused both by the use of BX as a vehicle for quasi-fiscal payments and by the mismanagement of the bank. Estimates indicate that nonperforming loans before the recapitalization of 1997 amounted to about $1.5 billion, and much of the off-balance-sheet and legal liabilities had been

143

incurred before then as well. It should also be emphasized that delays in the overall process may have increased the cost to the taxpayers by as much as several hundred million dollars.

144

ANNEX 7: UKRAINE: THE LIQUIDATION OF BANK UKRAINA171

Bank Ukraina (BU) was one of the four state-owned specialized banks that were spun off from the Soviet All-Union banks as Ukraine gained its independence from in 1991. Throughout the 1990s, BU remained one of the largest banks in Ukraine, employing (in October 2000) 16,600 people in a large network of 512 branches. The bank mainly concentrated its lending activities in the agricultural sector, while the other state banks specialized in industrial lending (Prominvestbank), social programs (Ukrsotsbank), and household savings (Oschadny Bank; see Box 5.2, above). A fifth state bank, Ukreximbank, was formed in 1992 to process Ukraine’s foreign trade payments.

All specialized banks but Oschadny and Ukreximbank were corporatized (and thus nominally privatized) in the course of 1992, primarily via “ownership transformation” whereby a number of large state-owned enterprises took substantial ownership shares in the banks that serviced their sectors. During the ensuing years, the ownership structure of BU and other similarly corporatized banks became even more complicated. This change was largely the result of a 1993 order by the Government that all state enterprise shares in these banks should be transferred to the Ministry of Finance. This prompted the banks to devise a method of transferring ownership through the distribution of shares to the employees of client enterprises, and of the banks themselves. Thus, ownership of BU and the other two former state banks became divided among tens of thousands of shareholders, most of them individuals.

Not surprisingly, decision-making within these banks was not controlled in a meaningful way by the shareholders. Most major policy and personnel decisions were still made by top managers of the state enterprises that were majority shareholders prior to the share redistribution. In the absence of a major outside entity owning a controlling stake, this meant that the state continued to exercise considerable indirect and informal influence in those three banks’ affairs. In the case of BU, the government also exerted direct influence on the decision-making process by retaining a residual state shareholding of at least 13 percent until 1998, when it was finally sold for cash (it was later alleged that the package was undervalued). Specifically, the government seemed to manage the institution through the Ministry of Agriculture, Ministry of Finance, and the National Bank of Ukraine (NBU). Additionally, regional authorities appeared to exercise some control over regional offices. This complicated governance structure was harmful to the financial position of the bank, and effectively turned the bank into a series of regional banks operating under the same name.

Personal links of BU with the government appeared to be, at the beginning of transition, a good source of financing and profit. Major decisions on channeling budget funds, financing individual projects, or attracting more lucrative enterprises as bank clients were made in agreement with, by recommendation of or under pressure from government authorities. However, despite a number

171 Alex Pankov is the primary author of this annex. Sources include (i) World Bank mission reports; (ii) A. Roe, et al. “Ukraine: The Financial Sector and the Economy,” World Bank Report, 2001; (iii) Intellinews reports; and (iv) Ukrainian News Agency.

145

of benefits that BU received from the Government during these years (e.g., free access to budget funds, state procurement contracts, government guarantees for trade finance deals, etc.), by the late 1990s, the BU ended up with a large proportion of bad assets on its balance sheet. This resulted from poor management quality, unmanageable liabilities, and serious external interference with BU’s daily banking operations and strategic decisions. To make things worse, the bank’s dire financial position was long obscured by inaccurate loan classification practices and low levels of loan provisioning that resulted in a serious overestimation of assets and capital.

The continuation of government-directed loans to non-viable state-owned enterprises that were approved under some form of “instruction” by local governments, ministerial resolutions, or presidential decrees proved to be particularly harmful to BU’s financial health. This was because most of the directed loans were never intended to be repaid. When the bank prepared a list of government-directed loans in 2000, only UAH 75 million ($13.8 million) were formally acknowledged by the government,172 leaving UAH 433 million ($80 million) unacknowledged. The analysis of the loan portfolio also revealed highly controversial lending to shareholders of the bank, to daughter companies, and to affiliated companies. As an example, two BU shareholders—Atlanta International (UAH 31 million) and Association Interagro (UAH 92 million—appeared on the list of government-directed loans as well.

The situation of the bank deteriorated dramatically from 1998. Its share of bad loans became unsustainable even by Ukrainian standards, and BU had to rely on central bank credits be maintain its liquidity position. The IMF-led diagnostic review of the bank in the same year confirmed its deep insolvency. The state-protected bank came to be seen as having wasted the country's financial resources by subsidizing loss-making industries and the largely unreformed agricultural sector. In November 1998, after considering renationalizing the bank, the authorities finally put it into a rehabilitation program, and in June 1999, the NBU instructed BU to sign a Commitment Letter aimed at bank recovery. However, the bank failed to comply with the targets of the recovery program. Given the high level of deterioration in its loan portfolio (approximately 75 percent of which was non-performing by the end of 2000, although estimated as high as 90 percent if reclassified under IAS), negative liquidity, and capital erosion, the bank's financial condition became dangerous enough to be a potential threat to the entire banking system. This resulted in the introduction of the Provisional Administration in the bank as of September 25, 2000.

After extensive analysis of BU’s situation conducted by the World Bank in early 2001, the government came to realize that there was no alternative to immediate and orderly liquidation. Under a recast IAS balance sheet at year-end 2000, BU's capital shortfall amounted to UAH 900 million, with UAH 650 million of loan loss provisions (about 56 percent of total assets) required to cover the level of non-performing loans. The World Bank concluded that there was no prospect of the bank becoming commercially viable or self-sustaining, even if it were re-capitalized to achieve minimum capital adequacy standards, as the temporarily improved income stream would be steadily reduced by continued and rising nonperformance. The government had no plans to cover the capital shortfall, and no potential third parties were willing to lend to BU or

172 This included Ukraine’s state-controlled energy company, NAFTA K, which was on the list of the biggest borrowers as well as on the list of government-directed loans. Total exposure to this company was UAH 179 million ($33.2 million).

146

provide additional capital. There was thus no viable means by which the bank could obtain the liquidity and capital necessary to continue its operations.

The World Bank analysis demonstrated that BU’s role as a provider of credit to Ukraine’s vital agricultural sector would not be an obstacle to bank’s liquidation. First, BU allocated only $25 million per year over 1999-2000 to the agricultural sector. This annual amount of credit was negligible, relative to the size of the Ukrainian agricultural sector (which has a turnover of about $8 billion per year). Second, BU had ceased to allocate these credits already, since the appointment of a provisional administrator in late 2000. BU simply did not have resources to provide new agricultural loans.

On July 16, 2001, after three years of delays and political infighting, NBU annulled BU’s banking license and effectively launched the bank liquidation procedure. In parallel, the Prosecutor-General’s Office launched a criminal investigation against the bank’s board of directors, who were suspected of abuse of office. The liquidation procedure, prepared with assistance from the World Bank, is coordinated by the Agency on Bankruptcy Issues. The World Bank’s technical assistance included advice on the deposit compensation process, loan workouts and debt recovery, staff retrenchment and compensation, and agricultural finance reform in the wake of BU’s demise.

The issue of dealing with 1.5 million individual deposit accounts (totaling UAH 271 million, or $50.2 million) represented a special challenge to the authorities in view of the potential for panic, and the impact this could have for the banking system as whole. According to the liquidation plan, individual depositors are entitled to full compensation from the Deposit Insurance Fund of up to UAH 500 ($92.6) per account. There stands to be a substantial loss for a small number of individual household depositors exceeding this limit, as well as corporate depositors and BU’s creditors, including the NBU, who will have to wait for proceeds from the debt recovery process that is expected to last at least two to three years.

Given the highly politicized nature of BU’s resolution process, it is too early to judge the success of the bank’s long-delayed liquidation, and its final cost to the taxpayers. The latter figure is likely to be in the tens of millions of dollars, as the total outstanding NBU credit to BU alone amounts to UAH 398 ($73.7 million), a result of state-directed credits to agro-enterprises and liquidity support for the bank from 1996-99. It also remains to be seen whether the authorities will be able to draw any lessons from the BU disaster. Already, vague plans have been voiced at the highest levels of the Ukrainian government to recreate a state agricultural bank using the branch network and infrastructure left by BU.

147

ANNEX 8: UKRAINE: OSCHADNY BANK173

Oschadny Bank, the traditional state savings bank, was formed as a specialized savings bank under the initial reforms introduced after Ukrainian independence in 1991. Prior to that, Oschadny was a part of the larger Gosbank system, specializing in deposit safekeeping, pension payments, transfers, utilities payments, and other services at the retail level. In contrast to the three other remnants of the Gosbank system (Bank Ukraina, Prominvestbank, and Ukrsotsbank) that were privatized (at least nominally) in the early 1990s, Oschadny remains fully state-owned to date.

Oschadny, although a small bank by international standards, is one of the largest banks in Ukraine. As of late 2000, in addition to headquarters (in two buildings), SB had 26 full service regional offices, 450 branches (full service district offices), 7,847 outlets, and 24 agencies. Staffing was 38,015, or 35,227 on a full time-equivalent basis. In many rural locations, the bank is the only formal financial institution providing basic payment services and deposit safekeeping. In this regard, Oschadny has about 25-30 percent of total household deposits in the banking system. Thus, on a comparative basis with other banks in Ukraine, Oschadny is the leading institution in mobilizing household deposits, mainly in local currency. Nonetheless, its importance in this area is mitigated by general monetary patterns indicating that only 9.8 percent of total deposits held with banking institutions are held with Oschadny. This suggests that most enterprise and foreign currency deposits are placed with other banks (or held outside the banking system), and that its aggregate deposit holdings in the end are not so significant that a major change in Oschadny’s operations and status would in any way destabilize Ukraine’s financial markets. To the contrary, continued loss-making operations, excessively risky lending, investment, and off-balance sheet financing (via guarantees, trade finance, etc.) to enterprises, and financing to third parties through subsidiaries, affiliates or other conduits represent far more serious risks to system stability than a reorganization of Oschadny under current circumstances. The total amount of household deposits in Ukraine was only about $330 million in late 2000, small for a country of 50 million. Thus, while perceived to be “large” in Ukraine, Oschadny is not really a significant bank in terms of aggregate intermediation, reflecting the comparative insignificance of banking and financial intermediation in the economy.

Similar to other state and quasi-state banks, Oschadny’s financial situation deteriorated in the late 1990s, due to the lack of restructuring earlier in the 1990s, as well as management and operating practices common to government-controlled banks in Ukraine that have proven to be financially unsustainable. The bank is characterized by a series of weaknesses, including (i) high operating costs associated with excessive branch structures and employment, and other operating inefficiencies; (ii) government-directed lending that has resulted in a substantially impaired portfolio, reduced levels of earning assets, and after-tax losses; (iii) ongoing governance problems associated with central and regional government involvement in decision-taking, and with the wholly non-transparent interventions of key business groups connected to the bank’s management and related authorities; (iv) the inability or unwillingness of the government to

173 Alex Pankov is the primary author of this annex. Sources include (i) World Bank mission reports; (ii) A. Roe, et al. “Ukraine: The Financial Sector and the Economy,” World Bank Report, 2001; and (iii) the Ukrainian News Agency.

148

honor financial obligations to the bank in the form of payments on guarantees and capital contributions; and (v) lagging performance in the upgrading of management systems and informational technologies to lower costs, and to enforce centralized policies on the bank’s regional offices.

Oschadny experienced after-tax losses of $22 million in 2000. There is a possibility that these and other cumulative losses from before (as well as since 2000) are understated due to questionable loan classification standards and overvalued collateral. Thus, the financial condition of Oschadny may be more serious than reported. Beyond that, there is the ongoing risk that Oschadny’s condition may worsen due to its efforts to aggressively increase lending, reflecting its desire to grow out of its problems. Oschadny was appointed by the government in mid-2000 to act as the authorized bank to service clearing accounts of electricity utility companies and their branches. Oschadny performed this responsibility until October 2001, and the potential losses from these operations and incremental lending remain to be seen.

Solving the problem of Oschadny depends largely on the willingness and capacity of the Ukrainian authorities to take decisive action in terms of governance and strategy. Although strict market logic argues for the bank’s closure, its central place in Ukraine’s social fabric makes this solution unlikely in the foreseeable future. To reverse current losses, the bank’s management has pursued a cost-reduction strategy by closing some non-viable offices and releasing staff. The bank closed 148 branches and 2,933 operational offices (agencies) from January 1, 1998 to December 31, 2000. However, the financial and institutional gap is still so large that it is questionable whether Oschadny can become competitive without a major restructuring and cost reductions, coupled with large volumes of financial assistance from the government. Preliminary estimates in early 2001 indicated that Oschadny would need to reduce its costs by about 40 percent to achieve breakeven.

The key condition for stabilizing the situation requires that Oschadny not be used as a quasi-fiscal institution, or as a vehicle for directed lending as it so often has been in the past. As shown by the case of energy companies’ accounts, the bank still has a severe problem in terms of corporate governance, with a high level of politicization of management decision-making. This subjects the institution and government (as its sole shareholder) to a high level of financial risk, and points to the need to promote more professional practices at the supervisory and management board levels. At a minimum, safeguards need to be put in place to ensure that Oschadny decision-making is grounded in commercial principles. During the restructuring period, any "social" or "governmental" activity should be insulated from the bank’s balance sheet and subject to commercial pricing. Any lending or investment should be explicitly guaranteed (in documented form) so that Oschadny is utilized strictly as an agent that assumes no risk.

There is a risk that, to avoid the social consequences of closure and liquidation, the government will permit SB to seek to grow out of its problems, attempting to leapfrog from its current status as a specialized savings bank to a full-service "universal" bank. This is an extremely dubious and high risk strategy, due to the weak financial condition and limited institutional capacity of the bank. Under such circumstances, it is almost certain that Oschadny would assume more risk than is prudent, to generate high earnings and fund its accumulated losses. Adverse selection becomes a strong possibility under such circumstances, especially since the bank does not have the systems or the experience to accurately measure risk and return.

149

Even with the best of intentions, the bank’s restructuring will be costly and time-consuming in terms of management, operations and finance. It will also be complex due to the weakness of information regarding its extensive branch network, the need for a more modern personnel management system, and the determination of retrenchment levels as part of the effort to reduce the cost structure of the bank. For these reasons, Oschadny’s restructuring can be expected to take years, consume significant management time and energy, and present a major cost to Ukraine’s taxpayers.

150

ANNEX 9: CZECH REPUBLIC: CESKA SPORITELNA174

In 2000, the Czech government sold Ceska Sporitelna, the state savings bank and the country’s second largest bank, to Erste Bank of Austria. The sale brought to a close one of the government’s most lengthy and difficult bank privatizations. It also helped pave the way for the privatization the following year of the Czech Republic’s largest remaining state-owned bank, Komercni Bank. However, the sale of Ceska Sporitelna came at great expense to the government as it followed several years of consolidation and restructuring that culminated in a massive government bailout during the final year before the sale. The government strategy proved far more costly than originally expected and, most likely, than if it had pursued privatization more vigorously at an earlier stage.

First Phase of Bank Privatization

Founded in 1825, Ceska Sporitelna was the state savings bank during the socialist era and remains the largest retail bank in the Czech Republic. As of 2000, it had $12 billion in assets, a 34 percent market share in retail savings, and a network of 934 branches. As part of the Czech government’s financial sector restructuring during the mid-1990s, Ceska Sporitelna was included in the first wave of Czech privatization programs.

Recognizing that the commercial banks created from the monobank system inherited significant stocks of non-performing loans from the central planning era, the government developed a two-staged program to financially restructure and then privatize the new banks.175 A total of 37 percent of the Ceska Sporitelna’s shares were offered for vouchers and 20 percent was sold to towns and municipalities, while 40 percent was retained by the state. The government pursued this policy as well for its other three major state-owned banks, including Komercni, Investicni, and Obchodni. By the end 1995, these four banks in total accounted for 62 percent of banking system assets, and were 47-63 percent divested by vouchers, with the state-owned National Property Fund retaining the largest block. While not fully privatized, these banks were effectively “corporatized” with the expectation that full privatization would occur after additional restructuring and as accession to the EU neared.

Acceleration of the Privatization Process: Toward a Strategic Investor

Despite these measures, Ceska Sporitelna languished in this quasi-privatized status until mid-1999, when the government began to speed up its planned sale of a majority stake in the bank. In the meantime, the bank’s prospects had been hurt by poor lending decisions that resulted in an increase in non-performing loans. By mid-1999, the bank was expected to lose $389 million—the equivalent of more than half of the bank’s capital. According to reports at the time, the bank needed to cover the loss by writing off part of its capital, which would then result in its capital adequacy ratio falling below the legal minimum set by the Czech National Bank. This forced the

174 Alex Gross is the author of this annex. Sources included (i) M. Borish, W. Ding, and M. Noel, On the Road to EU Accession: Financial Sector Development in Central Europe. World Bank Discussion Paper No 345, 1996; (ii) Economist Intelligence Unit, country reports (various), 1999 and 2000; and (iii) US Department of Commerce - National Trade Data Bank, November 3, 2000.175 Borish, Michael, Wei Ding, and Michel Noel, On the Road to EU Accession: Financial Sector Development in Central Europe. World Bank Discussion Paper No 345, World Bank, Washington DC 1996.

151

state to intervene to recapitalize the bank, and increased the incentive for the government to privatize the bank.176

Although several foreign banks expressed an interest in Ceska Sporitelna, the government felt that some of their offers insufficiently valued the bank’s franchise. Rather than launching a formal tender, the government entered exclusive negotiations with Erste. In the end, although Erste raised its bid, the bank paid only approximately 1.55 times book value for the Ceska Sporitelna.177 The state also was forced to make significant concessions. In particular, it gave Erste Bank five-year guarantees on about half of Ceska Sporitelna’s loans. Prior to the sale, approximately $1.1 billion of non-performing loans were transferred to the state, which also increased the bank’s share capital by $201 million to bolster attractiveness to foreign buyers.

The state had a strong interest in privatizing Ceska Sporitelna. Most importantly, in doing so, the state rid itself of responsibility for the bank’s losses before the bank’s market position slipped further, making it less appealing to a strategic investor and bringing on further losses. It marks government divestiture of one of its last two major state-owned banks—a hurdle that to date had stood in the way of EU accession.

Ceska Sporitelna has been a loss-making bank and will also need serious restructuring of its large network of branches, as well as a strategy for dealing with more than 16,000 employees. Erste Bank took on the bank primarily because it has significant long-term interests in the region, and was keen to establish a foothold in the Czech market. It was attracted by the potential cost savings that could occur with a rationalization of operations. Both Erste and Ceska Sporitelna are retail banks offering a similar range of products, including investment funds, leasing and insurance. In the deal, Erste also committed itself to a major capital increase at Ceska Sporitelna, setting aside $571 million for housing and small business programs, and providing an additional $29 million for venture capital.178

The Ceska Sporitelna privatization offered important lessons for the government as it pursued its last major privatization with Komercni Bank the following year. Like Ceska Sporitelna, Komercni Bank suffered huge losses from non-performing loans and required two massive government bailouts. Many potential bidders in the Ceska Sporitelna case were not interested because details of the bailout were unclear. As a result, the government entered a bidding process with only one major bidder. The Komercni privatization was pursued with such lessons in mind.

176 Economist Intelligence Unit, country reports (various), 1999 and 2000177 Ibid.178 US Department of Commerce - National Trade Data Bank, November 3, 2000

152

ANNEX 10: RUSSIA’S SBERBANK179

Sberbank, the Russian state savings bank, occupies a unique place in the banking system as it is by far largest bank in the country. With more than $20 billion in assets, nearly 200,000 employees, and 21,000 branches, Sberbank has the widest network of any bank throughout the country. It has a virtual monopoly of the county’s deposits, with an estimated three quarters of retail ruble deposits in 2000 and an estimated 50 percent of foreign currency deposits in late 2001.

Recent indicators show that Sberbank controls approximately 23 percent of the banking system assets. In addition, its share of total bank loans has grown rapidly since the 1998 crisis, increasing from 12 percent in 1998 to more than 25 percent in 2000. These figures reflect the bank’s considerable market power, accounting for up to 75 percent of commercial lending in some regions. In 2000 a ranking of international banks by “The Banker” placed Sberbank as 301st in the list of the world’s largest banks.

The Role of Sberbank in the Early 1990s

During the last decade, Sberbank has continued to play the role of a traditional state savings bank, despite the numerous rapid changes and developments in the banking sector. Sberbank was created as a joint-stock company in 1991 during the government’s reform and restructuring of the banking sector. It was at this time the government broke up the two-tier system, consisting of the Gosbank (the central bank) and five specialized banks that had existed in the Soviet Union since 1987. This reform allowed for private banks to exist for the first time, and led to the creation of some 800 new banks taking the capital of the previous state banks. Sberbank was the largest among these banks, and one of the first to be “privatized,” with its major shareholder becoming the Central Bank of the Russian Federation.

Sberbank’s Role Since the 1998 Financial Crisis

The dominance of state banks in Russia is high and has been increasing since the financial crisis in 1998. Sberbank emerged from the crisis with a stronger market position than ever before. While thousands of people lost their savings with the collapse of some of Russia’s leading banks, including Inkombank and SBS-Agro, Sberbank benefited from a government retail depositor protection scheme, whereby depositors were encouraged to transfer their deposits to Sberbank. This helped strengthen Sberbank’s public image as a secure financial institution, and reinforced the perception that the bank was “too big to fail.”

The 1998 crisis also led to a shift in some of the bank’s operations. Before this time, Sberbank invested most of its assets in government securities. Since the crisis, it has aggressively expanded its lending to the corporate sector. The bank has established itself as the dominant source of loans to large Russian corporations, such as oil, gas, chemicals, construction, and trade enterprises, and to state and municipal bodies.

179 Alex Gross is the author of this annex. Sources include (i) M. Fuchs, Building Trust Developing the Russian Financial sector (draft), World Bank, 2002; (ii) A. Aslund and R. Layard, Changing the Economic System in Russia, 1993; (iii) Sberbank Annual Report, 2000; and (iv) M. Builov, “Who Owns Russia,” The Russia Journal, 2002.

153

The Future of State Banking in Russia

Sberbank’s continuing dominance reflects the unwillingness of the government to address some fundamental policy issues that affect development of the financial sector. Sberbank has several advantages, including its (i) large branch network with a broad geographic distribution; (ii) perceived state guarantee on deposits; (iii) strong internal payment system; and (iv) key role in distributing state pension payments. However, its dominant position in the sector comes with significant costs to the system as a whole, not the least of which is the detrimental impact on competition in the banking sector. The government’s involvement in Sberbank also presents an inherent conflict of interest, as the Central Bank is not only the owner of the largest bank, but also the supervisor and the state authority responsible for monetary policy.

The Russian government’s involvement in the banking sector is not limited to Sberbank alone. A recent study commissioned by the government itself revealed that state organizations hold stakes in more than 469 banks throughout the country. While most of the shares are small, the state has blocking shares in 45 banks. The government has announced that it plans to divest ownership in all banks where the public share is less than 25 percent. While this will reduce the number dramatically, it means the government will leave Sberbank—and several of its other largest state-owned banks—largely untouched.

154

ANNEX 11: LATVIA: THE RESTRUCTURING AND PRIVATIZATION OF UNIBANKA180

The case of Unibanka represents a success story among transition countries. While it initially engaged in activities that undermined the quality of its loan portfolio and put bank capital at risk, its restructuring exercise proved successful, as reflected in its ability to withstand systemic weaknesses in the mid-1990s and to attract strategic investment in the second half of the 1990s.

The history of Unibanka is an integral part of the general transformation of the Latvian banking system from its earlier monobank roots to its current two-tier system. Following the breakup of the Soviet Union, Latvia found itself in much the same position as the other FSU countries. It inherited branches of the specialized Soviet banks: the Savings Bank (Latvijas Krajbanka), the Agricultural Bank, the Industry and Construction Bank, the Housing and Social Development Bank, and the Foreign Trade Bank. In addition to the inherited problems of large non-performing loan portfolios and management who were unused to lending along commercial lines, the branches were suddenly cut off from their former head offices. Moreover, the banks found that the authorities in Moscow were unwilling to pass on to the newly independent branches the assets needed to cover substantial portions of their liabilities.

Unlike most of the other newly independent countries that converted these branches directly into nationally owned specialized banks corresponding to the old Soviet banks, the Latvian government placed all of the branches of the specialized banks (except the branches of the Savings Bank) under the direct supervision of the Bank of Latvia (the Central Bank). These branches dominated the credit business, since the Savings Bank, initially, did not make loans to either private or public enterprises. As a result, at the end of 1991, the 45 branches controlled 83 percent of all credit to business and also held three-quarters of enterprises’ demand deposits.

In comparison to its Baltic neighbors, which generally kept the individual specialized banks separate, this strategy allowed the Latvian government a wide range of options. The government could sell the branches to the emerging private sector; privatize them either individually or in groups; or structure one or more state banks. However, it also gave the Bank of Latvia a great deal of responsibility at a time when both the sector and the Central Bank itself were undergoing dramatic transformation. In practice, the Bank of Latvia did not play an active role either promoting governance or encouraging the branch managers to run the banks according to strict commercial criteria. Therefore, the managers, who had little experience in commercial banking and little loyalty to their new managers at the Bank of Latvia, found themselves with a great deal of discretionary power during extremely difficult external conditions. The result was that the branches developed significant volumes of non-performing loans.

By 1993, the Government decided on a strategy for dealing with the remnants of the state banking sector, using a combination of the above mentioned approaches. It was decided to keep the Savings Bank in the public sector and submit it to considerable institutional development support and bring in new management prior to privatizing the bank in due course. The main

180 Alex Pankov is the author of this annex. Sources include (i) A. Fleming and S. Talley, “The Latvian Banking Crisis: Lessons Learned,” World Bank Research Paper, 1996; (ii) Fitch Credit Agency, Unibanka Rating Report, 2000; (iii) BankScope, Unibanka Reports, 1999-2001; and (iv) the Baltic News Service.

155

remnant was dealt with in three ways. First, nine branches were sold to private commercial banks. Second, 15 of the branches were consolidated into eight new private commercial banks. Finally, on September 28, 1993, the rump of 21 branches was structured into one state bank – the Universal Bank of Latvia, or Unibanka – and subject to intense institutional development supported by the World Bank, the Government of Switzerland, and the European Union. Most of the bad loans were concentrated in the branches that constituted this bank (40 percent of total assets in March 1994). As part of the rehabilitation process, these loans were taken off Unibanka’s books and replaced with seven-year Government bonds in the amount of LVL 25 million ($50 million). As a result of rapid growth in credit provided by private banks, Unibanka accounted for only seven percent of total credit by the end of 1994 and was the country’s second largest bank in terms of assets.

One of the main reasons cited by the Government for creating Unibanka was to provide an insurance policy against catastrophic failures in the private banking sector. This logic was put to serious test in the first half of 1995, as the insolvency of the country’s largest bank (Bank Baltija) triggered the systemic crisis in which about 40 percent of the assets and liabilities of the banking sector were lost, and seven banks, including three of the 10 largest banks, had collapsed. Although the crisis had a large effect on both large state banks, neither Unibanka nor the Savings Bank was directly involved in the crisis and neither bank needed to be closed or bailed out.

Unibanka, which was already healthier than the Savings Bank, was not as badly harmed as the Savings Bank. Since deposits from individuals accounted for the majority of deposits in the failed banks, these depositors probably lost confidence in the sector. As a result, the crisis had a greater effect on the Savings Bank than Unibanka because almost all deposits in the former institution were deposits by individuals. Between December 1994 and December 1995, individuals’ deposits in the Savings Bank fell by almost 17 percent (in real terms), and total deposits fell by 13 percent. At the same time, individuals’ deposits in Unibanka increased by 16 percent, while total deposits increased by nearly 30 percent. (Assets increased by 33 percent, and profits by 77 percent, in real terms). It appears plausible that Unibanka benefited from (and the Savings Bank suffered from) a flight to quality following the crisis, i.e., that depositors reallocated assets towards banks that appeared better managed, more strongly capitalized, and less risky in the composition of their portfolios. Surveys of Latvian banking professionals consistently rated Unibanka as the safest bank in Latvia.

In accordance with the Government's decision, privatization procedures were launched at Unibanka on October 3, 1995. The board of the Latvian Privatization Agency approved the bank's basic privatization regulations, which provided that Unibanka would be privatized in four years. During 1995, the first stages of Unibanka’s privatization were carried out. Share capital was increased to LVL 11.5 million (about $23 million), and a little over 50 percent of the shares were sold for privatization certificates. Twenty-two percent of shares were sold publicly; 13.5 percent were sold to customers of Unibanka; and 14.5 percent were sold to employees. The Privatization Agency held the remaining shares. In October 1995, the bank's shareholders decided to reorganize the bank into a joint-stock company, Latvijas Unibanka, and a new charter was approved for the bank. In January 1996, Unibanka became the first company to be listed on the Riga Stock Exchange official list.

156

According to the bank's privatization regulations, its share capital was increased during the next privatization round by attracting additional capital from a strategic investor. In May 1996, Unibanka’s share capital was raised by LVL 6 million (about $12 million), and the newly issued shares were purchased by the EBRD and Swedfund International AB. This gave the EBRD control of about 22.6 percent of total shares and Swedfund control of about 7.5 percent of shares. Over the next three years, most of the remaining state-owned shares were sold in the international market through a GDR program, and a part of the shares was sold through special auctions at the Riga Stock Exchange. Overall, the state received LVL 66.1 million (about $113.4million) by the time privatization was complete in late 1999, including LVL 21.3 million in cash and LVL 44.8 million in privatization vouchers.

By September 2001, Unibanka's paid-up share capital was LVL 37.1 million ($59.9 million). More than 98 percent of share capital belongs to the Swedish bank Skandinaviska Enskilda Banken (SEB). A major force in the general trend towards banking sector consolidation in the Baltics, SEB initially purchased a 23 percent interest in Latvijas Unibanka at a special auction held in the stock market in late 1998. It then steadily purchased shares from the other bank's shareholders, including the EBRD’s shares. Starting in early 2001, Latvijas Unibanka's shares are no longer quoted at the Riga Stock Exchange.

Unibanka is currently the second largest bank in Latvia and the fifth largest in the Baltics (by assets). As a universal bank, it provides a wide range of commercial and retail services, concentrating on the domestic market, where it has a solid franchise. SEB has helped Unibanka improve risk management, retail operations and implement credit controls. The bank’s performance since the completion of privatization has been very satisfactory, and positive economic forecasts for Latvia bode well for future growth in Unibanka’s operations.

157

ANNEX 12: INTERNATIONAL BANK OF AZERBAIJAN181

Ten years after the establishment of a two-tier banking system, the Azeri banking system remains at a critical stage of development. While the Government of Azerbaijan has made good progress in stabilizing its economy since 1995, its efforts to address a number of structural issues in the financial sector have had mixed results. While some advances were made in upgrading prudential regulations for banks, strengthening off-site supervision, and in regulating foreign exchange markets, the banking sector suffers from poor management and governance, limited technology, problem loans, and insufficient capital. It is estimated that only 10-20 percent of the total money in circulation passes through the banking system.

Tackling the issue of state-owned banks has proved to be one of the government’s most difficult and complex tasks. Despite early attempts to recapitalize, restructure and privatize state-owned banks since 1996, state ownership in the banking sector remains high, dominated by the International Bank of Azerbaijan (IBA). This bank was left with a near monopoly when the government consolidated its three other troubled state-owned banks into a single entity in 2000. While this step marked significant progress in reducing public ownership, the banking system remains highly concentrated and underdeveloped.

Consolidation Of Three State Banks

Throughout the government’s reform efforts during the mid-1990s, some of the biggest loss-makers in the sector were state-owned banks, including Amanat bank (the savings bank), Prominvest (the industrial investment bank), and Agroprom (the agro-industrial bank). Non-performing loans were particularly problematic at Agroprom and Prominvest, accounting for more than 90 percent of their portfolios. By late 1999, the government recognized that consecutive recapitalizations of these three banks had failed to improve performance, and that it needed to take more radical measures.

As a result, in February 2000, the government created the United State Industrial Bank from the merger of the viable operations of the three banks. The bank was later renamed United Universal Bank. The government issued a new limited license to the entity, allowing it only to collect deposits, perform foreign exchange activities, invest in government securities, and perform cash payment services for the Social Protection and Pension Funds and other budget entities. The terms of the license further prohibited the bank from engaging in lending for two years, so that it will in effect operate as a narrow bank.

The government plans to develop the operational structure of United Universal and to establish an effective lending capacity to help strengthen the bank, improve its efficiency, and create conditions for the eventual privatization of the bank through the sale of a controlling share to a strategic investor.

International Bank Of Azerbaijan Dominates The Banking Sector

181 Alex Gross is the primary author of this annex. Sources include (i) internal World Bank documents, 1999-2000; (ii) BankScope, Fitch IBCA; (iii) Economist Intelligence Unit (various country reports); and (iv) EBRD.

158

With the decision to create United Universal, the government solidified IBA’s position as the country’s leading and best capitalized bank. IBA has 75 percent market share of banking sector assets, and 40 percent of retail deposits. In 2000, bank assets stood at $614 million, and its loan portfolio grew by 22 percent. The bank has close links to many government departments and state organizations, including the important Oil Fund, and acts as an intermediary for government-guaranteed credit lines to Azerbaijan.

IBA was established in 1990 as a replacement for the Azerbaijan branch of Vnesheconombank, the former Soviet foreign trade bank. In keeping with its origins, the foreign trade operations are well established, but IBA also accepts deposits from and issues loans to Azeri firms. The bank is steadily increasing its retail operations. The bank has 32 branches and approximately 700 employees, providing it with a relatively large network given the small size of the country.

Towards Privatization

In 2001, the government reiterated its commitment to privatize IBA and issued a presidential decree to that effect. At present, the Ministry of Finance owns 50.2 percent of the bank’s shares. The EBRD, which has been providing support for the IBA in the form of a credit line targeted to small and medium enterprises, has indicated an interest in taking on a 20 percent equity stake. The remaining state shares are expected to be auctioned at a later date.

Despite its dominant position, IBA faces many governance and management problems and is plagued by inefficiency. Its profitability was weak in 2000, with after-tax

earnings of only $9 million. The slight increase in profit that the bank did see was primarily due to the net interest income earned on the placement of funds of the Azeri Oil Fund. These revenues are not expected to be repeated in 2001. Further, the level of overdue loans rose in 2000 (as did loan loss provisions), and IBA’s loan loss reserve cover was only 12.5 percent of gross loans at end 2000.

While the government’s recent moves related to IBA and United Universal represent progress, privatization is just one element of the financial sector reforms that are necessary. The broader challenge is to make banks more central to economic activity in Azerbaijan in terms of deposit mobilization, lending, and an increased array of services.

159

ANNEX 13: METHODOLOGY FOR THE STUDY ON STATE BANKS IN EUROPE AND CENTRAL ASIA

The team relied on primary data from the following sources as the main tools in assessing the current state of public sector banks in Europe and Central Asia:

Bank-specific data were generally taken from Bureau Van Dijk’s BankScope and are based on the banks’ official annual reports. It should be noted that the data for some banks in BankScope (primarily in CIS countries) are unqualified and/or unaudited. Internal World Bank data were used in a few instances (Bosnia-Herzegovina, Kyrgyz Republic, Uzbekistan, Yugoslavia).

Aggregate data for the banking sector were mostly derived from the IMF’s International Financial Statistics (IFS).

Data on stocks and flows of arrears were primarily sourced from official IMF reports and working papers. This was supplemented with data from externally distributed reports from the EBRD (Transition Reports), the European Union (EU-TACIS), and the World Bank.

Macroeconomic data were primarily sourced from the World Bank’s World Development Indicators database.

At the same time, the team used secondary data from the following sources to detail the history of state banking in transition economies over the past decade, including the case studies for selected state-owned banks:

Official and externally distributed publications by the World Bank and IMF; EBRD (Transition Reports for years 1998-2001); European Union (EU-TACIS Country Economic Trends); bank rating agencies (Fitch Research and Moodys); Economist Intelligence Unit (various country reports); and OECD (annual reports).

Publications and websites of government agencies, including the central banks (annual, quarterly, and monthly reports on the financial sector) and the state statistical committees.

Internal World Bank documents on the financial sectors of the Europe and Central Asia region.

Amounts presented in US dollar-equivalent terms are signified by the term “$” 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.

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.

160

Despite efforts to create comprehensive data sets for 1992-2000, gaps still remain. These particularly relate to both aggregate and bank-specific data for the first half of the 1990s, due to the poor data collection and accounting standards common to many countries in the early transition period.

161

Bibliography

Annual Report, Sberbank (Russia), 2000.

Aslund, Anders and Richared Layard, 1993. Changing the Economic System in Russia. New York, NY.

Bagratian, Hrant and Emine Gürgen, 1997. Payment Arrears in the Gas and Electric Power Sectors of the Russian Federation and Ukraine. IMF Working Paper, International Monetary Fund, Washington DC.

Bank of Slovenia, various reports. (http://www.bsi.si).

Banking Agency of the Republika Srpska, 2001. Report on the Condition of the RS Banking System in the Year 2000. Banja Luka, Bosnia and Herzegovina.

Barbone, Luca, and Domenico Marchetti, 1994. Economic Transformation and the Fiscal Crisis: A Critical Look at the Central European Experience of the 1990s. World Bank Working Paper 1286. Washington, DC.

Borish, Michael, Millard Long and Michel Noël, 1995. Restructuring Banks and Enterprises: Recent Lessons from Transition Countries. World Bank, Washington DC.

Borish, Michael and Michel Noël, 1996. Private Sector Development During Transition: The Visegrad Countries. World Bank, Washington DC.

Borish, Michael, Wei Ding and Michel Noël, 1996. On the Road to EU Accession: Financial Sector Development in Central Europe. World Bank, Washington DC.

Builov, Maxim, 2002. Who Owns Russia: Russia’s Banking Sector. The Russian Journal. Moscow, Russia..

Bureau Van Dijk's BankScope, 1996-2002. Bank reports (various). (http: //bankscope.bvdep.com/).

Central Bureau of Statistics, Republic of Croatia, 1993. Statistical Yearbook. Zagreb, Croatia.

Claessens, Stijn, 1996. Financial Sector Reform in Transition Countries. Policy Note. World Bank, Washington, DC.

Czech Narodni Banka, various reports. (http://www.cnb.cz).

Economic Policy Institute (Bulgaria), 2001. Restructuring and Development of the Banking Sector in Advanced Transition Countries: Lessons to be Learnt by Other Transition Economies. Sofia, Bulgaria.

162

Economist Intelligence Unit, 1998-2002. Country Reports (various). Economist Intelligence Unit, London, UK.

Essinger, James, 1994. Eastern European Banking. London, UK.

European Bank for Reconstruction and Development, 1994-2001. Transition Report (various years). European Bank for Reconstruction and Development, London, UK.

Federal Banking Agency of the Federation of Bosnia and Herzegovina, 2001. Information on the Banking System in the Federation of Bosnia and Herzegovina, Working Paper. Sarajevo, Bosnia and Herzegovina.

Federal Statistical Office, 1993. Statistical Pocket Book of Yugoslavia. Belgrade, Yugoslavia.

Fischer, Stanley and Ratna Sahay, September 2000. Taking Stock. Finance & Development. Washington, DC.

Fitch Credit Disk, 1999-2002. Bank Reports (various). (http://www.fitchratings.com/).

Fleming, Alex, Lily Chu and Marie-Renée Bakker, 1996. The Baltics—Banking Crises Observed. World Bank Research Paper, Washington, DC.

Fleming, Alex and Samuel Talley, 1996. The Latvian Banking Crisis: Lessons Learned. World Bank Research Paper, Washington, DC.

Fuchs, Michael, 2002. Building Trust Developing the Russian Financial sector (draft). World Bank, Washington DC.

International Monetary Fund, 1997. Kyrgyz Republic: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 1997. Ukraine: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 1998. Albania: Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 1998. Azerbaijan Republic: Recent Economic Developments and Selected Issues. International Monetary Fund, Washington DC.

International Monetary Fund, 1998. Georgia: Recent Economic Developments and Selected Issues. International Monetary Fund, Washington DC.

International Monetary Fund, 1998. Republic of Armenia: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 1998. Republic of Estonia: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

163

International Monetary Fund, 1998. Republic of Latvia: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 1998. Republic of Lithuania: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 1998. Republic of Uzbekistan: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 1998. Romania: Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 1998. Turkmenistan: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 1999 Armenia: Recent Economic Developments and Selected Issues. International Monetary Fund, Washington DC.

International Monetary Fund, 1999. Republic of Latvia: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 1999. Republic of Lithuania: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 1999. Turkmenistan: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 1999. Ukraine: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Azerbaijan Republic: Recent Economic Developments and Selected Issues. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Bosnia and Herzegovina: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Former Yugoslav Republic of Macedonia: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Georgia: Recent Economic Developments and Selected Issues. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Kyrgyz Republic: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Republic of Belarus: Recent Economic Developments and Selected Issues. International Monetary Fund, Washington DC.

164

International Monetary Fund, 2000. Republic of Croatia: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Republic of Estonia: Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Republic of Latvia: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Republic of Uzbekistan: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 2000. Russian Federation: Selected Issues. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Albania: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Bosnia and Herzegovina: Sixth and Seventh Reviews Under the Stand-By Arrangement and Request for Waiver of Performance Criterion; Staff Statement; and News Brief. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Bulgaria: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Georgia: Recent Economic Developments and Selected Issues. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Kyrgyz Republic: Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Republic of Armenia: Recent Economic Developments and Selected Issues. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Republic of Croatia: First Review of the Stand-by Arrangement – Staff Reports; Staff Supplements; News Brief on the Executive Board Discussion; and Statement by the Executive Director for the Republic of Croatia. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Republic of Estonia: Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Republic of Moldova: Recent Economic Developments. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Republic of Tajikistan: Statistical Appendix. International Monetary Fund, Washington DC.

165

International Monetary Fund, 2001. Romania: Selected Issues and Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Russian Federation: Post-Program Monitoring Discussions – Staff Report and Public Information Notice on The Executive Board Discussion. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Ukraine: Fifth and Sixth Reviews Under the Extended Arrangement – Staff Report; Staff Supplement and News Brief on the Executive Board Discussion. International Monetary Fund, Washington DC.

International Monetary Fund, 2001. Ukraine: Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2002. Republic of Belarus: Selected Issues. International Monetary Fund, Washington DC.

International Monetary Fund, 2002. Republic of Latvia: Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2002. Republic of Lithuania: Statistical Appendix. International Monetary Fund, Washington DC.

International Monetary Fund, 2002. International Financial Statistics. International Monetary Fund, Washington DC.

Jones, Colin, 1993. Banking and Financial Sectors in East and Central Europe. Financial Times Management Reports, London, UK.

Lane, David, 2001. The Evolution of Post-socialist Banks in Russia. Cambridge, UK

Lapidus, Mikhail, 1997. Understanding Russian Banking. Mir House International, Kansas City, KS.

Maletic, Cedo, 2002. Overview of Croatian Banking Sector: the Causes and Cost of the Two Recent Banking Crises. National Bank of Croatia. Zagreb, Croatia.

Merrill Lynch, 2001. Turkish Banking: In Depth Report. Global Fundamental Equity Research Department.

Moodys Investor Service, 1997-2002. Banking System Outlook (various). (www.moodys.com).

Narodna Banka Slovenska, various reports. (http://www.nbs.sk)

Organization for Economic Cooperation and Development, 1999-2001. Annual Reports (various). Paris, France.

Oxford Analytica Brief, 1995-2002. Various reports. (www.oxan.com).

166

Richard Roulier, 1995. Bank Governance Contracts: Establishing Goals and Accountability in Bank Restructuring. World Bank, Washington DC.

Russian-European Center for Economic Policy, various. Russian Economic Trends. Blackwell Publishers Journals, Oxford, UK.

Siegelbaum, Paul, Khaled Sherif, Michael Borish, and George Clarke, 2002. Structural Adjustment in the Transition: Case Studies from Albania, Azerbaijan, the Kyrgyz Republic and Moldova. World Bank Discussion Paper 429. Washington, DC.

Sundararajan, V., Arne Petersen, and Gabriel Sensenbrenner, 1997. Central Bank Reform in the Transition Economies. International Monetary Fund, Washington DC.

TACIS, Various. Armenia Economic Trends. TACIS, Brussels, Belgium. (http://www.economic-trends.org/).

TACIS, Various. Azerbaijan Economic Trends. TACIS, Brussels, Belgium. (http://www.economic-trends.org/).

TACIS, Various. Moldova Economic Trends. TACIS, Brussels, Belgium. (http://www.economic-trends.org/).

Tchaidze, Robert, 1999. Non-Cash Settlements of Expenditures in Baltics, Russia, and Other Former Soviet Republics. Mimeo, International Monetary Fund, Washington DC.

Ukrainian-European Policy and Legal Advice Center, 1998-2002. Ukrainian Economic Trends (various issues). TACIS, Brussels, Belgium. (http://www.economic-trends.org/).

World Bank, 1993. Azerbaijan: From Crisis to Sustained Growth. World Bank, Washington DC.

World Bank, 1993. Uzbekistan: An Agenda for Economic Reform. World Bank, Washington DC.

World Bank, 1994. Turkmenistan: A World Bank Country Study. World Bank, Washington DC.

World Bank, 1995. Hungary: Structural Reforms for Sustainable Growth. World Bank, Washington DC.

World Bank, 1998. Slovenia: Economic Transformation and EU Accession. Volume II: Main Report. World Bank, Washington DC.

World Bank, 2000. Kyrgyz Republic. Fiscal Sustainability Study. World Bank, Washington DC.

World Bank, 2000. Tajikistan Banking Sector Note. Draft. World Bank, Washington DC.

World Bank, 2001. Economic Trends – Croatia. World Bank Country Office, Zagreb, Croatia.

World Bank, 2001. World Development Indicators. World Bank, Washington DC.

167

Zoli, Edda, 2001. Cost and Effectiveness of Banking Sector Restructuring in Transition Economies. Working Paper 157, International Monetary Fund, Washington D.C.

168