Financial Sector Reform and Banking Crises in the Baltic Countries
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar
Close

Financial sector reform in the Baltic countries is reviewed in light of the banking crises that emerged during the reform period. It is argued that the crises had their roots in the structural deficiencies specific to planned economies and the financial environment that developed before and after these countries regained their independence, thus rendering them largely inevitable. Because of the low level of financial intermediation, however, even the failure of large banks had limited systemic effects and a minor negative impact on output and incomes. The crises slowed down the financial reform process, but brought about a desired consolidation of the banking sector.

Abstract

Financial sector reform in the Baltic countries is reviewed in light of the banking crises that emerged during the reform period. It is argued that the crises had their roots in the structural deficiencies specific to planned economies and the financial environment that developed before and after these countries regained their independence, thus rendering them largely inevitable. Because of the low level of financial intermediation, however, even the failure of large banks had limited systemic effects and a minor negative impact on output and incomes. The crises slowed down the financial reform process, but brought about a desired consolidation of the banking sector.

I. Introduction

1. As the Baltic countries embarked on their transition to a market system, after regaining independence, it was expected that the financial sector would play a critical role in the fundamental restructuring of the economy. In the absence of capital markets, the allocation of resources, the mobilization of savings, and the development of corporate governance at the enterprise level would rely heavily on the banking sector. A well-functioning banking system was also deemed essential for the efficient conduct of monetary policy in a market environment and thus for the achievement and maintenance of macroeconomic stabilization.

2. The banking system, however, having evolved under the Soviet Union’s command economy, was ill-suited to perform these roles. Banks had little experience of how to operate in a market economy and were insufficiently regulated and supervised. Close connection with their borrowers, together with uncertainties and risks associated with a changing economic environment, made it unlikely that lending decisions would be based on profit-maximization criteria.2 At the same time, the lack of effective monetary instruments and developed financial markets to transmit monetary impulses to the real economy were severe obstacles for the implementation of market-based monetary policy. These drawbacks underscored the need to implement major reforms in the financial sector as part of the overall transformation process.

3. Accordingly, financial sector reform has been an important item of the agenda in the Baltic countries from the beginning of the reform process. These reforms have involved not only the liberalization of a repressed financial system, but, more importantly, the build-up of institutions, expertise, markets, instruments, as well as significant changes in the regulatory and legal frameworks, and an overall change in attitudes. Although the process of financial deregulation has been largely completed, given the magnitude of the task, it is not surprising that financial sector reforms are still continuing with efforts to develop a more competitive and sound banking system, an effective regulatory framework, developed financial markets, and an efficient and safe payments system.3

4. During the course of reform, the Baltic countries experienced, to varying degrees and forms, banking crises which prompted different policy responses from the respective authorities. Nevertheless, a common theme across the three countries was that their banking problems evolved from microeconomic/structural roots specific to planned economies and to the financial and economic environment that developed during the first years after regaining independence. Accordingly, it is argued that the banking crises were almost inevitable given the lack of banking expertise, markets, institutions, effective banking supervision and, importantly, the extent of connected lending. On the other hand, due to the low level of financial intermediation in these countries, even the failure of large banks had limited systemic effects and only a minor negative impact on output and incomes.

5. The banking crises which emerged in the Baltic countries between 1992 and 1995 slowed down the financial reform process and temporarily reversed the trend of deepening financial intermediation. The crises have highlighted the importance of pressing ahead simultaneously with structural reform in all sectors of the economy, in particular of establishing a solvent and profitable banking system and a viable enterprise sector to support macroeconomic stabilization.

6. The paper is organized as follows. Section II describes as background the typical role of banks in a market economy and contrasts this to the role banks played under the command system. Section III evaluates the record of financial sector reform in the Baltics. Section IV assesses Baltic banks’ contribution to financial intermediation. Section V discusses the origins and triggers of recent banking crises, the authorities’ policy responses, and the effects of the crises. Section VI summarizes the main conclusions.

II. Background

A. Banks in a Market Economy

7. A stable, well-functioning, and efficient banking system plays a fundamental role in supporting macroeconomic stability in a market economy. Moreover, the banking system serves as the main conduit for mobilizing savings and transforming these into investment. Financial intermediation allows the savings decision of individual economic agents to be dissociated from the decisions of investors, with the market coordinating their aggregate behavior through interest rates. Banks facilitate the efficient allocation of resources by channeling funds to their most productive use. At the same time, banks assume, assess, and manage risk, and by monitoring borrowers closely, they participate in effective corporate governance of the enterprise sector. Banks also provide payment system services which are essential for facilitating transactions.

8. It needs to be recognized that, while playing these essential roles in the economy, the banking system is more apt to be vulnerable to shocks than other sectors. As financial intermediaries, banks transform maturities: their liabilities are normally short term and can be called on demand, while assets are longer term, illiquid, and difficult to value. The maturity mismatch between assets and liabilities, the non-transparent nature of assets, and their highly leveraged balance sheets make banks vulnerable to distress and failure. The possibility of sudden deposit withdrawals, together with the banks’ limited ability to liquidate assets quickly, may cause liquidity difficulties. Since deposit withdrawals are on a first-come-first-serve basis, and exercising this option is at the low cost of foregone interest compared with the possibility of, losing the capital value of deposits, there is an incentive for depositors to act quickly (run) if confidence problems arise. Due to the close linkages among financial institutions through the payments system and the interbank money market, liquidity problems in one part of the system are likely to spill over to other sound institutions, i.e. become systemic.4 Furthermore, isolated problems of individual banks may affect negatively depositors’ confidence in other banks or in the system as a whole. The opaque nature of assets makes it difficult for individual depositors to adequately evaluate the quality of banks’ portfolios. Unable to distinguish between good and bad banks and given the low cost of withdrawal to depositors, individuals have a strong incentive to withdraw funds from the system as quickly as possible.

9. The important role the banking system plays in a market economy, its special vulnerability and the severe negative impact that banking failures may have on output and incomes5 have led governments to closely monitor, supervise, and safeguard the banking system.

B. The Role of the Banking System Under the Soviet Union’s Command Economy

10. Before regaining their independence in 1991, the Baltic countries shared the common banking system of the Soviet Union, whose purpose and functions were very different from the role banks have in market economies. Until 1987/88 the (monobank) system was characterized by the lack of an institutional separation between monetary policy (typically conducted by a central bank in a market economy) and commercial banking functions.6 Specialized banks were closely linked to line ministries and their chief objective was to execute and implement the credit plan, which was supposed to mirror the production plan in the command economy. In sharp contrast to the role played in market economies, the banking system did not mobilize savings. Funds were allocated to banks through the budgetary process to finance investment projects which were also specified in the budget. In this system, there was no need to coordinate the behavior of savers and investors; therefore, interest rates did not play a role in guiding resource allocation.

11. In a system which was designed to serve principally as an institutionalized record-keeping device to monitor enterprise transactions,7 banks did not need to evaluate credit risk. The use of credit transfer instruments (demand orders) and banks’ practice to make funds available to creditors immediately allowed creditor enterprises to receive funds even when debtor enterprises did not have adequate funds on their current accounts. The resulting payments system arrears of enterprises and loans, particularly to the agricultural sector, were routinely written off at the end of each year.8 Not surprisingly, these soft budget constraints for banks and enterprises provided no incentive to enforce payment obligations. The system therefore did not require any banking supervision in the usual sense and consequently, supervision was limited to monitoring the fulfillment of the credit plan and auditing within institutions. Furthermore, bankers were not required to have the skills necessary for banking in a market economy.

12. The payments system was also distinct from those found in market economies. Financial flows in the command economy were strictly separated between enterprises and households. While enterprises relied on account transactions for making payments to suppliers, cash transactions were restricted to and were dominant for effecting payments between households and enterprises, and for wage payments.

C. Early Banking System Reform (1987-91)

13. The gradual banking system reform, which began in 1987/88, aimed at addressing some of the weaknesses inherent in the existing institutional structure, with the ultimate objective of strengthening macroeconomic management and improving resource allocation. The existing monobank system was replaced by a two-tier system, and the Gosbank U.S.S.R. was given the responsibility for conducting monetary policy and other typical central banking functions, while specialized state banks were established to perform all commercial banking functions. These banks received greater independence from line ministries and were expected to operate as profit centers.9 However, since this specialization effectively precluded competition between banks and there was little incentive to respond innovatively to the changing economic environment, their behavior in practice did not change much following the reform.

14. The new banking law introduced as part of the 1987/88 reforms permitted for the first time the entry of new banks and allowed them to operate independently from the central plan.10 Initially, cooperatives were allowed to establish cooperative banks, and subsequently state enterprises were allowed to found commercial banks.11 These institutions, however, were typically created to address the financial needs of enterprises which were dissatisfied with the slow and bureaucratic procedures of the existing banking system.12 Many of these banks had included in their charters the explicit purpose of financing their enterprise owners or related enterprises (“pocket banks”). Their resources consisted mainly of enterprise deposits, deposits of state banks, and own capital. Most of their lending was short term and often to the banks’ own shareholders.

15. Although by mid-1989 various prudential regulations had already been introduced, including minimum standards on liquidity, capital, risk concentration, auditing, and the securing of collateral, they were insufficiently strict and their enforcement was weak. Banking supervision by Gosbank was by and large ineffective. Internal control and risk management procedures in commercial banks were either nonexistent or insufficiently developed, as they were not generally designed to provide adequate reporting on banks’ financial condition and to monitor effectively and control the various risks of banking.

16. Low barriers to entry, in particular, liberal licensing policies and low minimum capital requirements, which in addition were subject to inflationary erosion, led to the proliferation of small banks in all Baltic countries. By 1991/1992, the number of commercial banks had increased to 42 in Estonia, 51 in Latvia, and 20 in Lithuania (Chart 1). State banks remained dominant, however, and continued to hold over 50 percent of banking system assets in 1991. Initially, limits were imposed on deposit interest rates for commercial and cooperative banks to protect the deposit base of the Savings Bank. However, even after the ceilings were removed in 1992, at first most of these banks were unable to attract a significant amount of deposits. Nevertheless, particularly in Latvia and to a lesser extent in Lithuania, a few of the new banks grew rapidly, drawing on the large profits obtained from transit trade financing, and were able to position themselves to become major players in the new banking system.13

CHART 1
CHART 1

THE BALTICS NUMBER OF BANKS 1/

Citation: IMF Working Papers 1996, 134; 10.5089/9781451855555.001.A001

1/ Excluding branches and representative offices of foreign banks; end-of-year.2/ End-June 1996.

III. Financial Sector Reform Since 1991

17. The second stage of financial sector reform in the Baltic countries began after they regained independence in 1991, and gained momentum following currency reform in 1992. Against a background of high inflation, significant deterioration in the terms-of-trade and large declines in output, the Baltic countries undertook to stabilize their economies based on prudent financial policies.14 However, these countries lacked institutional, regulatory and legal frameworks which could be adapted to the needs of a market system and thus support financial deregulation. Therefore, on various fronts, notably on institution-building, creation of markets, and the development of appropriate legal and regulatory frameworks, it was necessary to start from scratch: (a) central banks needed to be reorganized to conduct monetary policy; (b) the banking system needed to be restructured to encourage financial intermediation; (c) the payments system needed to be adapted to serve the market; (d) money, securities and capital markets needed to be developed to facilitate the operations of the banking system; (e) monetary instruments needed to be put in place for conducting market-based monetary policy; (f) banking supervision capacity needed to be built up and effectiveness of prudential regulations needed to be strengthened in order to discourage banks from excessive risk-taking and to ensure the viability and health of the banking system; and (g) bankruptcy and collateral legislation needed to be passed to allow banks and enterprises to enforce and secure contracts.

18. Although financial liberalization measures were implemented early in the transition process (Box 1), supporting structural reforms typically lagged behind, on account of their longer-term nature and redistributive implications, as well as social and political resistance. The resulting bottlenecks and imbalances are at the heart of the financial crises that emerged later.

Liberalization Measures

Liberalization measures adopted in the early stages of reform included the phasing out of directed and subsidized credits and interest rate controls, the introduction of current account convertibility, and the elimination of most restrictions on capital movements. Other measures, aimed at making the financial sector more responsive to market forces, included changes in monetary policy instruments, such as the introduction of obligatory reserve requirements and refinance facilities. Prudential regulations and reporting systems for banks were also introduced, but were insufficiently strict and enforcement was weak. In the absence of an appropriate regulatory framework, reliable banking supervision and the discipline imposed by well-functioning markets, some of these measures—such as the removal of interest rate and capital controls—may have contributed to raising financial fragility, as increased freedom encouraged banks to take excessive risks.

As demonstrated by the Latin American experience, high and volatile real interest rates are likely to emerge after deregulation,1/ while rapid capital account liberalization often has led to financial distress. In the case of the Baltics, specifically in Latvia and Lithuania, the liberal environment and high interest rates allowed banks to attract significant amounts of speculative foreign capital, notably from the Russian Federation and other CIS countries. In addition to the consequences of connected lending, the absence of effective prudential supervision, notably of limits on open foreign exchange positions, contributed to increase banks’ exposure to exchange risk, particularly in Latvia, and to encourage the accumulation of bad loans.

1/ See, for example, Lindgren, Garcia, and Saal (1996).

Toward a Two-Tier Banking System

After regaining independence in 1991, the Baltic countries moved rapidly to restructure the highly specialized and segmented financial sector inherited from the Soviet Union. In Estonia, the state-owned banks were transformed into joint-stock companies. By September 1991, there were 20 commercial banks; however, the four state banks still extended some 60 percent of credit. Ownership of the commercial banks was mainly concentrated in the hands of state enterprises and joint-stock companies, with some participation by local governments and cooperatives. In 1992, the government decided that the regional branches of Agroprom Bank should become independent commercial banks. Also, the Bank of Estonia was merged with the Estonian branch of Gosbank, which included the Estonian branch of the Savings Bank at that time. As of mid-1996, there was one branch and six representative offices of foreign banks in Estonia.

In Latvia, the regional branches of the former Soviet specialized banks (except for the Savings Bank, which remained under state ownership) were nationalized and taken over by the Bank of Latvia, which then became responsible for both commercial and central banking functions. During 1992, the commercial and central banking activities of the Bank of Latvia were gradually separated, and in 1993 the branches of the former state specialized banks were sold, closed or privatized. The Universal Bank of Latvia (Unibank) was then created from the merger of 21 former branches. The move toward a two-tier system resulted in a large number of private banks. Of the 38 commercial banks operating in Latvia at the end of 1995, three were state owned: the Latvian Savings Bank (LSB), the Unibank (now in the process of being privatized), and the Mortgage and Land Bank, accounting for about 20 percent of total bank assets (14 percent at the end of 1994). One foreign bank (Societé Generale) also operates in the country. The other banks are privately-owned or joint-stock companies with private and state capital. Several banks have foreign participation, with a predominance of Russian capital, attracted by the liberal system in Latvia as well as familiarity with the business environment and the possibility of using the Russian language.

In Lithuania, the regional branch of Gosbank and some of the regional branches of the former Soviet specialized banks were absorbed by the newly established Bank of Lithuania, which at first became responsible for both central and commercial banking functions. By 1992, the Bank of Lithuania separated itself from commercial banking activities. One of the characteristics of this separation was the mismatch of assets and liabilities in the new banks, which depended on the correspondent account balances with the Bank of Lithuania that the branches happened to have at the time of the separation; the State Commercial Bank (SCB), for example, which was formed by the majority of the branches of the Bank of Lithuania, was left with a large overdraft, while Aura Bank, which was formed by the Vilnius branch of the Bank of Lithuania, was left with a large positive balance due to substantial government deposits at that branch. The Agricultural Bank and the Savings Bank continued to function as nationalized state owned banks and, together with the SCB, remained the dominant banking institutions through the banking crisis which occurred in late 1995/early 1996. At its peak, in early 1994, the number of banks in Lithuania reached 28 and consolidation began even before the banking crisis two years later. By the middle of 1996, the three state controlled banks accounted for two thirds of total commercial bank deposits (46 percent in the Savings Bank). Three foreign banks have representative offices in Lithuania.

A. The New Banking System

19. The first step toward the liberalization of domestic financial systems was the establishment of a monetary authority. While central banks were established in all three countries in 1990-91,15 not until 1992 were commercial functions split off from the central banks and a true two-tier banking system created (Box 2). From early on, the Baltic central banks enjoyed different degrees of autonomy with respect to the conduct of monetary policy. A currency board was introduced in Estonia in June 1992,16 mainly with the objective of enhancing the credibility of its exchange rate peg, Latvia and Lithuania (the latter until April 1994) opted for a traditional central bank which, in the case of Latvia, was granted a large degree of independence. In April 1994, a currency board arrangement was introduced in Lithuania with the intention of enhancing the credibility of the authorities’ stabilization policies.17 Since experience in conducting monetary policy was limited, following a currency board rule did not require extensive skills in monetary policy formulation and implementation. It also allowed the central banks of Estonia and Lithuania to devote more of their resources to developing banking regulations and improving monitoring of the banking system. In all cases, fiscal responsibility was crucial for the successful implementation of monetary policy.

20. The next crucial step in the development of the new banking system was the reorganization and restructuring of state banks. These banks were first incorporated and converted into joint-stock companies in 1991. In some, control was transferred to their main customers through injection of new share capital, while in others the government maintained majority share holdings. Similarly to the newly-established commercial banks, the interlocking ownership and lending patterns between the former state banks and their shareholders, in particular state enterprises, created an environment in which banking decisions were not always based on market principles. Preferential access to credit was given to state enterprises—which were either shareholders of banks or closely related to them—and thus contributed to the delay of enterprise restructuring.

21. The accumulation of bad loans was at the core of the banking crises in the Baltics. When the state banks were incorporated, no effort was made to establish the true value of their assets nor to address any existing non-performing loan problem, with the exception of the respective Savings Banks and the Latvian Unibank. Part of the problem (the “stock” dimension) was addressed in Latvia and Lithuania by the issuance of long-term bonds to replace non-performing loans of state-controlled banks. Experience in all three countries confirms that recapitalization without simultaneous restructuring cannot be successful, since it addresses only the stock problem without creating the proper incentives for avoiding the accumulation of new bad loans (the “flow” problem).18 In general, following their incorporation, state banks remained undercapitalized and with weak loan portfolios. Their importance has declined significantly over the last years as a result of privatization, liquidation, and slow expansion. In early 1996, the share of state banks’ assets in the banking system had fallen to around 10 percent in Estonia and 20 percent in Latvia, and remained just over 50 percent in Lithuania.

22. Unlike other Central and Eastern European countries, where a large proportion of the newly-created banks were state-owned, private banks mushroomed in the Baltics. This was especially the case in Latvia, where the move to a two-tier system involved the sale/privatization of former branches of specialized state banks. As a result, a large share of financial intermediation is now undertaken by private banks. In the case of Estonia, the rapid increase in the number of small private banks was brought to a halt early in the reform process through tighter licensing requirements, stricter enforcement of prudential regulations, and a gradual increase in the minimum capital requirement. The authorities recognized early that, while a large number of banks did not contribute significantly to financial intermediation, they did absorb supervisory attention. Therefore, licensing requirements were strengthened through more thorough investigation of banks’ major shareholders and their business plans, and the minimum capital requirement was increased in stages. The initial increase of the minimum capital requirement in early 1993 reduced the number of banks from 42 to 23. Most of these banks were liquidated without loss to depositors. Further substantial increases of the minimum capital requirement reduced the number of banks to 15 in mid-1996. Although it started much later, consolidation of the banking system is also well under way in Latvia and Lithuania. Peaking at 63 at the end of 1993, the number of Latvian banks had declined to 35 by mid-1996. In Lithuania, the number of banks dropped from a high of 28 at the beginning of 1994 to 13 in mid-1996. Given the fragility of several institutions and the ongoing restructuring of the enterprise sector, the consolidation of the banking sector, through mergers and liquidations, is likely to continue.

23. The initial capital position for the banking system was low in all three countries. In Latvia it rose significantly, however, from less than 1 percent of GDP in 1992 to 6 percent in 1994, before declining to just over 4 percent in 1995 due to the banking crisis. In Estonia and Lithuania, the increase in banks’ capital did not keep up with nominal GDP growth, thus falling from about 4 percent to about 3 percent and from about 2 percent to about 1 percent of nominal GDP, respectively. The problem of undercapitalization was exacerbated by the fact that part of the modest recapitalization that took place during this period appears to have been funded through loans extended to shareholders. With capital consisting of borrowed funds, owners’ own funds were not placed at risk and therefore banks had no incentive to limit risk-taking.

24. Despite the large number of banks, competition remains limited and the market is still highly segmented. Prior to the banking crisis, the five leading Latvian banks accounted for 66 percent of deposits and 53 percent of total assets of the banking system. At the end of 1995, following the banking crisis, the four largest banks accounted for 60 percent of deposits and 43 percent of assets.19 There is even higher concentration in Lithuania, where the five largest banks accounted for 77 percent of total bank deposits and 71 percent of total bank assets in late 1995, and in Estonia, where the five largest banks accounted for almost 80 percent of deposits and about 60 percent of assets in the same period. Competition for deposits is strong among the larger banks, with wide branch networks, while the smaller banks have a limited market share and cater to a specific and narrow clientele. Several banks have specialized in services, often related to foreign exchange transactions, and provide little credit. A few banks concentrate on the lucrative transit business.

25. Although there is no legal impediment to the entry of foreign banks, at end-1995 there was only one foreign bank operating in Latvia, three representative offices of foreign banks in Lithuania, and two branch offices and two representative offices in Estonia. The reluctance of foreign banks to enter the Baltic countries might have been related to the small size of their domestic markets. Since the beginning of the reform process, foreign banks were encouraged to enter the Estonian banking market. It was hoped that their expertise and technology would generate positive externalities by creating a culture for banking in market economies, that they would encourage competition and contribute to the capitalization of the system. However, while already in 1992 some banks were established based on foreign capital, they were not related to any foreign bank. Their main purpose appears to have been to facilitate trade financing and payment transactions for their owners and not to become universal banks operating in the domestic market. Only in 1995 were the first branch offices of foreign banks established, but they have been specializing in trade financing so far. The entry of foreign banks has been encouraged also in Lithuania although changing and cumbersome regulatory procedures have not provided the necessary incentives.

26. The payments system inherited from the command economy, whose institutional setup implied the maintenance of large reserves of commercial banks at the central bank, was not adequate for operating in the new environment. First, transactions were recorded on paper and settlement was not same-day. Second, risks were not clearly assigned. Third, the systems did not allow for transferring large values with same day settlement and a second clearing stage was not available.20

B. Development of Money Markets and Monetary Control Procedures

27. As in other transition economies, the financial sector in the Baltic countries is dominated by banking institutions, with little activity in the money, securities and capital markets. One of the key tasks for central banks during the early stages of the transition to a market economy is to foster the development of the financial sector, in order to pave the way for the introduction of indirect market instruments and the development of financial intermediation. Structural weaknesses in the banking and enterprise sectors typically hinder financial market development and thus severely limit the central bank’s ability to conduct market-based monetary policy. Moreover, highly distorted credit markets tend to produce high and volatile interest rates, which have little value as signaling devices.

28. The development of the interbank money market has followed different paths in the Baltic countries. In Latvia and Lithuania, activity in the interbank market has been limited, with low trading volumes and a small number of participating banks. The average daily volume of transactions peaked at ½ percent of reserve money in Latvia and in Lithuania, in December 1994 and in the summer of 1995, respectively; in both countries activity declined thereafter as banking problems surfaced, although it picked up somewhat in Latvia in the first part of 1996. One basic stumbling block for the development of an active interbank market in these countries is the persistence of high credit risk, in turn linked to the structural weaknesses and fragility of the banking system as a whole. The banking crises contributed to the further reduction of activity, given the perceived risks involved in lending to some banks. In Estonia, by contrast, the interbank market has become an important vehicle for managing day-to-day liquidity. Initially, the Bank of Estonia issued certificates of deposit which were used as collateral in interbank transactions.21 Banks very quickly gained experience, however, and became more capable of evaluating counterparty risk; most lending in the interbank market is now uncollateralized. Already in 1994 the daily volume of transactions exceeded one percent of reserve money and interest rates followed closely the German interbank money market rate. Substantial funds were also made available to Social Bank during its crisis, against collateral.

29. Both Latvia and Lithuania have successfully established a primary market in treasury bills through an auction procedure.22 Stocks outstanding increased steadily since their introduction in December 1993 and July 1994, respectively, except in the context of banking crises, when there was a substantial weakening in demand for bills reflecting banks’ liquidity problems. In Latvia, after a significant weakening of demand in the second quarter of 1995, the treasury-bill market rebounded during the second half of the year due to a more flexible, interest rate policy and an improvement in the liquidity position of banks; a gradual move toward longer maturities took place subsequently. Central bank participation was initially modest, but has increased steadily, through purchases and sales of treasury bills at the secondary market window, and more recently also through repurchase agreements with banks. In Lithuania, as in Latvia, demand for treasury bills dropped sharply after the December 1995 banking crisis and yields doubled. The treasury bill market recovered in April 1996, however, after foreign buyers entered the market and domestic buyers, including individuals, re-entered. In both countries, treasury bills have been used primarily as an instrument for budget financing and the secondary market remains thin, although growing.

30. The effective conduct of monetary policy required the introduction of additional new instruments for monetary control. While bank-specific credit ceilings and selective credit allocations were used initially, they have been phased out, except for some politically directed credits (especially for agriculture and energy) in Lithuania. Although indirect instruments of monetary policy have been introduced as part of the reform process, money markets are still underdeveloped and the shift toward market-based monetary policy is by no means complete. With the adoption of a currency board arrangement, Estonia and Lithuania have decided to rule out discretionary monetary policy and to rely on a rule-based approach, according to which the central bank changes its balance sheet through passive sales and purchases of foreign exchange.

31. In Latvia, the main monetary policy instrument has been the purchase and sale of foreign exchange by the Bank of Latvia, with the objective of stabilizing the exchange rate and, since February 1994, of keeping the peg to the SDR. Although there has been some progress toward developing indirect monetary instruments, they have not actively been used for day-to-day and intra-day liquidity management. During the early stages of reform, the instability in money demand resulting from the financial liberalization process itself precluded complete reliance on indirect instruments and credit ceilings were established, as in the other Baltic countries. Subsequently, the lack of a competitive banking system and of developed money and interbank markets reduced the effectiveness of any attempts to use refinance policies to influence the level of banks’ reserves and market interest rates, and thus hampered the workings of the monetary policy transmission process. During the first three years following the introduction of the national currency, the Bank of Latvia adopted a hands-off approach regarding market development and there was little use of market-based instruments; for all practical purposes, since February 1994, the Bank of Latvia attempted to replicate a currency board arrangement, and tried to minimize the expansionary effects of persistent capital inflows by running a very tight credit policy. During 1995 the Bank of Latvia introduced new instruments to smooth out liquidity fluctuations (remunerated term deposits and repurchase agreements) but interest from banks has been limited so far, as they have concentrated their activity in the expanding treasury bill market. The development of the treasury-bill market, in turn, has affected the money multiplier (through changes in banks’ reserve/deposit ratio) and thus significantly affected overall liquidity.

C. Prudential Regulation and Banking Supervision

32. The experience of the Baltic countries confirmed the crucial role of strict prudential regulation and effective supervision in the early stages of financial liberalization. Lax supervisory and licensing policies in the early period of financial reform allowed the proliferation of unsound banks and encouraged imprudent behavior. The licensing process did not include proper investigation of banks’ shareholders and their business plans, increasing the scope for reckless behavior and fraud and thus contributing to the instability of the financial system. While regulations were tightened gradually, not until 1994/95 did the Baltic countries have fully fledged credit institutions laws and regulations to ensure that owners and managers would be “fit and proper”. Furthermore, although regulations similar to the Basle standards have been introduced in the Baltic countries, their implementation has been hampered by inadequate accounting standards.

33. In the Baltic countries, exposure to foreign exchange risk has been high, especially before the currency reform; frequently the bulk of banks’ income was generated through speculation on exchange rate changes or fees from foreign currency operations. In Estonia, for example, in the first half of 1992 these two sources accounted for 95 percent of income, while interest income accounted for less than 5 percent.23 This contrasts sharply to normal banking practices, where around 75 percent of income is generated through interest earnings. Even after the introduction of limits on foreign exchange exposure in 1994, in all three countries banks were frequently found in violation.

34. The sectoral specialization of banks and their close relationship to enterprises led to high concentration exposure in the Baltics. From the prudential point of view, the single most damaging problem affecting the banking sector in the new regime has been the widespread practice of insider and connected lending among commercial banks. As described in Section II, a large proportion of the new banks were established for the explicit purpose of providing credit to shareholders and connected parties. In addition, transfer of ownership of the state banks to their main customers undermined the establishment of arms-length relationships between banks and enterprises. Although insider and connected lending were formally limited in Estonia (January 1994), Latvia (October 1994), and Lithuania (December 1994),24 these limits have been either disregarded or actively circumvented.25

35. A number of steps have been taken recently to build up an effective supervisory capacity in the Baltic countries (the current regulatory framework is summarized in Table 1). Nevertheless, progress achieved in introducing prudential regulations and in monitoring financial institutions has not always been matched by adequate enforcement. Despite visible steps forward, effective banking supervision has been complicated by a number of factors, including the shortage of resources and expertise to deal with a growing number of problem banks. Although Baltic central banks have increasingly devoted resources to this area, staff training on the basic principles of commercial banking in a market economy and on banking supervision necessarily takes time. The staff constraint is compounded by the difficulties in the implementation of prudential rules before banks’ balance sheets have been cleaned up, the quality of loan portfolios improved and reliable accounting and information systems adopted at the enterprise level.27 In addition, bank reform is unavoidably linked to progress in enterprise restructuring and privatization, which in turn necessitates the introduction of adequate bankruptcy laws to promote financial discipline. Slow progress in these areas has hampered efficient financial intermediation and complicated the task of the supervisory authorities.

Table 1.

Prudential Regulations and Deposit Protection

article image

D. Legislation on Bankruptcy and Collateral

36. The absence of effective bankruptcy laws for enterprises and reliable registries for collateral has made it very difficult to enforce loan contracts. In Lithuania, although the 1994 banking law addresses bankruptcy proceedings, the effectiveness of the law is still being tested on the first cases of a number of very small banks. In Latvia, until the adoption of a new credit institutions law at the beginning of 1996, bankruptcy of banks was not treated separately, and the Bank of Latvia did not have the right to appeal to the courts for liquidation. Furthermore, the lack of a developed court system has hindered the bankruptcy process, as the commercial courts have limited capacity to deal with a growing number of bank failures. In Estonia, although a bankruptcy law and adequate procedural arrangements were introduced already in July 1992, banks have not made regular use of this instrument. The process has been found to be time-consuming and frequently enterprise managers have succeeded in stripping the assets of the defaulting enterprise before bankruptcy was declared.

37. The legal framework regulating the use of collateral for bank lending is not fully developed in Latvia nor in Lithuania, reflecting difficulties in registering property titles and enforcing property rights. Although a new law on collateral was introduced toward the end of 1993 in Estonia, the use of collateral did not become widespread immediately. Banks instead have relied increasingly on lending to leasing companies (owned by the larger holding business entities which own the lending banks themselves); this has the advantage that ownership rights are known to lenders from the start.

IV. Financial deepening and financial intermediation

38. Weakness in the financial sector strongly affects the behavior of public and financial institutions, with significant implications for the stability of money demand and the monetary policy transmission mechanism.28 The deepening of the financial intermediation process typically includes an increase in the number of financial institutions, in the ratio of money to GDP, and in the holdings of financial assets by the private sector; at the same time, a decline in the ratio of currency to deposits and in banks’ reserve/deposit ratio are often observed, leading to an increase in the money multiplier. Moreover, the early stages of reform may be accompanied by a period of high real interest rates as well as by a higher rate of growth in credits than in deposits, reflecting the repressed demand for credit in the pre-reform period. As discussed below, financial sector reform in the Baltic countries during the period leading up to the banking crises had been characterized by many of these features. The resulting financial system and its weaknesses were discussed in Section III; the role of banking crises in slowing down or reversing the effects of financial reform and disrupting previous trends in behavioral relationships (such as the money multiplier or velocity) is examined in Section V.

39. Financial liberalization and macroeconomic stabilization resulted in a steady decline in the velocity of broad money29 in Latvia since mid-1992 (when interest rates were liberalized) and in a substantial increase in the holdings of financial assets by the public. In contrast, there has been no significant financial deepening in Estonia since 1992, nor in Lithuania since 1994 (Charts 2 and 3).

CHART 2
CHART 2

THE BALTICS INDICATORS OF FINANCIAL INTERMEDIATION I

Citation: IMF Working Papers 1996, 134; 10.5089/9781451855555.001.A001

CHART 3
CHART 3

THE BALTICS INDICATORS OF FINANCIAL INTERMEDIATION II

Citation: IMF Working Papers 1996, 134; 10.5089/9781451855555.001.A001

40. Despite early and fast financial deregulation in the Baltic countries, the role taken on by the banking sector in financial intermediation has been modest throughout the reform period. Although some increase in intermediation has taken place more recently, progress has been slow and protracted. Banks are far from playing a comparable role to those in industrial countries or even those in the other transition economies of Central and Eastern Europe (Table 2).

Table 2.

Indicators of Financial Intermediation (end-1995)

article image

Pazarbasioglu and van der Vossen (1996)

41. The low level of financial intermediation is foremost reflected in the high share of currency in broad money. While this may be indicative of the low trust the public has in the banking system, an increasing share of currency in broad money may also be typical of cash-based economies during remonetization. The continued predominance of cash may also reflect the slow pace of change in households’ behavior and the desire of some economic agents to keep transactions unrecorded, due to their illegal nature or with a view to avoiding taxes. In Estonia, the share of currency in broad money remained at about 40 percent following the 1992-93 banking crisis and declined slightly starting from mid-1995. Before the banking crises, currency in circulation accounted for almost one-third of the money supply in Latvia and Lithuania, and rose to 40 percent following the crisis in Latvia.30 Improved liability management and the development of a government securities market led to a gradual decline in banks’ reserve-to-deposit ratios. However, the expected increase in the money multiplier during the reform process was dampened by the persistence of high currency holdings (Chart 4).

CHART 4
CHART 4

THE BALTICS MONEY MULTIPLIER

Citation: IMF Working Papers 1996, 134; 10.5089/9781451855555.001.A001

42. Another feature of the financial system in the Baltic countries has been the persistence of the pattern of segmentation observed during the pre-reform period. In the absence of extensive branch networks, retail banking to individuals is limited and most banks cater to medium and large enterprises. In an environment in which banks can mobilize funds through large enterprise deposits and external loans, there is little incentive to raise high-cost small deposits. This is reflected in the high minimum balance requirements for time deposits, low interest rates on demand deposits, and high bank fees charged for transaction services by most banks. Households find it too expensive to effect transactions via the banking system, and since most banks do not make efforts to attract deposits in Estonia and Lithuania, a large portion of household deposits are still held at the respective Savings Banks. By contrast, in Latvia, other large banks, notably Unibank (and Bank Baltija before its collapse), have large branch networks and compete for the retail market.

43. Both credits and deposits rose sharply in the wake of financial deregulation, reflecting the rapid growth of monetary aggregates during the reform. Deposit and credit growth was balanced in Estonia, while in Latvia and Lithuania, as in other countries undergoing financial liberalization, growth in credit (repressed in the pre-reform period) initially exceeded growth in deposits (Chart 5). This imbalance gradually subsided as the reform process evolved

CHART 5
CHART 5

THE BALTICS CREDIT AND DEPOSIT GROWTH

(In percent)

Citation: IMF Working Papers 1996, 134; 10.5089/9781451855555.001.A001

44. Domestic credit as a share of GDP has remained at about 10 percent in Estonia and Lithuania since 1992. This compares to shares of over 100 percent on average for the major industrial countries and up to 50 percent for most Eastern and Central European transition economies. Only Latvia showed a significant increase in this figure, which peaked at about 20 percent in 1995, before it fell back to around 12 percent following the banking crisis. The low ratio of credit to GDP reflects not only the low monetization of these economies, but also the low share of loans in banks’ assets. In 1993, loans on average accounted for less than 50 percent of banking assets. Banks typically held liquid assets, especially in the form of non-interest bearing deposits at the central bank and in correspondent accounts abroad, partly to facilitate the operations of an inefficient payment system, but also because there was little experience in managing liquidity and determining an appropriate liquidity level.

45. The share of private sector credit in GDP in the Baltic countries, albeit low, had been on a rising trend which was reversed following the banking crises. In Estonia, credit to the private sector has been on the rebound since 1993 and it exceeded total domestic credit at end-1995, owing to a surplus position of the general government and the rapid privatization of state enterprises (Chart 6, Table 3). In Latvia, the share of private sector credit in GDP exceeded 16 percent before falling to about 7 percent in the aftermath of the banking crisis; similarly, in Lithuania it rose to around 12 percent of GDP before declining to 8 percent in early 1996. Credit to the household sector rose from about 3 percent of domestic credit in 1992 to 8-15 percent in 1995. This low share may be a reflection of several factors, including banks’ primary focus on the enterprise sector, insufficient mortgage legislation, very recent house privatization, banks’ limited branch networks, and low credit demand by households in view of high interest rates. Although almost all credit resources of banks have been extended to enterprises, bank credit has not been a major source of financing for the enterprise sector. Enterprises instead have relied on retained earnings for financing investment.

CHART 6
CHART 6

THE BALTICS CREDIT TO NON-GOVERNMENT SECTOR

(as a percent of GDP)

Citation: IMF Working Papers 1996, 134; 10.5089/9781451855555.001.A001

Table 3:

Composition of Bank Credit

(As a percent of total bank credit) 1/

article image

It includes credit to government on a net basis.

46. In the early stages of reform, almost all bank credit to enterprises was short-term and trade related, and longer maturities have been offered only gradually. Banks have been reluctant to extend longer term credit because of a perceived lack of good investment opportunities. In a changing and highly volatile macroeconomic environment, given the short credit history of enterprises and inadequate accounting standards, longer term and investment lending was viewed as high-risk. In addition, banks did not have the necessary skills for assessing credit risk, property rights were not well defined, making it difficult to secure adequate collateral, and short term credit, which was routinely rolled over, could be used as a disciplining instrument. The reluctance to extend longer term credit and the significantly lower interest rates charged for such investment projects, may also indicate that banks rationed credit toward preferred customers.

47. Progress in enterprise reform has significantly influenced banks’ behavior in extending longer term credit. Estonia, whose enterprise sector is in a more advanced stage of restructuring, exhibits a wide spectrum of longer maturities, with the share of longer term loans in newly extended loans reaching between 40 and 50 percent in 1995. In Latvia and Lithuania, although the share of short-term loans in banks’ portfolios has declined significantly in the past two years (Chart 7), unsecured short-term credits continue to represent a large part of bank financing. In Latvia, the post-crisis drop in bank credits affected short-term maturities particularly strongly, reflecting the large share of short-term loans in the portfolio of Bank Baltija and other large failed banks.

CHART 7
CHART 7

THE BALTICS SHORT TERM CREDIT 1/

(as percent of total credit)

Citation: IMF Working Papers 1996, 134; 10.5089/9781451855555.001.A001

1/ Up to one year.

48. The sluggish response of nominal interest rates to a decline in inflation is not uncommon, in particular when interest rates are liberalized rapidly and early in the reform process, before significant progress has been made on macroeconomic stabilization and enterprise reform and while effective banking supervision is still not in place.31 The persistence of high nominal deposit and lending rates and large interest rate margins after interest rates were deregulated—despite the significant decline in inflation—contributed to financial instability and impacted negatively on investment and growth. This was especially relevant for Latvia and Lithuania. The high spreads between deposit and lending rates in the Baltic countries during the post-independence reform period reflect a combination of high default premiums, banks’ attempts to recapitalize, high operating costs, and large stocks of non-performing assets in banks’ portfolios. The adoption of rules for loan-loss provisioning also contributed to high margins. Real lending rates have been mostly negative in Estonia since mid-1993 but were positive and high in Lithuania and, especially, in Latvia during the period 1993-94 when inflation was declining; in Lithuania, lending rates became briefly negative in 1995 before again becoming positive (Chart 8). This variability may have reflected uncertainty under the changing financial and economic environment; the high real lending rates reflected both the structural weaknesses of the financial and enterprise sectors—translated into high credit risk—and the exceptional opportunities to finance highly profitable short-term transactions, such as trade, created by the existence of controlled prices on certain commodities in countries of the former Soviet Union; it may also have reflected the low credibility of the monetary arrangement.32 In Latvia, furthermore, significant inflows of capital in the wake of interest rate and capital account liberalization encouraged risky lending by banks. The high rates charged by a few large banks heavily involved in the lucrative transit trade pushed interest rates upwards and resulted in significant misallocation of resources, with the more efficient borrowers crowded out by the riskier ones. Therefore, a combination of high real lending rates, high interest margins, and declining loan-deposit ratios was observed in 1993-94. Interest rates have declined gradually since then but lending rates continue to be high in real terms, reflecting persistence of high credit risk.

CHART 8
CHART 8

THE BALTICS INTEREST RATES

Citation: IMF Working Papers 1996, 134; 10.5089/9781451855555.001.A001

1/ The spread is defined as the difference between the nominal lending rate and the nominal deposit rate.2/ Real rates have been deflated by the six-month forward Inflation rate.
CHART 9
CHART 9

THE BALTICS FOREIGN CURRENCY DEPOSITS

(as a percent of total deposits)

Citation: IMF Working Papers 1996, 134; 10.5089/9781451855555.001.A001

V. Banking Problems, Crises and Policy Responses

A. Origins of Problems and Crises

49. As a result of insufficient progress in structural reform and unsupervised growth in the banking sector, a financial system vulnerable to market failure and to exogenous shocks emerged from the monobank system in the Baltic countries. Financial deregulation increased the potential exposure of financial and nonfinancial institutions to risk and, in the absence of an effective regulatory framework, set the stage for the banking crises which emerged in the subsequent years. Unlike in many other countries facing banking problems, macroeconomic shocks did not play a central role in the banking crises of the Baltic countries. The crises were mainly determined by the structural deficiencies of a fast-growing banking sector, compounded by the lack of markets, legal and institutional infrastructure, know-how, and sound investment opportunities reflecting the slower growth in the enterprise sector. Although prudential regulations had been gradually introduced, standards were not tight enough and enforcement was weak. These factors together generated a state of latent insolvency in the banking system, which inevitably burst out in the form of financial crises. The timing and magnitude of the crises were to a large extent determined by the speed and sequencing of reform and the capacity to enforce discipline in the banking and enterprise sectors, as well as the depth of financial intermediation in these countries.33

50. Liberal licensing requirements allowed the proliferation of commercial banks without fit and proper owners, appropriate capital, banking skills, accounting methods and internal control procedures. The lack of market discipline and effective banking supervision in the new, deregulated environment, combined with the moral hazard associated with the generalized perception of implicit government guarantees, increased the scope for risk-taking behavior, unsound lending policies and fraud. At the same time, slow progress in privatization and enterprise reform narrowed sound profitable investment opportunities for the banking sector. These in turn led to the build-up of weak portfolios and increasing losses, which eventually translated into banking crises. In both Latvia and Lithuania, the crises were caused by the bad management, poor credit policies and fraud (including false accounting and reporting) of private banks, of the “too-big-to-fail” type.34 In the case of Latvia, these problems were exacerbated by the persistence of large inflows of foreign exchange into the banking system during 1993-94, which, given the lax supervisory standards, further encouraged risk-taking by banks.35 In Latvia and Lithuania, the attempt to deal with the weak portfolio problem by requiring banks to provision against losses helped uncover the fragile situation of many banks.36 In Estonia, the 1992 crisis was mainly due to external circumstances (the freezing of deposits abroad), but also reflected poor management. The Social Bank crisis of 1994 was caused by mismanagement and fraud.

B. Triggers

Estonia

51. Banking problems surfaced in November 1992 when the state-owned North Estonian Bank (NEB), the Union Baltic Bank (UBB), and the Tartu Commercial Bank (TCB) exhibited serious liquidity problems and delayed payments by three weeks. The NEB and the UBB had lost access in early 1992 to US$76 million deposited at the Moscow Vneshekonombank, but were able to postpone the onset of serious problems by attracting new deposits. The TCB ran into difficulties when one major loan of EEK 73 million was not repaid. The situation was aggravated by losses on foreign exchange operations and nonperforming loans extended to connected parties.37

52. The second crisis in Estonia took place in early 1994, when the government reduced the level of its deposits from the Social Bank. Starting in February 1994, when budgetary deposits alone accounted for one third of deposits in the Social Bank, the government withdrew over 80 percent of its deposits from that bank. These withdrawals, which were heavily publicized, precipitated the withdrawal of local government deposits and led to serious liquidity problems. Initially, these were addressed by the bank through the reduction of excess reserves and foreign assets. However, despite the very short maturity structure of its loan portfolio, the bank was unable to reduce its outstanding loans to enterprises. It became apparent that a considerable part had been rolled over automatically and had to be considered non-performing. At the same time, loans to shareholders and their shell companies accounted for roughly one third of all loans and, when it became unlikely that the bank would be bailed out, shareholders appear to have transferred assets abroad. The bank, which accounted for 15 percent of total banking system assets at end-1993, was finally closed in March 1995, after a failed rescue attempt by the Bank of Estonia (Box 3).

Estonia: Social Bank Crisis-Summary of Developments

In March 1994, the government acted to lower the level of its deposits from the Social Bank, Estonia’s second largest bank, reducing these deposits by EEK 255 million or 83 percent by end-July. In addition’ other government entities and state enterprises started to withdraw deposits. Already in early spring after an on-site inspection, the Bank of Estonia had recognized the potential for problems at the Social Bank. The Bank of Estonia requested a change in the management of the Social Bank and action to limit loans to shareholders. The Bank of Estonia concluded that the shareholders were unduly influencing management, were involved in making credit decisions, and received a large portion of the loan portfolio. In addition, a substantial part of the loan portfolio had repeatedly been rolled over and some loans had to be written off immediately since these loans had been extended to companies which had been declared bankrupt already. In response to the withdrawal of deposits, the Social Bank reduced its reserves held at the Bank of Estonia, reduced its interbank deposits, liquidated its foreign assets and increased borrowing from the interbank money market, but it failed to reduce its loan portfolio. In fact, long term loans were increased from EEK 99 million in January to EEK 170 million in July. Closer scrutiny of the loan portfolio also revealed that the amount of non-performing loans considerably exceeded the amount reported to the Bank of Estonia and indicated the existence of a solvency problem.

In July, the Social Bank got permission from the Bank of Estonia to draw down required reserves to zero without penalty. On July 15, the Bank of Estonia ordered the bank to stop issuing new shares, limited its loan portfolio as a share of total assets, and required that funds from repaid loans be held in liquid assets. To alleviate the solvency problem the Bank of Estonia suggested that the Social Bank divest some of its branches. While the rolling over of loans had been identified as problematic, the Bank of Estonia was unable to impose sanctions as the law on commercial banks at that time did not provide for this possibility. At end-July, the Social Bank defaulted on a EEK 75 million interbank overnight loan from North Estonian Bank and a EEK 30 million loan from the Bank for Industry and Construction. The inability of the Social Bank to obtain loans in the overnight interbank money market and a backlog of payment orders required action to be taken by the Bank of Estonia. Initially the Bank of Estonia refrained from declaring a moratorium since its experience had been that banks under moratorium were unable to attract new deposits. Instead the Bank of Estonia opted for another change in management and ownership. On August 2, over 50 percent of shares were sold for a nominal amount to the sharerholders of Arengubank, a very small commercial bank, and its chairperson assumed the leadership at the Social Bank. The Bank of Estonia announced its support for the new management and stood ready to provide emergency liquidity assistance. In addition, several commercial banks promised liquidity support and it was expected that the liquidity problem could be addressed by selling the Social Bank’s branches. On August 4, the Bank of Estonia announced it had made a deposit of EEK 70 million at the Social Bank as emergency liquidity assistance (collateralized by branches of the Social Bank). In addition, the commercial banks promised liquidity support of EEK 80 million.

After the sale of five of its branches on August 15, 1994 and the imposition of a moratorium on the Social Bank, the new balance sheet of the Social Bank was established with assets almost halved. On September 21, the Bank of Estonia decided against the liquidation option and the continuation of the moratorium, and for reopening the Social Bank. The Bank of Estonia expected that a higher proportion of bad loans could be recovered through a working bank than through the bankruptcy court. The Bank of Estonia was also concerned about pressures that would be brought to bear from public sector entities which would lose money following the closure of the bank. The moratorium on the Social Bank was lifted on September 26, 1994. Equipped with the remaining emergency liquidity deposit of the Bank of Estonia and a new liquidity credit, the Social Bank faced heavy deposit withdrawals during the first week after the end of the moratorium. In the following week the situation at Social Bank partially stabilized, including through further liquidity support by the BOE. Total support by the BOE eventually reached 6 percent of reserve money. In March 1995, the BOE closed the bank and concluded agreements to sell part of the bank’s assets and to begin work on turning the remainder into a loan recovery agency.

Latvia

53. The growing distress in the Latvian banking system turned into a full-fledged crisis in April 1995, after the failure of some banks to complete audited reports for 1994 and the subsequent closure of Latvia’s largest commercial bank, Bank Baltija (Box 4). The published audited reports exposed the fragile situation of the banking system, as two-thirds of the audited banks recorded losses in 1994 and lacked an adequate capital base to provide a cushion against these losses. The losses were generally a result of poor lending decisions and disregard for prudential regulations. Particularly important was banks’ excessive lending to insiders, in several cases aggravated by large open positions in foreign exchange, with the expectation of an eventual depreciation of the lats. Furthermore, as a result of the liberalization of energy prices in Russia and the real appreciation of the Russian ruble, profit margins on trade financing had started to narrow from the end of 1993, exposing banks highly dependent on high risk short-term trade financing.

Latvia: the Collapse of Bank Baltija

Bank Baltija (BB) was founded in 1989 and, before collapsing in May 1995, became the largest Latvian commercial bank, accounting for about 16 percent of the total assets and 30 percent of total deposits of the banking system. Since mid-1993, when BB bought several of the former commercial branches of the Bank of Latvia, the bank had adopted a very aggressive policy to increase its market share, by offering higher interest rates and lower minimum deposits than its competitors. At the same time, the bank had concentrated its lending on high-risk, high yield, trade-financing loans and kept a large open position in foreign currency, betting on a devaluation of the lats. As a result of unsound lending policies, lack of internal controls, neglect for prudential regulations—in particular, significant insider lending, as well as plain fraud, its financial situation deteriorated rapidly and by early 1995 it became insolvent. The prolonged state of distress increased the incentive for further risk-taking, while the bank’s size (”too-big-to-fair) encouraged expectations of a government bail-out. Despite its precarious situation, in March 1995 the bank announced plans to merge with two other large insolvent banks. Liquidity problems surfaced in the subsequent months and an emergency credit from the Bank of Latvia was obtained in May in an effort to keep the bank afloat. Since this was not enough to meet its short-term obligations, the bank requested another Bank of Latvia loan, which was refused. On May 23, 1995, the operations of Bank Baltija were suspended. The government took full control of the bank and started to consider a rescue program, including partial compensation of household deposits.

By this time, however, most of BB’s assets had been stripped, leaving negative net worth of around US$ 400 million, equivalent to 8 percent of GOP. A substantial part of its loan portfolio (about 60 percent) was transferred to third parties at distress prices just prior to closure. Deposit liabilities accounted for around LVL 138 million (excluding bank deposits), with household deposits accounting for 75 percent of the total. In addition to bad banking practices, fraudulent and criminal activities were uncovered, notably lending to its own shareholders through the cover of unidentified off-shore companies. Both the main shareholder and the president of the bank have been charged with fraud and economic sabotage.

Following the authorities’ intervention, the Bank of Latvia assumed the responsibility for running the bank, but handed it over to a court-appointed administrator after insolvency was declared by the Commercial Court at the end of June 1995. The process of liquidation of the bank has moved very slowly, reflecting the complex connections surrounding the bank and the deficiencies of the court system. Some pressure groups have consistently argued for rehabilitation as the only way to recover depositors’ money, despite the general belief that the prospects for recovery of assets are virtually nonexistent. In December 1995 the Commercial Court declared BB bankrupt and the Bank of Latvia revoked its license; this decision was appealed by creditors and annulled by the Riga District Court in February 1996, on procedural grounds. BB was again declared bankrupt by this higher court in April 1996, but subsequently several appeals have been lodged to the Supreme Court on behalf of BB’s creditors. Until the Supreme Court decides the fate of the bank, the court-appointed liquidators cannot begin the task of tracing BB’s assets.

54. After steady growth for two consecutive years, bank deposits declined by two percent in February 1995, in the wake of a scandal involving one of the medium-sized commercial banks.38 The failure of several large banks (including Bank Baltija) to submit their audited 1994 reports by the March 31 deadline further undermined confidence in the banking system, leading to capital flight and additional deposit withdrawals. The collapse of Bank Baltija, several weeks later, came at a time when uncertainty about the forthcoming Parliamentary elections and fiscal policy promoted further currency substitution and a switch back to cash.

Lithuania

55. The deteriorating situation in the Lithuanian banking sector culminated into a crisis following the publication of on-site inspection results, in December 1995, for the largest and third largest private banks, Innovation Bank and Litimpeks Bank, and the subsequent suspension of these banks’ operations. During 1995, bank capital had diminished further from an already low base, losses increased substantially, and a growing number of small banks filed for bankruptcy. More importantly, it became unclear whether the government would show the same resolve in dealing with the larger banks that it had shown with problematic small banks. In the summer of 1995, a medium-sized bank was supported heavily by the authorities when it faced liquidity problems as a result of deposit withdrawals which followed rumors about the bank’s insolvency. In the fall of 1995, the government gave liquidity support to another medium-sized bank, which faced problems again in December, at which time it had its operations partially suspended and was put under administratorship.39

56. The large increase of recorded bank losses during 1995 was partially due to the introduction of loan loss provisions, which rendered losses explicit, although such provisioning was partial and likely underestimated the volume of non-performing loans. Banks’ weak capital base did not afford protection against losses, which in addition to reasons common to the three Baltic countries, were often the result of political direction in lending. Connected lending and political involvement played a big role in the problems facing Innovation Bank, with which the problematic energy sector was closely involved, both as shareholder and as borrower. Following protracted negotiations on a merger in the fall of 1995 between Innovation Bank and Litimpeks Bank, on-site inspections ordered by the Bank of Lithuania revealed in December 1995 that both banks were insolvent, and deposit withdrawals followed; subsequently, the Bank of Lithuania put the two banks under moratorium: their activities were confined to collecting loans and other assets, and their deposits were blocked.40 Both the closure of the two banks and the way with which the actions were carried out were the focus of ample criticism by the political opposition, the press, and the public. Statements followed in the press questioning the solvency of the three state-controlled banks, which comprised two-thirds of total commercial bank deposits. Under the circumstances, a systemic crisis became possible.

57. Immediately following the closure of the two banks, widespread and steady bank deposit withdrawals took place from the banks that continued to operate and large foreign exchange outflows occurred through the currency board against a background of political uncertainty. Commercial bank reserves fell below the minimum requirement and the interbank market virtually ceased to operate. Interest rates on government securities doubled from 20 percent in mid-November 1995 to 40 percent in late February 1996.

C. Policy Responses

Estonia

58. When the first banking problems surfaced in 1992, the Bank of Estonia provided some liquidity support to the NEB, since liquidity difficulties were initially perceived to be temporary. On November 17, 1992, the NEB, the UBB, and TCB41 were closed when it became clear that there were more serious underlying problems which could not be addressed through liquidity injections. The authorities decided to liquidate the TCB, since its problems had been caused by imprudent lending decisions. Depositors were partially paid from liquidation proceeds. However, losses were substantial as the government decided not to bail out creditors. At the same time, a rescue package was designed for the recapitalization of the NEB and the UBB as their problems were attributed to political circumstances deemed to have been out of the banks’ control. Both banks were merged and all assets corresponding to the frozen deposits at Vneshekonombank were taken off the balance sheet. On the liability side, shareholder claims were written off, and the government and Bank of Estonia injected new capital in the order of EEK 400 million (about US$31 million). Further support in the form of liquidity loans was provided by the Bank of Estonia.42

59. During the Social Bank crisis in 1994, despite initial emergency liquidity support by the Bank of Estonia, the bank’s liquidity problems escalated and, in August, the bank was put under a moratorium. Although there were clear signs of insolvency and no fully fledged plan for restructuring, the bank was reopened in September with a public commitment by the Bank of Estonia for further liquidity support and a guarantee to depositors. Nevertheless, almost 50 percent of the remaining deposits were withdrawn within a short time, and the bank’s financial position deteriorated further. When the bank was finally closed, performing assets and deposit liabilities of individuals and enterprises were transferred to the North Estonian Bank, while the remaining balance sheet was converted into a loan recovery agency in March 1995 (Box 5). On the asset side of the balance sheet of the loan recovery agency there were mainly nonperforming loans, while on the liability side, there where claims of the government and the Bank of Estonia. Support by the Bank of Estonia had reached 6 percent of base money. With the exception of shareholders, all other creditors were bailed out with losses expected to be shared between the government and the Bank of Estonia.

Estonia: Loan Collection Agency

Estonia provides the only case in the Baltic countries of experience with employing a centralized loan workout agency (Lithuania established such an agency in the summer of 1996). The Bank of Estonia (BOE) unwound the remaining balance sheet of the Social Bank by dividing the Social Bank’s assets into what were considered performing and non-performing assets. By thus making the value of assets more transparent, the total valuation of all assets by risk-averse agents could be maximized. Creditors were partially compensated through the sale of performing assets and the government and the BOE obtained a contingent claim on any net proceeds recovered from the non-performing assets.

A loan collection agency was established to recover as much as possible of the non-performing assets. However, as has been the experience in other countries, the benefits from the loan collectionagency turned out to be considerably lower than expected, as centralized recovery agencies are typically not able to manage and work out loans as effectively as individual banks which have more detailed knowledge of their clients. There was also an added incentive problem. Since management and workers were paid fixed salaries, there was little incentive to recover more assets than was necessary to cover administrative costs.

Some banks pressed to transfer their non-performing assets to the loan collection agency, while writing down their capital by an equivalent amount. In return, these banks asked to receive a contingent claim on the now centralized loan collection agency (”bad loans bank”). However, the authorities decided not to go ahead with this scheme since it would be sending a wrong signal to both domestic and foreign business and financial communities about the health of the banking system. This step could also have been interpreted as an admission of systemic problems in the banking system, which could thwart the access of banks to international capital markets. Moreover, the establishment of such an institution could very easily have become a vehicle for increasing pressures toward government recapitalization of the participating banks or a bailout of the loan collection agency itself.

It was recognized that even if the banks themselves saw merit in a centralized approach to collecting their non-performing loans and they were to organize this institution on their own initiative, possible benefits would remain uncertain. A centralized recovery agency might well yield economies of scale, since the claims of several banks are apparently held on the same clients, and therefore the recovery agency might operate more cost-effectively than individual banks. Since banks would be expected to bear the operating costs, there would also be no fiscal impact. As pointed out above, however, the benefits from the establishment of loan collection agencies often turn out lower than expected. Moreover, the costs of establishing a centralized agency are considerable in the first place since very specialized skills would be required.

Latvia

60. Until the Baltija collapse, the Bank of Latvia had dealt effectively with small problem banks, with a view to achieving an orderly consolidation of the banking industry. The licenses of 8 commercial banks were revoked in 1994; most of these were small institutions, with negligible impact on the system. However, when problems with larger and influential banks surfaced in early 1995, the Latvian authorities were not prepared to handle the situation in such a way as to avoid a large-scale crisis. The combination of complex political constraints, poor coordination between the monetary and fiscal authorities and the absence of an adequate legal framework to deal with major bank failures resulted in delays and helped to undermine confidence in the banking system. In the case of Bank Baltija, there were also pressures to delay action despite previous knowledge of the bank’s precarious financial situation.43 In addition, the problem in Bank Baltija included alleged criminal activities and close political connections. The delay in intervention, until after the bank became insolvent, allowed enough time for asset-stripping and thus increased sharply the resolution costs of the crisis. On the other hand, given the extent of the fraud involved, it is hard to conceive of supervisory measures capable of preventing the bank’s debacle at that late stage.

61. To restore confidence in the banking system, given the systemic threat posed by the Baltija collapse, the government promised to compensate depositors of failed banks in the amount of LVL 500 (around US$1,000) per depositor over a three-year period. In the event, only a minimal amount of compensation was extended during 1995 on the basis of recovered assets. A package of banking legislation, including laws on deposit insurance, on the establishment of a bank rehabilitation agency, and a new law on credit institutions, was submitted to Parliament after the banking crisis to strengthen the regulatory framework for banking and provide the legal basis for deposit compensation and liquidation of banks. After heated debate, only the law on credit institutions was adopted, as Parliament decided that more time was needed to consider the introduction of deposit insurance and a rehabilitation agency.

62. After the Baltija collapse, the Bank of Latvia took a number of steps to strengthen banking supervision, including a more aggressive approach to monitoring problem banks and enforcing prudential regulations.44 The overall approach was to apply various types of restrictions to the operations of banks which failed to comply with prudential regulations and ultimately revoke their licenses, thus working toward the build-up of a smaller and sounder banking sector. In June 1995, the number of commercial banks eligible to accept household deposits was limited to 16. These banks, which would form the “core” of the revamped banking system, were required to undergo more frequent external audits and publish their balance sheets on a quarterly basis. Increased transparency contributed to build up market discipline and increase confidence. Other measures taken to strengthen the regulatory framework included the tightening of key prudential regulations; an increase in the minimum capital requirement from LVL 100,000 (about US$200,000) to LVL 1 million (about US$2 million); increased monitoring of problem banks by the Bank of Latvia with the assistance of external auditors; and changes in the organization of the banking supervision department at the Bank of Latvia, including an increase in the number of staff and improvements in the use of available information (an early-warning system was established at the end of 1995).

63. Despite the various possibilities initially contemplated for a comprehensive restructuring program and compensation of depositors in failed banks, in retrospect, the Latvian solution to system risk was based on enhanced regulation and supervision, and increased transparency (to instill market discipline), with minimum lender of last resort support from the central bank, no deposit insurance and no bail-outs.45

Lithuania

64. With the exception of delays caused by the inadequacies of the legal framework and some bottlenecks resulting from the short supply of trained personnel, the Bank of Lithuania dealt with the cases of 14 small problem banks before the end of 1995 by promptly halting or restricting their operations.46 As problems surfaced in larger banks, however, in the third quarter of 1995, the authorities were reluctant to let them fail, and found ways, in the presence of the currency board arrangement—which effectively precluded the extension of credit by the central bank—to support these banks, including with government credit, government purchases of bank shares, and the transfer of government deposits.

65. The suspension of the operations of Innovation Bank and Litimpeks Bank on December 20, 1995 was not followed up by the implementation of measures to strengthen the banking system. Instead, the ensuing political turmoil, a number of contradictory laws passed by Parliament, and the inaction of the authorities on the affected banks contributed to further undermining the public’s confidence in the financial system. In particular, Parliament passed a law guaranteeing full protection of creditors in the two closed banks, while the deposit protection law, which had been passed in early December 1995, provided for limited insurance (capped at LTL 4,000, or US$1,000, per depositor for litas deposits and excluding foreign currency deposits).47 A law providing for the extension of government guarantees of one year’s duration for interbank loans up to a total of LTL 300 million (US$75 million) was ineffective in restoring interbank lending.

66. In view of the systemic nature of the liquidity squeeze on banks and the danger of further eroding their capital base, the Bank of Lithuania waived penalties on reserve requirement shortfalls. As a result, the payments system was not affected in a detrimental way. Waiving of penalties for not paying taxes for enterprises with frozen bank deposits and the accumulation of short term arrears, including for energy consumption, helped weather the liquidity crunch.

67. The already limited technical capability of the Bank of Lithuania was further reduced as a number of key officials, including the head and deputy head of the banking supervision department and the head of the international department, resigned. Facing a loss of parliamentary support, the Bank of Lithuania Governor resigned. Under criticism for his role in the handling of the closure of the two banks, the Prime Minister’s resignation followed soon after. The new Bank of Lithuania management introduced significant changes into the organization of the Bank of Lithuania with a view to better preparing it in exercising the functions of a market economy central bank.

68. The authorities delayed acting on a plan to deal with the problem banks. Despite loss of capital and allegations of fraud, the affected banks mounted a public campaign to prevent liquidation and to postpone reorganization for several months. Although the three closed banks were insolvent, they were allowed gradually in the spring of 1996 to resume certain operations, excluding lending and deposit taking. Each of the banks made repeated attempts to have their shareholders’ restructuring plans approved by the Bank of Lithuania, although the plans were judged not to be viable. In the event: (a) Innovation Bank had its share capital reduced and was taken over and recapitalized by the government, which became majority owner; (b) Litimpeks Bank reopened for full business in June 1996 after some of its deposits were converted to capital and the Bank of Lithuania was satisfied that it met all prudential regulations; (c) Vakaru Bank entered liquidation procedures; and (d) Aura Bank was restructured into an asset management company which is to act as a centralized loan collection agency. Time will show the degree of effectiveness of a centralized recovery agency in Lithuania and whether the rate of recovery will be higher than it was in Estonia (Box 5).

D. Effects of Banking Crises

Estonia

69. Although deposit liabilities amounting to 40 percent of broad money were affected in the 1992 crisis, the crisis did not spill over to the remainder of the banking system. Contagion was limited since banks did not operate an interbank money market and liquidity problems did not spread through the payments system. In addition, most depositors were bailed out. While there was a general loss of confidence in the banking system, which was evidenced by an increasing cash-to-deposit ratio, there appear to have been no significant negative repercussions for the real economy. Indeed, after economic decline during 1991-92, Estonia’s output started to rebound in mid-1993.48 However, the crisis may have arrested temporarily the development of deeper financial intermediation.

70. In the 1994 Social Bank crisis, the effects were isolated and limited to the concerned bank. None of the liquidity problems spread through the payments system or the interbank money market. Remarkably, during the crisis, lending to the failing bank through the interbank market increased dramatically. Banks, however, charged higher interest rates and required collateral for their loans. Depositors shifted funds to healthy banks and no runs on other banks were observed. There is no evidence that the Social Bank crisis had any measurable negative effects on output.

Latvia

71. The banking crisis impacted strongly on the public’s confidence in the banking sector and dented international confidence in the country. The failure of Baltija and other large banks in the spring of 1995 led to the withdrawal of foreign funds invested in Latvia, a sharp decline in the demand for money, substantial portfolio shifts, and a severe contraction in the banking system’s role in intermediation. The low level of activity in the interbank market reduced the probability of contagion and no massive bank runs took place. Nevertheless, the crisis significantly altered the behavior of monetary and credit aggregates. Total deposits declined by 5.4 percent during the first half of 1995, mostly reflecting the withdrawal of deposits by households and state enterprises. If deposits blocked in insolvent banks are excluded, total deposits fell by 35 percent during that period.49 Households shifted from term-deposits to demand deposits and cash, while enterprises switched to foreign exchange-denominated deposits. Although initially deposit withdrawals were restricted to those banks which failed to submit their financial statements, following the Baltija collapse several banks faced large withdrawals of deposits caused by the spreading of malicious rumors about their financial health. Other banks, perceived to be stronger, were able to capture part of these deposits (i.e., run to quality). Although deposit rates had become significantly more negative in real terms, banks were reluctant to raise them in order to attract more deposits because they feared, after the experience with Bank Baltija, that higher deposit rates would be associated by the public with higher risk. Interbank market activity declined sharply, reflecting lack of trust among banks.

72. After experiencing a liquidity squeeze and negative reserve positions for a few months, the banking system started to accumulate excess reserves again in June 1995. The group of “core” banks (representing two-thirds of total assets) was able to generate profits at the end of the semester (although some of these banks reportedly had troubles) and again at the end of the year, but by then their number was reduced to 11. On the other hand, the banking crisis and tightened banking supervision led banks to become more cautious about their lending, while high yields on Treasury bills, reflecting the government’s growing borrowing requirement, encouraged banks to increasingly invest in these securities.50 These developments, together with the downsizing of the banking system, resulted in a sharp decline in credit to the private sector (from 16 percent to 8 percent of GDP) in 1995.

73. The banking crisis led to a reversal of the post-liberalization trend in various financial indicators (discussed in Section IV), exacerbating the instability in the money multiplier and in the money demand/supply process typical of transition processes. The Bank of Latvia’s intervention in the foreign exchange market succeeded in arresting capital flight and in restoring a degree of stability to financial markets, at the expense of a substantial loss of international reserves. As a result, monetary aggregates started to grow again by mid-year. However, some elements of a financially repressed system reemerged, as the crisis encouraged dollarization, increased use of cash, and the growth of the underground economy.51 It also weakened revenue collection, and thus exacerbated the government’s own financial crisis. The deflationary impact of the banking crisis was less severe than suggested by the sharp contraction in domestic liquidity, in view of the limited role of bank intermediation and the high degree of dollarization in the Latvian economy.

74. The banking crisis also resulted in political casualties. First, the Minister of Finance resigned a few weeks after the Baltija collapse. Second, the Governor of the central bank had to face a parliamentary no-confidence vote. Although the motion failed to pass in Parliament, it was repeatedly rescheduled and weakened the central bank’s independence at that time. Finally, the fiscal crisis and the banking crisis reinforced each other and created an overall atmosphere of uncertainty and lack of confidence in the government and contributed to the perpetuation of a “non-payment culture”.

75. One year after the banking crisis, the financial sector had achieved relative stability as a result of tighter supervision and enhanced enforcement. The number of problem banks declined and compliance with prudential regulations improved. The publication of banks’ quarterly accounts increased transparency and thus also contributed to a gradual build-up of confidence in the banking system. Only five banks failed to meet the increased minimum capital requirement by the April 1, 1996 deadline, and had their activities suspended. In contrast to the previous year, the audited reports for 1995 were submitted on time by all but three banks, revealing aftertax profits of about LVL 1 million (about US$2 million) for the banking system (and LVL 5 million or nearly 1 percent of assets for the “core” banks).

76. Notwithstanding notable progress in improving the regulatory framework and in developing an effective supervisory capacity, financial fragility remains and effective financial sector reform is still a long way ahead. The profits posted by commercial banks during 1995 were to a large extent the result of their increased holdings of Treasury bills. As a result, the private sector was severely crowded out of credit markets. Despite a significant tightening of fiscal policy by the new government, credit to the private sector continued to fall through the first quarter of 1996. At the same time, the demand for money, partly as a result of increased disintermediation, remained low until April 1996 as nonbanks also increased their holdings of treasury bills.

Lithuania

77. It might be too early to assess the full macroeconomic impact of the banking crisis in Lithuania, but its effects so far appear to have been rather limited; growth and incomes seem to have been only slightly affected. The acute stage of the crisis lasted between 10 and 12 weeks, as already in March 1996 interest rates on government securities had declined and the direction of capital flows was reversed with the beginning of a few large foreign investors’ participation in the domestic treasury bill market. Deposit withdrawals stopped two months after the closure of banks, as households came to believe that the government would not close the state controlled banks, and enterprises opened new accounts in banks known to be better capitalized.

78. The financial system continued to function even at low liquidity levels. The temporary waiving of reserve requirements worked as a stabilizer under the currency board arrangement in that it offset the increase in the cash-to-deposit ratio, which reflected the withdrawal of deposits, leaving broad money largely unchanged.

79. In early 1996, a comprehensive banking sector restructuring plan was developed by the authorities, with help from the World Bank and the IMF, but its implementation was delayed repeatedly. Despite the slow follow-up action by the authorities, the banking crisis had limited contagion effects and effects on the economy as a whole. Rapidly rising interest rates immediately following the crisis seem not to have acted as an indicator of high risk, as new foreign investors, who entered the government securities market driving interest rates back down, remained in the Lithuanian market.

80. As in Latvia, the banking crisis led to a reversal of the post-liberalization trend for various financial indicators, exacerbating the instability in the demand for money typical of the transition process. By the end of the third quarter of 1996, the currency-to-deposit ratio, which followed an upward trend after the crisis, had remained high as economic agents continued to rely heavily on the use of cash for transactions and deposits remained roughly unchanged since the beginning of the year. Banks avoided granting new credits and invested in treasury bills, bringing the money multiplier down to below 1993 levels. The deflationary impact of the banking crisis was not as severe as the reversal of financial deepening suggests and preliminary indications are that macroeconomic activity in 1996 seems to have been only slightly affected The situation in the financial sector remains fragile, however.

VI. Concluding Remarks

81. As part of the overall transition process to a market economy, the Baltic countries introduced major reforms in the financial sector, involving, in addition to the liberalization of a repressed financial system, the build-up of institutions, expertise, markets, instruments, and changes in the legal and regulatory framework. Liberalization of the existing financial system was accomplished early on in the post-independence reform period: (a) subsidized and directed credits were almost completely phased out; (b) controls on interest rates were removed; (c) current accounts became convertible; and (d) almost all restrictions were eliminated from capital movements. However, the build-up of an effective market-based financial system to replace the old command system, although it began concurrently with liberalization, is still not complete. Toward this end, the Baltic countries have worked on several fronts: (a) they have taken steps to develop a competitive and sound banking system, including the establishment of a two-tier banking system; the incorporation and reorganization of state banks; and a move toward universal banking by mobilizing savings for all banks and developing skills for credit risk evaluation and portfolio selection; (b) they have built up considerable capacity for banking supervision so as to establish an effective regulatory framework; (c) they are in the process of developing money, securities and capital markets in order to facilitate the operations of the banking system and enhance monetary control; and (d) they are developing efficient payments systems in which financial transactions can be effected quickly and safely. Additional progress in these areas, however, requires, to a large extent, an acceleration of structural reform in other sectors of the economy, including enterprise restructuring, the development of markets for land and real estate, and improvements in the legal and regulatory framework, notably regarding bankruptcy procedures, the transfer of ownership, and securitized lending.

82. Recent experience in the Baltic countries has confirmed that banking problems and crises strongly influence the pace of financial system reform. In particular, banking crises lead to disruptions in the trends of various financial indicators and thus represent a step backwards on the road to financial sector development, which may take some time to be corrected. On the other hand, the reform process itself can contribute to the vulnerability of the banking and enterprise sectors, and thus set the stage for banking crises.

83. The banking crises which took place in Latvia and Lithuania in 1995/96 have demonstrated that failure to recognize the links between the various aspects of structural reform and to act decisively to address structural bottlenecks could compromise the effectiveness of stabilization policies. The slow pace of structural reform and lack of progress in enterprise restructuring created a weak operating environment for fast-growing banking systems. Furthermore, the interlocking ownership/lending pattern between banks and enterprises, often seen in transition economies, led to fragility in both sectors. As a result of the influence of enterprises, banks’ decisions were often taken which were not based on profit maximizing criteria; on the other hand, access to credit was often difficult for many enterprises, especially new and/or small ones, as banks tended to concentrate their lending to their shareholders and related risky business. Against this background, failures in internal governance, market discipline and external governance (regulation and supervision) translated into banking crises. In the absence of effective banking supervision, speculative capital inflows aggravated banking problems.

84. In view of the long-term nature of structural reform and of the unchecked growth of the banking sectors in the three Baltic countries, the development of banking problems in recent years was to a large extent inevitable. The emergence of crises, their magnitude and overall impact depended significantly on the way they were handled, that is to say, the policy response. The Baltic experience suggests that minimizing bail-outs is the most effective way of addressing systemic risk, as it (a) reduces moral hazard by helping to contain the “flow”problem, (b) contributes to boost confidence in the fledgling banking system, and (c) avoids the fiscal impact of fully-fledged recapitalization. Recapitalization efforts in an environment of close ties between banks, enterprises and the government are susceptible to moral hazard problems and might not address fundamental issues of internal and external governance of banks. Experience has shown that bank restructuring was most successful when insolvent banks were swiftly liquidated and prudential regulations were strictly enforced. The no-bail-out approach promotes the consolidation of banking sectors, which is highly desirable in view of the limited role banks play in financial intermediation and in attracting savings in these economies. From that perspective, the crises can provide an opportunity for building up stronger and healthier banking systems.

85. Although stricter enforcement of prudential regulations and closer monitoring could have reduced the magnitude of banking problems, it is unlikely that strong banking supervision alone could have prevented the emergence of crises in the Baltic countries. That said, given the close ties between banks and enterprises, strict enforcement of insider/connected lending limits would have gone a long way toward minimizing banking problems. During the crises, the damage could have been much smaller if central banks had started to apply restrictions to the activities of problem banks sooner. More generally, the crises underscored the need for the development of early-warning systems and for speedy action once problems were identified. A crucial aspect is the resolution of problems of medium and large banks; the closure of small banks has proven to be relatively easy and was dealt with competently by Baltic central banks. However, action toward larger banks was often influenced by vested interests and political constraints, and the resulting delays and hesitations significantly contributed to undermining confidence in the banking system and to increasing the resolution costs of crises.

86. The Baltic banking crises brought about a significant contraction in the banking system’s role in credit intermediation, as well as increased incentives for disintermediation, tax evasion and underground activities. On the other hand, although the real effects of the crises remain difficult to quantify, their impact on output and incomes has been small compared to the magnitude of bank failures involved. Nonetheless, credit risk remains high, banks remain fragile and further crises cannot be ruled out until the appropriate supporting institutions, regulations, markets, expertise, as well as a viable enterprise sector have been developed. The necessary ingredients to reduce credit risk and induce banks to resume private sector lending include: (a) continued fiscal restraint, to free resources to the private sector; (b) progress in enterprise restructuring, to open up sound investment opportunities for banks; (c) improved market discipline, to encourage internal governance; and (d) the development of an appropriate legal framework for conducting business.

References

  • Bennett, Adam, 1992, “The Operation of the Estonian Currency Board”, IMF Paper on Policy Analysis and Assessment 92/3 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Bennett, Adam, 1994, “Currency Boards: Issues and Experiences”, IMF Paper on Policy Analysis and Assessment 94/18 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Berengaut, Julian, and others, 1996, Republic of Lithuania—Recent Economic Developments, IMF Staff Country Report No. 96/72 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Calvo, Guillermo A., and Manmohan S. Kumar, 1993, “Financial Sector Reforms and Exchange Arrangements in Eastern Europe”, IMF Occasional Paper No. 102 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Camard, Wayne, 1996, “Discretion with Rules? Lessons from the Currency Board Arrangement in Lithuania”, IMF Paper on Policy Analysis and Assessment 96/1 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Caprio, Gerard, and Daniela Klingebiel, 1996, “Bank Insolvencies: Cross Country Experience”, unpublished (Washington: World Bank).

  • Galbis, V., 1994, “Sequencing of Financial Sector Reforms: A Review”, IMF Working Paper No. 101, September (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Gallik, David, Cestmis Jesina and Stephen Rapaway, 1968, “The Soviet Financial System: Structure, Operation, and Statistics” (Washington: U.S. Department of Commerce, Bureau of Census).

    • Search Google Scholar
    • Export Citation
  • Geadah, Sami, and others, 1995, Republic of Estonia—Recent Economic Developments, IMF Staff Country Report No. 95/40 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Genberg, Hans, 1991, “On the Sequencing of Reforms in Eastern Europe”, IMF Working Paper No. 13, February (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Grossman, Gregory, 1963, “Notes for a Theory on the Command Economy”, in Soviet Studies, Vol. 15, No. 2, October.

  • Hansson, Ardo, and Jeffrey Sachs, 1994, “Monetary Institutions and Credible Stabilization: A Comparison of Experiences in the Baltics,” Paper presented at a conference on Central Banks in Eastern Europe and the Newly Independent States, University of Chicago Law School, April 2223.

    • Search Google Scholar
    • Export Citation
  • Hansson, Ardo, and Triinu Tombak, 1996, “Banking Crises in the Baltic States: Causes, Solutions and Lessons”, Paper prepared for the “Second Dubrovnik Conference on Transition Economies.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 1983, ”Interest Rate Policies in Developing Countries”, Occasional Paper No. 22, October, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, the World Bank, OECD, and EBRD, 1991, A Study of the Soviet Economy.

  • Knöbl, Adalbert, and others, 1993, Republic of Estonia, IMF Economic Review No. 4 (Washington: International Monetary Fund).

  • Lingren, Carl-Johan, Gillian Garcia, and Matthew I. Saal, 1996, Bank Soundness and Macroeconomic Policy, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • McKinnon, Ronald L, 1991, The Order of Economic Liberalization: Financial Control in the Transition to a Market Economy (Baltimore, Maryland: Johns Hopkins University Press).

    • Search Google Scholar
    • Export Citation
  • Pazarbasioglu, Ceyla and J.W. van der Vossen, (forthcoming), “Design of Bank Restructuring Strategies in the Baltic States and the CIS: Main Issues and Challenges”, IMF Occasional Paper (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Quirk, Peter, and Owen Evans, et. al., 1995, ”Capital Account Convertibility, Review of Experience and Implications for IMF Policies”, IMF Occasional Paper No. 131, October 1995 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Saavalainen, Tapio O., 1995, ”Stabilization in the Baltic Countries: Early Experience”, in “Road Maps of the Transition”, IMF Occasional Paper No. 127, September, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Sundararajan, V., 1990, “Financial Sector Reform and Central Banking in Centrally Planned Economies”, IMF Working Paper No. 90, December (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Sundararajan, V. and Tomas J.T. Balino, 1991, “Issues in Recent Banking Crises” in Banking Crises: Cases and Issues (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Van Arkadie, Brian and Mats Karlsson, 1992, Economic Survey of the Baltic States (New York: New York University Press).

  • van der Mensbrugghe, Emmanuel, and others, 1995, Republic of Latvia—Recent Economic Developments, IMF Staff Country Report No. 95/125 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Villanueva, D. and Mirakhor, A., 1990, “Strategies for Financial Reforms”, IMF Staff Papers, Vol. 37, No. 3, September (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Wolf, Thomas, and others, 1994, Republic of Estonia, IMF Economic Review No. 7 (Washington: International Monetary Fund).

1

The authors thank, without implication, Julian Berengaut, Odd Per Brekk, Dennis Jones, Henri Lorie, Ceyla Pazarbasioglu, Christopher Ryan, and Emmanuel van der Mensbrugghe for their comments and suggestions, and Alexander Keenan for his research assistance.

3

For a discussion of the elements of financial sector reform and sequencing, see Galbis (1994) and Genberg (1991).

5

See Caprio and Klingebiel (1996) for an estimate of losses from banking crises in industrial and developing countries.

6

For further reference on the banking system before the 1987/88 Soviet reforms, see Gallik, et.al. (1968).

10

On July 17, 1987, the CPSU Central Committee and the USSR Council of Ministers issued Resolution # 821, Item 121 “On the improvement of the banking system in the country and boosting its effect on the enhancement of economic efficiency”.

11

Cooperative and commercial banks were subject to different prudential regulations.

13

The first commercial bank to open under the new banking law in 1988 was Tartu Commercial Bank (Estonia). While the bank went under in the 1992 banking crisis, one of its branch offices, which became independent, evolved into Hansabank and became eventually one of the largest banks in the Baltics.

14

See Saavalainen (1995) for a discussion on the Baltics’ experience in stabilization during 1992-94.

15

Central banks were established in Estonia and Lithuania in 1990, and in Latvia in 1991.

16

For the operation of the Estonian currency board, see Bennett (1992); for issues and experiences with currency boards in three countries, including Estonia, see Bennett (1994).

17

Camard (1996) describes the process leading up to the establishment of the currency board arrangement and Lithuania’s early experience with it. See also IMF (1996) for Lithuania’s experiences under the currency board arrangement.

18

See Calvo and Kumar (1993) for a discussion of the different approaches in dealing with bad loan porfolios in Eastern Europe.

19

Although the string of bank failures in Latvia took place in the spring of 1995, the break in the statistical series was recorded in December 1995, when the licenses of several large banks, including the largest, Bank Baltija, were revoked. Until then, deposits in these banks had been frozen.

20

Clearing of net balances in a single clearing process required banks to hold large reserves at their correspondent account with the central bank in order to accommodate volatile payment flows. When a second clearing stage was added, banks with negative balances after the first clearing stage and insufficient funds on their correspondent account could acquire funds in the interbank money market, thus permitting a reduction of excess reserve balances.

21

For a detailed description see IMF (1994).

22

Treasury bills have not been issued in Estonia where there has been little need for budget financing domestically.

23

However, inadequate accounting procedures may have partially influenced this picture. Valuation gains may have been counted as income, while standard international accounting treatment would require to only identify realized gains as income.

24

Insider lending is a narrower concept than connected lending, with insider lending defined as lending to managers, employees, shareholders, usually of 20 percent or more, their relatives and companies owned or controlled by them. Connected lending includes the smaller shareholders, usually of 10 percent or more. Most countries, including Estonia and Lithuania, only have a limit on connected lending.

25

During the collapse of the Estonian Social Bank in late 1994 it became evident that over 30 percent of loans had been extended to shareholders which had circumvented connected lending limits by establishing an elaborate system of shell companies. Insider lending was one of the main factors behind the debacle of Bank Baltija in Latvia in May 1995, and the banks, Aura, Innovation and Vakaru in Lithuania in late 1995.

26

Foreign exchange risks are covered by the regulation on capital adequacy.

27

Major Latvian banks were required to follow IAS standards for their 1993 financial accounts; the requirement was extended to all banks in the context of the new credit institutions law which was passed by Parliament in October 1995 and took effect in January 1996. This requirement was introduced in Estonia in 1995 and is scheduled to be applied in Lithuania from end-1996.

29

Throughout the paper, broad money is defined to include residents’ foreign currency deposits.

30

During the Lithuanian banking crisis, deposit withdrawals translated into outflows through the currency board and, as a result, the ratio of currency to broad money did not exhibit a large increase (see Section V.D.).

31

For a survey of interest rate policies in the context of financial sector reforms, see Villanueva and Mirakhor (1990) and IMF (1983).

32

See Hansson and Sachs (1994) for a detailed discussion.

33

See Hansson and Tombak (1996) for a discussion on the reasons for and lessons learned from the banking crises in the Baltic countries.

34

In Lithuania, although the crisis was precipitated by private banks, the three-state controlled banks, which dominated the market, were also known to have weak portfolios at the time of the crisis.

36

The requirement for loan loss provisions was introduced in Latvia in December 1993 and in Lithuania, in December 1994.

37

See IMF (1993) for a more detailed description.

38

In February 1995, the Ministry of Interior ordered armed intervention in the Lainbanka, charging its management with abuse of authority and fraudulent behavior. The case received ample press coverage.

39

The two banks which received liquidity support from the authorities in the third quarter of 1995, Aura Bank and Vakaru Bank, comprised eight percent of total commercial bank deposits.

40

Innovation Bank and Litimpeks Bank accounted for 15 percent of commercial bank deposits.

41

Deposit liabilities in these three banks accounted for 40 percent of broad money.

42

For more detail see IMF (1993).

43

Baltija’s auditors had encountered difficulties in producing the bank’s 1993 financial statement; the report was completed only in the fall of 1994. In the meantime, Bank of Latvia examinations had detected serious problems, nevertheless, the bank’s expansion continued until its collapse.

44

Another 15 banks, including some of the largest ones, had their licenses revoked and two banks had their operations suspended in 1995, reducing the number of operating banks to 38 at year-end. As in the previous year, the main reasons for closing these banks was insider lending and poor credit policies.

45

A limited amount of liquidity support was extended to problem banks by the Bank of Latvia.

46

The impact of the failure of these banks on the banking system was negligible as their deposits totaled less than four percent of commercial bank deposits.

47

Deposits at the three state-controlled banks are fully guaranteed by the government through end-1996 in accordance with the Civil Code.

48

IMF (1995).

49

Blocked deposits corresponded to 32 percent of total deposits at end-June 1995.

50

The outstanding stock of T-bills rose from 1 percent of GDP in June 1995 to 4 percent of GDP in March 1996.

51

The share of foreign currency deposits increased from 40 percent in December 1994 to 49 percent in December 1995 (Chart 9); the currency/deposit ratio increased from 45 percent to 67 percent during the same period.

  • Collapse
  • Expand
Financial Sector Reform and Banking Crises in the Baltic Countries
Author:
International Monetary Fund