Cost and Effectiveness of Banking Sector Restructuring in Transition Economies
Author: Ms. Edda Zoli

Contributor Notes

Author’s E-Mail Address: ezoli@imf.org
Publication Date:
01 Oct 2001
eISBN:
9781451857498
Language:
English
Keywords:
WP; cost; banking sector; bank; banking sector problem; restructuring operation; banking sector feature; restructuring policy; liquidity position; restructuring program
Download PDF (1.5 MB)

The paper analyses the cost and effectiveness of bank restructuring policies in 11 transition countries during 1991-98. It argues that country-specific banking sector features, the size of bad loans inherited from the centrally planned system, and weaknesses in the restructuring policies implemented were the main factors affecting the overall fiscal costs, with the latter two being more significant. The paper finds no significant relationship between the size of restructuring costs and overall improvement in banking sector performance for the sample countries as a whole.

Abstract

The paper analyses the cost and effectiveness of bank restructuring policies in 11 transition countries during 1991-98. It argues that country-specific banking sector features, the size of bad loans inherited from the centrally planned system, and weaknesses in the restructuring policies implemented were the main factors affecting the overall fiscal costs, with the latter two being more significant. The paper finds no significant relationship between the size of restructuring costs and overall improvement in banking sector performance for the sample countries as a whole.

I. Introduction

The development of a sound and efficient banking sector has been a major challenge facing countries in transition from a centrally planned to a market economy. Virtually all countries of Central and Eastern Europe (CEE) and the former Soviet Union went through episodes of extensive banking sector problems, which necessitated substantial restructuring of the banking sector, that, in turn, generated sizable cost for the governments and central banks involved.2

Both the magnitude and composition of bank restructuring costs varied widely across these countries (Figure 1).3 In 6 of the 11 countries considered, the estimated total cost of banking sector restructuring was more than 9 percent of GDP (in four cases, it exceeded 15 percent of GDP); in the remaining five countries it was under 5 percent of GDP. The cost for depositors’ compensation was typically moderate in all sample countries, with the exception of the former Yugoslav Republic of Macedonia. Although the bulk of the restructuring costs was typically borne by governments, in Bulgaria, the Czech Republic and the Kyrgyz Republic, central banks incurred sizable expenditures as well. In Estonia and the Kyrgyz Republic, governments transferred funds to central banks to cover the losses generated by their support to troubled banks4.

Figure 1:
Figure 1:

Costs of Banking Sector Problems in Percent of GDP (1991–98) 1/

Citation: IMF Working Papers 2001, 157; 10.5089/9781451857498.001.A001

Source: Author’s calculations.1/ These total costs are the sum of the present value at end-1998 of annual costs during 1991–98, as a percentage of 1998 GDP. For details see Appendix I. The estimates do not take into account cost recoveries.2/ Sum of the amount of bonds issued for bank recapitalization, cash transfers to banks, and expenses for called loan guarantee (when applicable), excluding government transfers to the central bank.3/ Sum of losses and provisions on credit extended to banks, excluding losses covered by government transfers.4/ Sum of the amount of bonds issued and of cash transfers for deposit compensation.5/ For Macedonia, FYR the expense for deposit protection includes the obligations related to the compensation for foreign currency deposits. See text for explanation and discussion.

What factors contributed to the emergence of such costs? Did countries that incurred higher bank restructuring costs improve their banking sector more than did countries that spent less? These are the two key questions that this paper tries to answer in the context of the experiences of the 11 transition economies. With regard to the second issue, the paper finds no significant relationship between total banking sector restructuring costs and overall improvement in banking sector performance for the sample countries as a whole.

To address the first issue, the paper considers several factors that might account for the different outcomes, and analyzes, in particular, the role played by the specific features of the banking sector in the countries concerned; the size of the non-performing loans inherited from the former centrally planned regime; and the weaknesses in the design and implementation of the restructuring policies. To assess the latter, the paper develops indices of the effectiveness of bank restructuring, capturing the key features of the restructuring and reform policies expected to affect the total cost. The empirical analysis seems to support the proposition that restructuring programs entailing repeated recapitalizations, significant involvement of the central bank in the restructuring scheme, inadequate institutional restructuring, and the existence of more than one restructuring agency amplify the total cost of banking sector restructuring.

The paper is organized as follows. Section II provides a brief overview on banking sector problems in the countries under consideration. It also discusses the main causes of these problems and the actions taken by the authorities concerned to address them. Section III examines the factors affecting the total cost of banking sector restructuring. Section IV presents the econometric analysis. Section V discusses the relationship between bank restructuring costs and indicators of banking sector performance. Concluding remarks and policy implications are presented in Section VI.

II. Background

A. Origins of Banking Sector Problems

The 11 countries studied repeatedly faced banking sector problems over the period 1991–98, adversely affecting the solvency and liquidity of both state-owned and private banks.5 Previous studies on the experience of these countries have identified the following five factors as the main reasons for the emergence of such problems.6

(1) External shocks. All countries in Europe and Central Asia suffered from the collapse of the Council for Mutual Economic Assistance trade system, which had previously isolated the conditions of foreign trade within socialist countries from those prevailing in the rest of the world. Furthermore, some countries were hit by idiosyncratic external shocks. Specifically, a few banks located in countries of the former Soviet Union lost access to part of their assets held in Russia after the collapse of the Soviet empire. In the mid-1990s, bank profitability in Latvia and Lithuania was sharply reduced by the decline of trade financing opportunities resulting from price liberalization in Russia.7

(2) Macroeconomic conditions. The transition process and external shocks led to severe output contractions, which often precipitated banking sector problems. In addition, in certain countries, stabilization programs involving tight monetary policies forced nominal interest rates up and reduced inflation, thus raising real interest rates and adversely affecting borrowers’ ability to service their debt8

(3) The transition process. The transition process made the banking sector vulnerable in different ways. First, internal and external liberalization and the removal of enterprise subsidies curtailed the profitability of enterprises and reduced their ability to repay loans. Second, both enterprises and the newly commercialized banks lacked experience operating with a profit-oriented approach. Third, the commercial banks carved out from former state banks inherited poor quality loans created under the central planning system. Finally, in Georgia, Kazakhstan, Kyrgyz Republic, the continued use of directed credit and on-lending under government instruction contributed to the emergence of weak bank portfolios.

(4) Deficiencies in supervision and in the legal framework. In most transition countries of the early 1990s supervisory systems were inadequate and prudential regulations were missing, so credit institutions often engaged in risky lending, prompting a deterioration of the quality of their portfolios.

(5) Poor internal governance. In some countries, corruption, insider lending, fraud, and inadequate disclosure further weakened the banking system. In fact, management misconduct and insider lending were behind the formation of bad loan portfolios in two Estonian banks (liquidated respectively in 1992 and 1994), in the largest Latvian bank (liquidated in 1995), and in the second largest bank in Hungary, which in 1997 suffered from substantial liquidity problems.

B. Authorities’ Responses: Bank Restructuring and Deposit Compensation

Different instruments have been employed for financial restructuring in the 11 countries studied.9 In most cases the government boosted bank’s net worth through the transfer of consolidation bonds (Table 1). Occasionally, governments resorted to the transfer of cash or property, the reduction of bank’s liabilities, the provision of public guarantees on outstanding loans and repurchase agreements. Central bank’s assistance to the banking sector entailed mainly liquidity support (Table 2).

Table 1.

Instruments of Financial Restructuring and Government Assistance to the Banking Sector

article image
source: Tang, zoil, and klytchniKova (2000)

For directed credits extended under government instruction.

In exchange the government received an equity position in the enterprises.

Table 2.

Instruments of Central Bank Assistance to the Banking Sector

article image
Source: Tang, Zoli, and Klytchnikova (2000).

In most of the countries depositor compensation was limited. Only in Bulgaria; Macedonia, FYR; and Lithuania was the cost higher than 1 percent of GDP.10 In Macedonia, FYR the cost was estimated to be very high, because the government granted explicit guarantee for approximately DM 1.1 billion foreign exchange deposits (about 17.5 percent of GDP), lost after the dissolution of the of the former Socialist Republic of Yugoslavia.11

III. Factors Affecting Restructuring Costs

The 11 countries faced similar problems in their transition from a centrally planned system to a market economy. Why then was the result in government and central bank costs for banking sector restructuring so different? This section analyzes the contributions of three factors to this outcome, namely the country-specific banking sector features that affected the authorities decision on whether to liquidate or restructure ailing banks; the size of non-performing loans inherited from the socialist era; and the design weaknesses and implementation slippages of the restructuring programs.

A. Liquidation versus Restructuring

To address banking sector problems, authorities in the CEEs implemented extensive restructuring and recapitalizations, whereas in the Baltics and the CIS they pursued a combination of liquidations and restructuring. Since deposit compensation was limited, the fiscal costs associated with bank liquidations were very small. The different strategies adopted to deal with ailing banks, then, partly explain the variance in the total cost of resolving banking sector problems. The authorities decision on whether to liquidate or restructure banks, in turn, was affected by the country-specific features of the banking sector.

In the CIS and, to a lesser extent, the Baltics, the banking sector consisted primarily of many newly created small and often undercapitalized banks. Although in some countries the number of credit institutions was large, financial intermediation was very low, as indicated by the ratio of credit to the private sector to GDP and by the ratio of M2 to GDP (Table 3). Consequently, the authorities were able to respond to banking sector problems by closing the insolvent institutions, without disrupting the shallow financial system. In these economies the resolution of banking problems resulted in the consolidation of the banking sector, hi fact, as shown in Figure 2, the sharp increase in the number of banks in most of the CIS and the Baltics in the early 1990s was followed by a sharp drop in the number of banks per capita from the mid 1990s.

Figure 2.
Figure 2.

Number of Banks

(Per million inhabitants)

Citation: IMF Working Papers 2001, 157; 10.5089/9781451857498.001.A001

Source: EBRD Transition Report (2000)
Table 3.

Features of the Banking Sector in Sample Countries

article image
article image
Source: EBRD Transition Report (2000); IMF International Financial Statistics.

Percent of GDP.

In the CEEs, instead, the number of new small banks was lower, and the quality of the new banks was better than that in the CIS and the Baltics. New entrants, in fact, also included foreign banks, which were generally sounder than the domestic private banks (Table 3). Furthermore, financial penetration was much deeper than in the CEEs in the other two groups of countries, as measured by the ratio of credit to the private sector to GDP and by the ratio of M2 to GDP (Table 3). In this context, liquidation of problem banks could mean wiping out most of the banking system, imposing huge economic and political costs. As a consequence, in the CEEs, with the exception of Bulgaria,12 the resolution of banking problems did not result in any significant change in the number of banks in the system.

B. Inherited Non-Performing Loans

The problem of the non-performing loans outstanding at the inception of the two-tier banking system was significant especially in Bulgaria, Hungary, and the Czech Republic, prompting those countries to implement special bank restructuring programs to resolve for selected inherited bad assets.13 The way the authorities handled these non-performing loans added to the total cost of banking sector problems over the 1990s in two ways. First, governments issued large amounts of bonds to recapitalize banks. Second, the drawbacks of the consolidation programs carried out to deal with inherited bad debt played a role in exacerbating successive banking sector problems, raising the total cost of bank restructuring.

The origin of inherited non-performing loans can be traced to the credits extended during the socialist period to inefficient state-owned enterprises, without any preliminary credit risk assessment, and passed on to the commercial banks carved out of the monobank. Whereas the CIS countries; the Baltics; Macedonia, FYR; and Poland experienced hyperinflation in the early 1990s, which greatly reduced the real value of their bad debt, in Bulgaria, Hungary, and the Czech Republic inflation was not as high (Appendix III), so the real value of inherited bad debt was only partially eroded.

Table 4 reports the estimated fiscal costs of the initial banking sector restructuring operations, carried out in Bulgaria, the Czech Republic and Hungary at the beginning of the 1990s to clean bank portfolios from inherited bad credits.14 These figures represent an appraisal of the fiscal costs for inherited bad loans, but they have to be interpreted with caution, as these costs might include some expenses related to the clean-up of non-performing loans created during the very first years of the transition. The estimates indicate that the restructuring packages implemented in the early 1990s contributed substantially to the total cost of banking sector problems during 1991-98.

Table 4.

Estimated Fiscal Costs of the Consolidation Programs for Inherited Bad Loans

article image
Source: Author’s estimates.

Calculated as the sum of the present value at end-1998 of annual costs, as a percentage of 1998 GDP. For details see Appendix I.

Furthermore, the weaknesses of these consolidation programs contributed to successive banking sector problems, thus boosting the total cost of bank restructuring. There were three major drawbacks associated with the early consolidation programs, which are discussed below.

Inadequate financial restructuring at the beginning

The loan consolidation packages did not always sufficiently strengthen banks’ net worth and liquidity position. Inadequate bank recapitalization resulted in recurrent bank problems as undercapitalized banks faced distorted incentives in extending new loans and were prone to excessive risk taking.15 The transfer of government bonds to banks in some cases also failed to improve liquidity as the securities carried a minimum selling price or paid below-market interest rates.16

Insufficient institutional and operational restructuring.17

The loan consolidation schemes to remove inherited bad loans from banks’ portfolios did not affect the culture of banks or their operating procedures. Until the late 1990s in all the three countries the supervisory regime was lax and the regulatory and legal frameworks were inadequate;18as a result the misuse of credit was widespread. Moreover, without adequate operational restructuring, many banks continued to rollover outstanding loans to state-owned enterprises.19

Lack of credibility and expectations of further bailouts.

Government actions created expectations of further bailouts and encouraged moral hazard in banks’ lending decisions. For example, in Hungary, after announcing that inherited bad loans would be covered only by a partial government guarantee, the authorities launched two successive far-reaching consolidation programs. Furthermore, the government acted as a soft negotiator by extending the date for credits eligible for bailout20, by reneging on the terms of the consolidation bonds in favor of banks21 and by expanding the number of state-owned enterprises to be bailed out. Such government behavior encouraged excessive risk taking and creation of new bad loans.22 Also, in the Czech Republic, while implementing the first consolidation program (1991–93), the government retained a controlling stake in major banks, creating the expectations of future bailout.

C. Implementation Procedures of Bank Restructuring

Design weaknesses and implementation slippages of bank restructuring policies (in short, “bad policies”) may aggravate the associated overall costs. Table 5 provides a synopsis of the main features of the restructuring policies implemented in the 11 transition countries that may have ultimately affected the magnitude of total government and central bank costs. In order to quantify and assess the impact of bad policies on the total cost of banking crises, it may be helpful to develop appropriate bank restructuring effectiveness indices (REIs) to capture and evaluate the key cost-critical features of bank restructuring policies.

Table 5:

Key Aspects of Bank Restructuring (1991-98)

article image
Souce: IMF Country Reports, OECD Economic Surveys, various issues.

Only for banks involved in the Consolidation Program II.

As a first step, a 0-1 dummy is assigned to selected cost-critical features of the restructuring policies described in Table 5,23 subsequently, all the dummies are summed up to create a REI. To check for the robustness of the results obtained when one or two sensitive components of the REI are dropped, four alternative version of the same index are constructed (EFFECTIVE 1 to EFFECTIVE4)24 Large values of REIs indicate high effectiveness of restructuring policies and are expected to be associated with lower restructuring costs. The REIs are meant to evaluate only the effectiveness of the restructuring programs in minimizing the fiscal costs and not to provide an overall assessment of the effectiveness of the bank restructuring programs.

The four REIs are based on the following assumptions, which will be tested in the empirical analysis below: (i) financial restructuring operations entailing repeated recapitalizations are expected to intensify moral hazard and amplify total fiscal costs; (ii) central bank liquidity support to insolvent institutions is expected to increase total costs, by generating losses, that will be borne either by the monetary authority directly or, eventually, by the government;25 (iii) the bail-out of insolvent banks that are considered “too big to fail” is expected to create moral-hazard and magnify the fiscal costs of bank restructuring; (iv) consolidation programs that do not provide adequate recapitalization from the start and that fail to effectively improve bank liquidity will likely require further intervention, raising total fiscal costs;26 (v) insufficient institutional restructuring is expected to contribute to higher fiscal expenditures; (vi) the existence of more than one restructuring agency may duplicate operating costs, jeopardize the transparency of the restructuring program, and ultimately generate unforeseen budget costs, thus it is expected to increase total costs;27 and (vii) a large involvement of the central bank in consolidation schemes may create conflict of interest and moral hazard since the monetary authority may end up having to supervise banks in which it has a direct economic interest28. This may lead to delay in the resolution of banking problems and is, therefore, expected to increase the associated costs.29

Figure 3 plots the total cost of banking sector problems against the REI EFFECTIVE2, for the 11 countries. As expected, there appears to be a negative relationship between the index and total costs. But how important was the role of bad policies in determining the total cost of bank restructuring? To try to evaluate the impact of the REIs on the cost of banking sector problems against other factors that might have exacerbated these costs, cross-country and panel data analyses are performed in the next section. It has to be stressed, however, that owing to the small number of observations and other data limitations,30 the following econometric results should be seen as merely suggestive rather than conclusive.

Figure 3.
Figure 3.

Cost of Banking Problems and REI

Citation: IMF Working Papers 2001, 157; 10.5089/9781451857498.001.A001

Note: For details on costs see Appendix I. The costs related to the compensation for foreign currency deposits in Macedonia, FYR is deducted.

IV. Econometric Analysis

A. Cross-Section

The discussion above suggests that three main factors seem to have contributed to the total cost of banking problems: financial depth prior to the emergence of the problems, the size of inherited non-performing loans, and the weaknesses of bank restructuring programs. In the following statistical analysis three variables are used alternatively as indicators of financial depth: the ratio of credit to the private sector to GDP and the ratio of broad money to GDP before the eruption of bank problems31(INITCREDIT and INITM2, respectively), as well as the average of these two variables (INITCREDM2)32The share of inherited non-performing credits in total credits is adopted as a measure of the size of inherited bad loans (INHERITED).33Three different variables are used to assess the cost of banking sector problems: the total cost (TOTAL), the total cost less the cost for deposit compensation (RESTRUCT), and the total cost less the cost for inherited bad loans (NET COST).34

Tables 6 and 7 show the correlation coefficients and the Spearman rank correlation coefficients between the different variables measuring the costs of banking problems and the variables that might have affected these costs. As expected, the four REIs are negatively correlated with the cost variables. Also, the rank correlation coefficients have the expected sign and are statistically significant. The different indicators of initial financial depth and INHERITED are positively correlated with the cost measures in Table 6. However the rank correlation coefficients on these variables are not statistically significant and occasionally they even exhibit the wrong sign (Table 7).

Table 6:

Correlation Coefficients

article image
Table 7:

Spearman Rank Correlation Coefficient

article image
Note: Absolute value of t statistics in parenthesis. *, **, and *** imply significance at 10%, 5% and1% level, respectively.

Also the results obtained by regressing the cost measures on the REIs, on INHERITED, and on the indicators of initial financial depth suggest that the drawbacks of restructuring policies significantly contributed to the total cost of banking sector problems in the 11 countries (see Table 8). It appears that inherited bad credits also had a positive and significant impact on the cost of banking sector problems. On the other hand, the coefficients of the variables measuring the scale of intermediation prior to the emergence of banking problems have the unexpected sign and are insignificant.

Table 8:

Cross-Country Regression Results

article image
Note: Estimates obtained using OLS. Absolute value of t statistics in parenthesis. *, **, and *** imply significance at 10%, 5% and 1% level, respectively. The REIs have been normalized, so that they assume values between 1 and 10.

B. Panel Data

Panel data analysis allows to expand the sample size, which is very limited in the above cross-sections. Although other studies have used econometric models to investigate the determinants of fiscal costs of banking crises using cross-country regressions, to the author’s knowledge this is the first attempt to address the issues using panel data.35 Since the length of banking sector problems varies across countries, an unbalanced panel is constructed (see Appendix IV for details).

The dependent variables are the total cost of banking sector problems (TOTAL), the cost excluding deposit compensation (RESTR), and the total cost excluding the expense for inherited bad loans (NET COST) for each year, as a percentage of GDP.36 Among the explanatory variables, in addition to those considered in the cross-section regressions, the model includes the European Bank for Reconstruction and Development index of enterprise reforms (ENTERPRISE). This index evaluates progress in privatization, governance, and enterprise restructuring in transition countries.37 Extensive reforms in the enterprise sector (a higher ENTERPRISE index) are expected to increase firm profitability, and reduce the quantity of non-performing credits, thus diminishing banking sector problems and their costs. To account for the macroeconomic shocks that, as discussed in Section II, caused or exacerbated banking sector problems, several macroeconomic indicators are also included among the regressors. These indicators are GDP growth (GROWTH), real interest rates on loans (REALINT), the current account balance as a percentage of GDP (CAB), the fiscal account balance as a percentage of GDP (FAB), the change in the terms of trade (TOTCHANGE).38 The assumption is that higher growth, lower real interest rates, a more favorable external and fiscal position and an increase in the terms of trade reduce banking sector problems and lower costs.39 Finally, the regressors also include time dummies (for example, T1991, T1992) as well as a dummy for the first year in which banking sector problems erupted (START).

The results of the regressions (Table 9) support the previous finding that the indices assessing the effectiveness of the restructuring policies appear to have had a significant impact on the cost of banking sector problems. The coefficients of the variable accounting for inherited non-performing loans and of the index measuring the progress in enterprise reforms have the expected sign and are statistically significant. Surprisingly, the coefficients of INITCREDIT are significant, but with the wrong sign. The coefficients of the other indicators of initial financial depth are instead insignificant. Among the variables controlling for the impact of macroshocks, only the change in the terms of trade, the fiscal account balance as a percentage of GDP, and the real interest rates show significant coefficients with the expected sign. Finally, only the time dummy T1994 has a significant coefficient.

Table 9:

Panel Data Regressions

article image
1/ Estimates obtained using OLS. Absolute value of t statistics in parenthesis. *, **, and *** imply significance at 10%, 5% and 1% level, respectively.2/ The t statistics are obtained using the White heteroskedasticity-consistent standard errors.3/ The REIs have been normalized, so that they assume values between 1 and 10.

In sum, after controlling for several factors that might have exacerbated the cost of banking sector problems, both cross-country and panel regression results suggest that weaknesses in the design and implementation of the restructuring programs aggravated the associated costs. Specifically, these empirical findings seem to support the paper’s assumption that restructuring programs entailing repeated recapitalizations, significant involvement of the central bank in the restructuring scheme, inadequate institutional restructuring, and the existence of more than one restructuring agency amplified the total cost of banking sector restructuring.

V. Bank Restructuring Costs and Banking Sector Performance

Did countries that incurred high bank restructuring costs improve their banking sector more than countries that spent less? The paper considers different aspects of banking sector conditions to assess the success of the restructuring operations undertaken in the 11 countries.40 The ratio of credit to the private sector to GDP and the ratio of M2 to GDP are used as indicators of the degree of financial depth. These variables are expected to enlarge when the banking sector is effectively restructured. An increase in credit to the private sector, however, may also reflect bad lending practices. Therefore, the share of non-performing loans41 is included as an additional indicator of banking sector performance. Interest rate spreads (lending minus deposit rates) and central bank credit to banks are taken as measures of the efficiency of intermediation. An increase in interest rate spreads may indicate that banks are facing riskier borrowers, and hence charging them with higher rates, or that banks need to cover larger expenses due to loan losses. So a decline in the spreads is interpreted as an improvement in efficiency. Also lower reliance of banks on central bank credit indicates progress in the performance of the system. The ratio of Ml to M2 and the currency-to- deposit ratio are used as measures of confidence in the banking sector. As these variables reflect the degree to which the banking sector is able to mobilize long-term liabilities, a decline in these indicators is interpreted as a sign of increased confidence in the system.

Table 10 shows the change in the indicators of banking sector performance after the resolution of banking sector problems, relative to the years before the eruption of the problems. All 11 countries show a significant decline in the share of non-performing loans and central bank credit to banking institutions. In certain countries, however, some variables, such as the ratio of M2 to GDP, and the ratio of Ml to M2 point to a weakening in the intermediation capacity of the banking sector and a decline in the public’s confidence in the financial system.

Table 10.

Improvements in Banking Sector Performance 1/

article image

Percent change between performance before and after the resolution of banking sector problems.

Lending minus deposit rate (percent per annum).

To investigate whether there is any relation between the cost of bank restructuring and improvement in banking sector conditions, an overall measure of progress in banking sector performance is constructed. This measure is defined as the unweighted average of the increase in credit to the private sector, increase in the ratio of M2 to GDP, decline in central bank credit to banks, decline in the currency-to-deposit ratio, decline in the Ml-to-M2 ratio, and decline in the share of non-performing loans. 4 A simple look at the data underscores no significant relationship between total banking sector restructuring costs and overall improvement in banking sector performance for the sample as a whole (Figure 4). This is probably because progress in banking sector conditions was largely affected by macroeconomic as well as idiosyncratic shocks. Furthermore, part of the costs was not discretionary, but was incurred in response to impending crises, so it could not be well targeted toward improving the performance of the banking sector.

Figure 4.
Figure 4.

Banking Sector Restructuring Costs and Banking Sector Performance

Citation: IMF Working Papers 2001, 157; 10.5089/9781451857498.001.A001

1/ Total costs less costs for deposit compensation. For details, see Appendix I.2/ Average of the increase in credit to the private sector, increase in M2-to-GDP ratio, decline in central bank credit to banks, decline in the currency-to-deposit ratio, decline in Ml-to-M2 ratio, decline in the share of non- performing loans following the resolution of banking sector problems.3/ The vertical and horizontal lines mark the median.

Figure 4 divides the sample countries into four subsets by drawing a vertical and a horizontal line to correspond with the two median countries, Macedonia, FYR and Bulgaria. Countries in the first group (Bulgaria, Hungary, the Czech Republic and the Kyrgyz Republic) undertook expensive restructuring programs that significantly improved banking sector conditions. Countries in the second group (Kazakhstan and Macedonia, FYR) made little progress in banking sector performance relative to the other countries in the sample, despite the relatively high cost of the restructuring policies. For countries in the third group (Georgia and Latvia), low banking sector restructuring costs are associated with relatively small improvements in banking sector performance. Finally, the countries in the fourth group (Estonia, Poland and Lithuania) appear to be those in which restructuring policies produced significant progress in banking sector conditions despite their relatively low cost.

VI. Concluding Remarks and Policy Implications

This paper analyzed the banking sector restructuring costs incurred by the government and the central bank in 11 transition economies. It pointed out that although all these economies confronted similar problems, the size and composition of the overall bank restructuring costs varied greatly among them. In this connection, the paper examined the contribution of three main factors in explaining the different outcomes, namely, the specific features of the banking sector in each country; the size of non-performing loans inherited from the previous centrally planned regimes; and the design weaknesses and implementation slippages in the restructuring policies put in place. The results of empirical analysis appear to support the hypothesis that inherited non-performing loans and weaknesses in the restructuring operations had a significant impact on the cost of banking sector problems. This finding also holds after controlling for other factors, such as macroeconomic shocks, that might have exacerbated those costs. The paper finds no relationship between total bank restructuring cost and the ex post overall improvement in banking sector performance for the sample countries as a whole.

The analysis of the experiences of the 11 transition economies supports the following policy implications.42

  • To achieve a successful and sustained resolution of banking sector problems and to minimize the associated costs for the government and the central bank, financial, operational, and institutional restructuring involving improvements in banking supervision and regulation would need to be implemented simultaneously and comprehensively.

  • Financial restructuring must provide adequate recapitalization, and effectively strengthen banks’ liquidity position to curtail the risk of moral hazard inherent in expectations of repeated injections of public funds.

  • Consistency and credibility of the authorities’ actions and announcements are key to containing expectations of future bailouts.

  • The policy of bailing out banks that are regarded as too big to fail may induce moral hazard and magnify the cost of bank restructuring.

  • Significant involvement of the central bank in bank restructuring operations may generate conflict of interests reduce transparency, and ultimately result in unexpected costs for the government budget.

  • The existence of more than one centralized restructuring agency may produce duplication of operating costs, reduce the transparency of restructuring policies, and create unanticipated expenses for the government budget.

References

  • Anderson, Ronald, and Chantal Kegels, 1998, “Transition Banking—Financial Development of Central and Eastern Europe” Oxford: Clarendon Press.

    • Search Google Scholar
    • Export Citation

  • Capek, Ales, 1994, “The Bad Debts Problem in the Czech Economy,” Moct-Most, No. 3, pp. 59–70.

  • Czech Republic National Bank, 1998, Annual Report (Prague: Czech Republic National Bank).

  • Daniel, James, 1997, “Fiscal Aspects of Bank Restructuring,” IMF Working Paper No. 97/52 (Washington: International Monetary Fund).

  • Davis Jeffrey M. and Timo Välilä, 2000 “Fiscal Support in Financial Sector Restructuring: Analytical Issues” (Unpublished; Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Dziobek, Claudia, and Ceyla Pazarbasioglu, 1997, “Lessons from Systemic Bank Restructuring: A Survey of 24 Countries,” IMF Working Paper No. 97/161(Washington: International Monetary Fund, December).

    • Search Google Scholar
    • Export Citation

  • Enoch Charles, Barbara Balwin, Olivier Frecaut, and Arto Kovanen, 2001 “Indonesia: Anatomy of a Banking Crisis—Two Years of Living Dangerously—1997–99” IMF Working Paper No. 01/52 (Washington: International Monetary Fund, December).

    • Search Google Scholar
    • Export Citation

  • European Bank for Reconstruction and Development, 2000 Transition Report, (London: EBRD).

  • European Bank for Reconstruction and Development, Transition Report, Various issues (London: EBRD).

  • Fleming, Alex, and Samuel Talley, 1996, “The Latvian Banking Crisis—Lessons Learned,” World Bank Policy Research Working Paper, No. 1590 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

  • Frydl, Edward, 1999, “The Length and Cost of Banking Crises,” IMF Working Paper No. 99/30 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Honohan, Patrick and Daniela Klingebiel, 2000, “Controlling Fiscal Costs of Banking Crises,” World Bank Policy Research Working Paper No. 2441 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund, Staff Country Reports, various issues (Washington).

  • International Monetary Fund, 1997, Hungary, IMF Staff Country Report No. 97/103, (Washington).

  • International Monetary Fund, 2000, Former Yugoslav Republic of Macedonia, IMF Staff Country Report No. 00/76, (Washington).

  • Kawalec, Stefan, 1999, “Banking Sector Systemic Risk in Selected Central European Countries,” (Unpublished).

  • Organization for Economic Cooperation and Development, 1996, Czech Republic, OECD Economic Surveys (Paris: OECD).

  • Organization for Economic Cooperation and Development, 1998, Czech Republic, OECD Economic Surveys (Paris: OECD).

  • Pazarbaşioğlu, Ceyla, and Jan Willem van der Vossen, 1997, “Main Issues and Challenges in Designing Bank Restructuring Strategies,” in Central Bank Reform in the Transition Economies, ed. by Vasudevan Sundararajan, Arne B. Petersen, and Gabriel Sensenbrenner Sensenbrenner (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Tang, Helena, Edda Zoli, and Mna Klytchnikorva, 2000, “Banking Crises in Transition Economies—Fiscal Costs and Related Issues,” World Bank Policy Research Working Paper No. 2484(Washington: World Bank).

    • Search Google Scholar
    • Export Citation

  • World Bank, 1999, Czech Republic. Toward EU Accession, A World Bank Country Study (Washington: World Bank).

APPENDIX I

Table A1.

Cost of Bank Restructuring and Deposit Compensation for the Government (1991–98)

(Percent of GDP)

article image
article image
article image

Sum of present value at end-1998 of annual costs during 1991-98, as a percentage of 1998 GDP.

The government swapped foreign currency bonds held by Mineral Bank and Economic Bank for local currency denominated securities. The figure shown is net of the bonds withdrawn.

Assistance to the banking sector provided by the National Property Fund and various asset management companies: Konsolidacni Banka (KoB); Ceska Inkasni (CI); Ceska Financni (CF)

Issued by the National Property Fund

Extended by the National Property Fund

The government assumed the obligation for the repayment of househols foreign currency deposits that were lost after the dissolution of the former Socialist Republic of Yugoslavia.