This Selected Issues paper examines the risks and structural weaknesses in Bosnia and Herzegovina. The paper provides an estimate of the current account adjustment required to stabilize net foreign liabilities. It uses the external sustainability approach of the Consultative Group on Exchange Rate Issues (CGER) methodology for exchange rate assessment. The paper analyzes the impact of the newly introduced borrowing rules on the longer-term debt dynamics. An overview of salient facts about unemployment in Bosnia and Herzegovina is also presented.

Abstract

This Selected Issues paper examines the risks and structural weaknesses in Bosnia and Herzegovina. The paper provides an estimate of the current account adjustment required to stabilize net foreign liabilities. It uses the external sustainability approach of the Consultative Group on Exchange Rate Issues (CGER) methodology for exchange rate assessment. The paper analyzes the impact of the newly introduced borrowing rules on the longer-term debt dynamics. An overview of salient facts about unemployment in Bosnia and Herzegovina is also presented.

VI. Credit Growth, Bank Soundness, and Banks’ Foreign Liabilities31

Bosnia & Herzegovina has experienced rapid credit growth in recent years. Foreign-owned banks, which account for of the bulk of banking system assets, are the main sources of credit growth. These banks have been largely relying on long-term foreign funding to finance the credit expansion.

Rapid credit growth, coupled with weaknesses in asset classification and provisioning as identified by the 2006 Financial Sector Assessment Program (FSAP), may mask the underlying credit risks. In addition, the reliance on potentially volatile foreign funding to finance credit growth may also create a risk for both financial stability and the macroeconomy, especially in the context of the currency board arrangement.

This paper examines the relationships between credit growth, bank soundness, and banks’ foreign liabilities. It also tests for the effects of the policy measures that have been implemented by the Central Bank of Bosnia & Herzegovina (CBBH) to curb credit growth in the past few years. The empirical investigation is done using bank-level and macroeconomic data from the second quarter of 2003 to the third quarter of 2006. The measure of bank soundness used in this paper is the “distance-to-default”, which measures the probability that losses exceed equity and thus captures the risk of insolvency.

The results show the following:

  • There is no evidence that stronger banks expand credit faster.

  • The policy measures of the CBBH have had no significant impact on credit growth.

  • Credit growth improves bank soundness, possibly through boosting profits. Controlling for other factors, large foreign banks appear to run a higher risk of insolvency. This is because foreign bank subsidiaries keep capital-to-asset ratios low in order to raise return on equity. Insurance against this risk is provided by parent banks, which typically can be expected to provide additional capital when needed.

  • Credit expansion, robust economic activity and strong profitability are the main determinants of banks’ foreign borrowing. The tightening of the maturity matching requirements has also contributed to the increase in banks’ foreign liabilities.

Since policy measures to curb credit growth have not been effective, it is even more important to strengthen prudential supervision to minimize the risks to financial stability. As large foreign subsidiaries tend to keep minimum amount of capital, supervisors should make sure that they build up adequate reserves that reflect the underlying risk through tightening provisioning requirements or requesting additional capital buffers. The results also highlight the importance of cooperating with home-country supervisors to closely monitor local subsidiaries of foreign banks.

Relaxing the maturity matching requirements could help lessen the reliance on foreign borrowing to some extent.

A. Introduction

111. Bosnia & Herzegovina has experienced rapid credit growth in recent years (Figure 1). Real private sector credit growth averaged 22.8 percent between 2001 and 2006, and credit to GDP ratio more than doubled during the same period. This experience is not unique to Bosnia & Herzegovina: financial deepening and rapid credit growth have been prominent features of recent economic developments in the Central and Eastern European (CEE) and Southeastern European (SEE) countries. In Bosnia & Herzegovina, the credit expansion so far seems to be in line with the regional trends, given the low initial level and healthy consumer demand growth.

Figure 1.
Figure 1.

Bosnia and Herzegovina: Credit Growth, January 2001–07

Citation: IMF Staff Country Reports 2007, 269; 10.5089/9781451804942.002.A006

Sources: CBBH and staff estimates

112. Foreign-owned banks, which are the main sources of credit growth in Bosnia & Herzegovina, rely on long-term foreign funding to finance this expansion. Most of the large foreign bank are subsidiaries of European banking groups, which set ambitious return on equity (ROE) targets for their CEE and SEE subsidiaries, including ones in Bosnia & Herzegovina. These subsidiaries finance their credit expansion by borrowing from abroad, mostly from their parent banks. In addition, in Bosnia & Herzegovina, the 2006 Financial Sector Assessment Program (FSAP) also suggested that the strict maturity matching requirements may have contributed to the reliance on foreign funding because the short maturity of local deposits limit their role in funding credit growth.

113. The rapid credit expansion, coupled with its foreign funding, raises concerns about both prudential and macroeconomic risks. During a credit boom, commercial banks’ credit assessment capacity may be overstretched and credit risks could be underestimated. Although the entry of foreign banks has improved the soundness of the financial system (IMF (2006a)), they have introduced a new macroeconomic risk, which could materialize if there were a sharp fall in their foreign borrowing or a reversal of capital flows. Such sharp fall could, for example, be a result of a shift in risk appetite of foreign banks. With the high current account deficit, a sudden stop or a slowdown of bank-related capital flow could put pressure on the balance of payment. In response to these risks, the authorities have implemented a number of measures aimed at curbing credit growth.

114. This paper empirically analyzes the relationships between credit growth, bank soundness, and banks’ foreign liabilities, and assesses the impact of the policy measures taken. Specifically, the paper addresses three main questions:

  • What is the relationship between credit growth and bank soundness? Do sound banks tend to lend more? Do those that lend more become less sound?

  • What are the determinants of banks’ accumulation of external liabilities? What is the relationship between credit expansion and bank-related capital inflows?

  • What have been the effects of the policy measures that were implemented with an intention to curb credit growth? Have they been effective in reducing rate of credit growth?

115. The rest of the paper is organized as follows. Section B describes recent developments in the banking sector in Bosnia & Herzegovina and sets them in the regional context. The Section also discusses prudential and macroeconomic risks associated with the credit boom. Section C describes the measures taken by the authorities to curb credit growth. Section D discusses the empirical model specification, describes the data, and presents the results. Section E concludes and provides policy implications.

B. Recent Developments in the Banking Sector

Developments in credit and market structure

116. The recent rapid credit growth mainly reflects the financial deepening process (Figure 3). Many studies suggest that countries in the CEE and SEE regions are going through a financial catch-up, where credit to GDP ratio is converging to the levels consistent with their incomes (Schadler and others, 2004; Backe and Zumer, 2005; Cottarelli Dell’Ariccia and Vladkova-Hollar, 2005, and Hilbers and others, 2005). In the case of Bosnia & Herzegovina, credit to GDP more than doubled between 2000 and 2006, and reached 52 percent in 2006. In fact, Bosnia & Herzegovina was categorized as a “late riser” in terms of credit growth, as its turning point only came in 2001, while credit in many CEE countries started its expansion path since late 1990s (Cottarelli, Dell’Ariccia and Vladkova-Hollar, 2003).

Figure 2.
Figure 2.

Central and Eastern Europe: Credit to GDP Ratio and Credit Growth, 2000–06

Citation: IMF Staff Country Reports 2007, 269; 10.5089/9781451804942.002.A006

Source: IMF.

117. Credit expansion has also been supported by favorable macroeconomic conditions. Improved economic prospects have boosted financial intermediation. Demand for credit has been fostered by economic growth, low inflation, and healthy wage growth. On the supply side, foreign-owned banks have been taking advantage of benign global conditions and low world interest rates to finance credit expansion with low cost funds.

118. Privatization and reforms of the financial sector have strengthened confidence in the banks. As elsewhere, privatization and reforms in the financial sector have far surpassed those in the corporate sector. The EBRD index for banking reform has improved 1.7 points since the late 1990s, compared with a 1 point increase for enterprise reform. With institutions such as the currency board arrangement and the deposit insurance in place, there is rising public confidence in the banking system. Deposit growth took off in 2001 with the annual rate of 65 percent, and maintained an average rate of 22 percent up to 2006. The ratio of broad money to GDP rose from 22 percent in 2001 to 57 percent in 2006.

119. Foreign participation and market concentration have increased markedly (Figure 3). The financial sector attracted about 40 percent of the foreign direct investment stock until end-2005. By end-2006, foreign-owned banks accounted for 94 percent of the system assets. In particular, four banking groups from the EU countries account for 74 percent of the system assets, following consolidation both at the country and the European level. The Herfindahl index (HI), which measures the degree of market concentration, shows that the market reached the upper end of “moderately concentrated” by end-Q3 2006. This development in Bosnia & Herzegovina is similar to the experience of many CEE and SEE countries. Figure 3 shows that the share of foreign-owned assets was greater than 50 percent in all but four selected CEE and SEE countries in 2004.

Figure 3.
Figure 3.

Bosnia & Herzegovina: Developments in Market Structure

Citation: IMF Staff Country Reports 2007, 269; 10.5089/9781451804942.002.A006

1/ A HI between 1,000 and 1,800 indicates moderate concentration. A HI above 1,800 indicates high concentration.Sources: CBBH; FBiH Banking Agency; RS Banking Agency; Mihaljek (2006); and staff estimates.

120. As in most CEE and SEE countries, the rapid credit growth has been led by lending to households (Figure 4). In Bosnia & Herzegovina, the average real growth of credit to household between 2001 and 2006 was about 50 percent, while it was only 13.5 percent for the credit to enterprises. Starting from a very low base, credits to household now account for 47 percent of total credit to private sector. Corporate lending only started to pick up in 2003–04. Most household loans are long term, and about 25 percent is housing-related credit, although commercial banks note that some consumer credits are also used for housing-related activities. Anecdotal evidence suggests that part of the household credit expansion has been led by a housing boom, although no aggregate housing price data are available. Competition to gain market share has led to a fall in lending interest rates.

Figure 4.
Figure 4.

Bosnia and Herzegovina: Key Features of the Credit Expansion

Citation: IMF Staff Country Reports 2007, 269; 10.5089/9781451804942.002.A006

Sources: CBBH; and staff estimates.

Financing of the credit expansion

121. At the aggregate level, both customer deposits and foreign borrowing have been the main sources of funding, but the large increase of net assets at the CBBH shows banks’ inability to intermediate much of short-term deposits (Figure 5). On the asset side, private sector credit to GDP grew by 25 percentage points while net assets at the CBBH grew 9 percentage points. A simple accounting exercise shows that the main sources of growth in assets are private sector deposits, which account for about 19 percentage points, and net foreign liabilities, which account for about 8 percentage points. Government net deposits and other items together account for 7 percentage points.

122. With cheap long-term funding from abroad, the existing maturity structure of loans and deposits, and tight maturity matching requirements, banks prefer to rely on long-term foreign funding to finance long-term credit. The regulations require that almost 100 percent of short-term (less than one year) liabilities must be matched with assets of the same maturities (see more details in Section C). While short-term deposits accounted for 56 percent of total deposits at end-Q3 2006, short-term lending accounted for only 24 percent of total lending. As the FSAP suggested, banks therefore put short-term deposits at the CBBH as reserves or invest abroad in short-term foreign assets. Long-term credits, which is 76 percent of total lending, have to be financed mainly by long-term foreign borrowing.

123. Bank-related inflows have been a major source of the increase in international reserves. Bank-related inflows averaged 2.7 percent of GDP from 2001 to 2006, and gross foreign liabilities reached 25 percent of GDP by end-2006. This is comparable to the level observed in Croatia, Slovenia and Lithuania. Foreign funding, typically comes in a form of euro-denominated loans or non-resident deposits, is exchanged for KM at the CBBH as most lending is in KM indexed to the euro. As a result, these inflows has contributed directly to the international reserves.

Key risks

124. Financial soundness indicators (FSIs), as illustrated in Table 1, suggest that the banking system is in good health. Profitability has improved over the years, capital adequacy ratios are high and nonperforming loan ratios are moderate and falling. Banks are highly liquid.

125. However, rapid credit growth, coupled with weaknesses in asset classification and provisioning, as identified by the FSAP, may mask the underlying credit risks. With the weaknesses in the regulation of credit risk management, commercial banks may currently underestimate the extent of credit risks. If these credit risks materialize, they would have a major adverse impact on banks’ profitability and capital adequacy, forcing banks to adjust their operating strategy accordingly. A downgrade in asset quality might also lead parent banks to curb lending to their Bosnian subsidiaries, which in turn could jeopardize financial and external stability.

Figure 5.
Figure 5.

Bosnia and Herzegovina: Financing of Credit Expansion

Citation: IMF Staff Country Reports 2007, 269; 10.5089/9781451804942.002.A006

Sources: CBBH; FBiH Banking Agency; RS Banking Agency; IMF; and staff estimates.
Table 1.

Bosnia and Herzegovina: Financial Soundness Indicators, 2001–06

article image
Source: CBBH.

126. In addition, indirect credit risks may also be significant. Most loan contracts contain adjustable interest rates with links to euro interbank offered rate (Euribor). Anecdotal evidence suggests that many borrowers may not realize the implications of such provisions in their contracts. Commercial banks also admit that with fierce competition to gain market share, they have not adjusted the lending rates, although the Euribor has risen recently. Moreover, some banks have started offering Swiss franc-indexed loans, besides the common euro-indexed loans. This could expose banks more to exchange rate-induced credit risks. Movements in exchange rates or interest rates could have a significant impact on borrowers’ capacity to pay.

127. Finally, the reliance on potentially volatile foreign funding to finance credit growth has introduced an additional risk to both financial stability and the macroeconomy. A fall in foreign funding could lead to a fall in international reserves. Under the currency board arrangement, this would have an immediate impact on the economy and could potentially lead to a sharp increase in non-performing loans. The stress tests performed by the FSAP based on June 2005 data suggest that a slowdown in credit growth and capital flows, coupled with a shift from KM into euro assuming declining confidence in the currency, could have a negative impact on financial stability. However, the system at the time appears resilient to such shock.

128. These risks may be aggravated by the fact that there is only a limited number of foreign parent banks operating in the CEE and SEE countries, raising the possibility of cross-border contagion (Table 2). The parent and subsidiary relationship is largely asymmetric for a small country like Bosnia and Herzegovina; Bosnia and Herzegovina’s portfolio only accounts for a very small share of parent banks’ total portfolio. Therefore, a small change in the parent bank lending policies can have a big impact on capital flows and lending in Bosnia and Herzegovina. IMF (2006b) examines the determinants of capital flows to CEE and SEE countries and how changes in these flows may be related to a foreign bank group’s overall exposure to the region.

Table 2.

Bosnia & Herzegovina: Main Foreign Banks

article image

End-Q3 2006

Sources: Banks’ websites; CBBH and staff calculation.

Measures taken to curb credit growth

129. In response to concerns about credit growth, the authorities have introduced several measures aimed at tightening the regulatory framework. First, banks were required to phase in tighter limits on their foreign exchange exposure from about 120 percent of capital in mid-2003 to 30 percent of capital by mid-2004. The definition of foreign currency assets and liabilities was also amended to include KM assets and liabilities indexed to foreign currency, in line with international practice. Second, banks were required to comply with the existing prudential regulations on maturity matching by June 2004, although some banks were allowed to delay compliance by September 2004 (Table 3).

Table 3.

Bosnia & Herzegovina: Maturity Matching Requirements

article image

130. The CBBH also tightened its monetary stance by expanding its required reserves base and increasing the reserve requirement (Table 4). In fact, the reserve requirement was raised three times between 2004 and 2005. Remuneration on excess reserves was also cut to 1 percent to discourage excess reserves accumulation.

Table 4.

Bosnia & Herzegovina: Changes in Reserve Requirements

article image
Source: CBBH

C. Empirical Investigation

Data and methodology

131. The relationships between credit growth, bank soundness, and banks’ external liabilities are explored using bank-level data. Bank-level data were obtained from the CBBH, the Banking Agency of the Federation of Bosnia & Herzegovina (FBiH Banking Agency) and the Banking Agency of the Republika Srpska (RS Banking Agency). The data are unpublished and include individual banks’ financial statements, ownership, and detailed classifications of assets and liabilities, as well as some of the prudential reports submitted to the regulators. Annual data are available from 2001–2005 (except for interest rates, data on which are only available starting in 2002) and quarterly data are available from Q2 2003 to Q3 2006. For more details on data definitions and sources, see Appendix I.

132. Bank soundness is measured by the risk of insolvency, or “distance-to-default”. This measure—widely used in measuring bank soundness—is directly related to the probability of loss exceeding equity capital, and thus captures the risk of insolvency. The measure is based on the option pricing model by Black and Scholes, and Merton.32 It can be summarized as:

DD=k+μσ

where µ is average return on assets, k is equity capital as percent of assets and σ is the standard deviation of returns on assets, a proxy for return volatility. Daily market data on equity combined with annual accounting data are typically used to calculate the market value and the volatility of assets. Without liquid equity market, however, market values may be biased. In this case, a simpler measure of distance to default based on balance sheet and income statements is used.

133. Credit growth is modeled as a function of bank soundness, macroeconomic and bank-specific factors, and policy variables, following Igan and Tamirisa (2006), Cihak and Tamirisa (2006), and Moreno-Badia (2007), who examine similar questions in the new EU member states and the Euro area. The baseline specifications for credit growth and bank soundness are:

CrediGri,t=αi+β1CreditGri,t1+β2DDi,t+β3DDi,t1+β4Macroi,t1+β5BankSpecifici,t1+β6Policyt+μi+εi,t(1)
DDi,t=αi+β2DDi,t1+β1CreditGri,t1+β3Macroi,t1+β4BankSpecifici,t1+β5Policyt+μi+εi,t,(2)

where i indexes bank, and t indexes quarters. CrediGri,t is real credit growth, DD is the measure of bank soundness—distance to default—for each bank in each observed quarter; Macro is a set of macroeconomic variables;33 BankSpecific is a set of bank-specific variables;34 Policy is dummies for policy measures; µi are bank-specific fixed effects and εi,t is a serially uncorrected error term. In the distance-to-default equation, the credit growth variable starts at its first lag; this is to capture the nonimmediate effect of credit growth on bank soundness. Both quarter-on-quarter and year-on-year credit growth are used for the credit growth variable. All control variables are in their first lags to avoid endogeneity.

134. The banks’ external liabilities are modeled as a function of credit, macroeconomic and bank-specific factors, and policy variables. The model is based on the empirical work on determinants of international bank lending (see Jeanneau and Micu (2002) for example). The baseline specification is

FLi,t=αi+β1FLi,t1+β2CreditGri,t1+β3Crediti,t1+β4Macroi,t1+β5BankSpecifici,t1+β6Policyt+μi+εi,t(3)

where FL is log banks’ foreign liabilities and Credit is log private sector credit. The two main variables are in log format so that the coefficients can be easily interpreted as elasticity.

135. The equations are estimated with the generalized method of moments (GMM) system estimator (“system GMM”). The estimator addresses the problems with independent variables that are not strictly exogenous, with fixed effects, and with heteroskedasticity and autocorrelation within individuals. System GMM as developed in Arellano and Bover (1995) and Blundell and Bond (1998) estimate a dynamic equation system by combining the original equation into the differenced equation, and uses the lagged difference variables as instruments for level variables. It was developed to improve efficiency in the difference GMM in Arellano and Bond (1991), or “difference GMM,” where the first difference of each variable is used to eliminate the fixed effect, and the lagged levels of variables are used as instruments.

136. One advantage of system GMM over differenced GMM is that it can capture the full effect of policy changes. When differenced, the policy dummies, which are equal to one only after the policy implementation, and equal to zero otherwise, will only show up as one only in the quarter that the policy is implemented. Difference GMM would therefore only estimate the initial effect of the quarter that policy was implemented. System GMM, by bringing back the level equation, is able to explain the full impact of policy measures.

137. Unit root test for panel data shows no sign of specification problems. Maddala and Wu (1999) tests for panel unit roots show that no series used in the empirical investigation has a unit root.

Results

138. Tables 5 to 7 report the estimation results. The two-step robust estimates with standard errors that are corrected for finite sample bias (the Windmeijer correction) are reported. All lags are used for each of the endogenous variables. All regressions show no sign of second-order serial correlation, and the Hansen test for overidentification is not rejected. Adding the second lag of the explanatory variables (for credit growth in equations (2) and (3) in particular) does not add any more information, and is therefore not included in the equations.

Credit growth

139. Although the coefficients are small and not statistically significant, sounder banks seem to have higher credit growth, (Table 5).35 Both distance-to-default and its first lag have positive but statistically insignificant coefficients. Surprisingly, the real interest rate has a positive and significant impact on credit growth, while more robust economic activity, proxied by real retail sales growth, has a negative, though insignificant, impact. Although these results seem to contradict conventional wisdom, they explain the initial jump in credit growth at the beginning of the period when interest rates were high.

140. The policy measures do not have a statistically significant effect on credit growth. The results confirm the nonbinding nature of each increase in reserve requirement observed in Figure 5. The dummy variables for the policy changes in September 2004 (which include both the tightening of maturity matching and the increase in reserve requirements), December 2004, and December 2005 have positive but statistically insignificant coefficients.

Bank soundness

141. The annual credit growth seems to have no significant impact on bank soundness, while the more immediate quarter-on-quarter credit growth seems to boost bank soundness (Table 6 and 7). In both sets of equations, bank soundness is also found to have positive persistence—banks that are sound today are likely to be sound tomorrow. As expected, a higher NPA ratio has a negative impact on distance-to-default, although this result is statistically significant only in one specification. Higher interest rates have a negative although statistically insignificant impact on bank soundness, perhaps reflecting higher indirect credit risk in the face of higher rates that is not offset by higher profitability. Bank vulnerability decreases with an increase in inefficiency, or cost-to-income ratio, and the result is statistically significant.

142. Controlling for other factors, large foreign banks, appear to have higher risk of insolvency. Large foreign banks have higher risk of insolvency, especially banks that are subsidiaries of European banking groups. The interaction terms between foreign ownership and size, and between EU bank ownership and size show negative and statistically significant coefficients. A possible explanation is that these subsidiaries, though enjoying strong profits, can afford to keep capital-to-asset ratios low in order to raise reported ROE, since the parent banks typically can be expected to provide additional capital when needed (IMF, 2006a). In turn, the positive and statistically significant coefficient for the state-ownership dummy reflects the high capital ratios of state-owned banks, although their profitability is low.

143. The increase in the reserve requirement in December 2004 may have had a negative impact on bank soundness. Other policy changes, however, have had no significant effect on banks’ risk of insolvency.

Table 5.

Credit Growth Regressions

article image

indicates significant at 10 percent;

significant at 5 percent

significant at 1 percent.

Table 6.

Distance to Default Regressions, with year-on-year Credit Growth

article image

indicates significant at 10 percent;

significant at 5 percent.

significant at 1 percent.

Table 7.

Distance to Default Regressions, with Quarter-on-Quarter Credit Growth

article image

indicates significant at 10 percent;

significant at 5 percent.

significant at 1 percent.

Banks’ Foreign liabilities

144. Credit expansion, robust economic activity, and strong profitability all positively affect banks’ foreign borrowing (Table 8). If credit grows by 1 percent both in the last quarter and in the current quarter, banks’ foreign liabilities will increase by between 0.3–0.7 percent. In other words, the elasticity of foreign liabilities to credit is about 0.3–0.7. This finding confirms the notion that European parent banks are seeking higher returns by investing low-cost funds from home into their Bosnian subsidiaries where interest rates and profit margins are higher than what they could get in the EU market. Controlling for other factors, ownership does not have a significant impact on the level of foreign liabilities.

145. In addition, the empirical results show that the tightening of maturity matching requirements has contributed to the increase in banks’ foreign liabilities. Controlling for other factors, the dummy for this policy measure shows positive and significant coefficients. The coefficient, however, is not large, compared with the effects of other economic factors such as credit, economic activities, interest rate, and interest margin.

146. The results yield an empirical estimate of the link between credit, foreign liabilities, and international reserves. This estimate is useful for its macroeconomic implications. If credit growth slows down to zero in two quarters, foreign liabilities would fall by around 7–16 percent. Assuming a one-to-one fall in international reserves and taking the maximum value, this would translate to a decline of €864 million in international reserves, or from 5.1 months to 4.3 months of imports (end-2006 figures). The magnitude of this shock is similar to the assumption used in the FSAP stress tests.

Table 8.

Banks’ Foreign Liabilities Regressions

article image

indicates significant at 10 percent;

significant at 5 percent

significant at 1 percent.

The results in regional perspective

147. Unlike the empirical results from similar investigations in the new EU member states (NMS) and Bulgaria, the results for Bosnia & Herzegovina show that credit growth seems to have a positive and immediate impact on bank soundness (see Appendix Table). Igan and Tamirisa (2006) find that credit growth has statistically insignificant effect on bank soundness in the NMS, while Herderschee and Ong (2006) find that banks with positive loan growth strategies have higher solvency risk on average in Bulgaria. The results from Bosnia & Herzegovina are similar to the findings for Greece and the euro area, where the first lag of credit growth has a positive impact on bank soundness (Moreno-Badia (2007)).

148. Large foreign banks in Bosnia & Herzegovina seem to run a higher solvency risk than their domestic counterparts, again unlike those in the NMS and Bulgaria. According to Igan and Tamirisa (2006), foreign ownership has a positive impact on distance-to-default for banks in NMS. The same is true for Bulgaria, where locally-owned private banks are more vulnerable to shocks than foreign-owned subsidiaries.

D. Conclusions and Policy Implications

149. The process of financial deepening in Bosnia and Herzegovina—in line with the regional trends—has been supported by favorable economic conditions and the consolidation and competition in the banking sector. The banking system is well-capitalized and profitable, although inherent risks exist in an environment of rapid credit expansion.

150. Since policy measures to curb credit growth have not been effective, it is even more important to strengthen prudential supervision. The empirical investigation shows that the increases in the reserve requirement have had no statistically significant impact on credit growth. The authorities thus have only prudential tools to minimize the risks to financial stability.

151. As large foreign subsidiaries tend to keep the minimum amount of capital, supervisors should make sure that they build adequate reserves to cover the underlying risks. This could be done by tightening provisioning requirements or requesting additional capital buffers. The banking agencies should urgently address the weaknesses in asset classification and provisioning rules as recommended by the FSAP. Moreover, they could request additional capital requirements for market risks, such as foreign exchange risks (see examples in Appendix III).

152. The results also highlight the importance of cooperation with home-country supervisors. Because large foreign banks are keeping a minimum amount of capital and rely on capital injections from their parent banks if the risks materialize, it is important that the supervisors closely monitor the health of the parent banks. Although all parent banks of large subsidiaries in Bosnia & Herzegovina are reputable and have high credit ratings, future capital injections would depend on the financial performance of the parent banks at that time.

153. The links between foreign borrowing, credit expansion, and the role of maturity matching requirement are empirically established. Credit expansion, robust economic activity, and strong profitability all positively affect banks’ foreign borrowing. In addition, the empirical results also show that the tightening of the maturity matching requirements have contributed to the increase in banks’ foreign liabilities.

154. Relaxing the maturity matching requirement could thus help lessen banks’ reliance on foreign funding. The maturity matching requirement is strict by international standards (see Appendix III for more details), and it has not encouraged domestic intermediation. Relaxing it could help lower foreign borrowing to some extent, as such borrowing is driven mainly by other economic factors.

Appendix I—Data Sources and Definition

Bank-level data were obtained from the CBBH, the FBiH Banking Agency and the RS Banking Agency. Macroeconomic data were taken from the Federation Statistics Office and the RS Institute of Statistics.

To ensure that the analysis was not affected by potential mismeasurement and misreporting, observations on the tails of the distributions of credit growth and distance to default were dropped.

article image
article image

Appendix II— Regional Comparison of Empirical Results

article image

Appendix III— Prudential Regulation: International Examples

As recommended by the FSAP, comprehensive guidelines on foreign exchange risk and other market risks should be introduced in Bosnia & Herzegovina. The foreign exchange exposure limit is in line with the practice in other countries with dollarized financial systems (Appendix Table 1). However, the supervisors should also require capital for foreign exchange exposure. This practice has become increasingly popular among regulators as it makes it more difficult for weakly capitalized banks to take on new risks. Capital requirements give a bank greater flexibility in choosing the risks it will accept by allowing managers to allocate a bank’s capital between credit and market risk, including foreign exchange risk (Cayazzo and others, 2006).

Table 1.

Foreign Exchange Risk Requirements: Selected Countries

article image
Source: Cayazzo and others (2006)

The maturity matching requirements in Bosnia and Herzegovina are strict by international practice. The banking agencies in Bosnia use a minimum liquidity ratio and maturity matching requirements to regulate banks’ liquidity. For the minimum liquidity ratio, banks have to maintain average 10-day minimum liquidity in cash funds up to at least 10 percent of short-term funds sources according to book value. Country practices vary but most countries with partially dollarized banking systems have laxer requirements (Appendix Table 2). Only Romania seems to have a system as strict as Bosnia’s while Chile has a 100 percent requirement but it only applies to assets and liabilities of up to 30 days. In Slovenia, although the coverage must be at least 100 percent, demand deposits are given a weight smaller than 100 percent.

Table 2.

Liquidity Requirements: Selected Countries

article image
Sources: Cayazzo and others (2006); and Hayward and Byskov (2005).

References

  • Arellano, M. and S. Bond, 1991, “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations,Review of Economic Studies, 58, 277 -297.

    • Search Google Scholar
    • Export Citation
  • Arellano, M. and O. Bover, 1995, “Another Look at the Instrumental-Variable Estimation of Error-Components Models”, Journal of Econometrics, 68, 29 -51.

    • Search Google Scholar
    • Export Citation
  • Backe, P. and T. Zumer, 2005, “Developments in Credit to the Private Sector in Central and Eastern European EU Member States: Emerging from Financial Repression—A Comparative Overview,” Focus on European Economic Integration No. 2/05 (Vienna: Austrian National Bank).

    • Search Google Scholar
    • Export Citation
  • Blundell, R. and S. Bond, 1998, “Initial conditions and moment restrictions in dynamic panel data models,” Journal of Econometrics, 87(1), 115 -143

    • Search Google Scholar
    • Export Citation
  • Bond, S., 2002, “Dynamic Panel Data Model: A Guide to Micro Data Method and Practice,” CEMMAP Working Paper No. 09/02 (London: Center for Micro Data Methods and Practice).

    • Search Google Scholar
    • Export Citation
  • Cayazzo, J., A. Pascual, E. Gutirrez, and S. Heysen, 2006, “Toward an Effective Supervision of Partially Dollarized Banking Systems,” IMF Working Paper No. 06/32 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Chan-Lau, J., 2006, Fundamentals-Based Estimation of Default Probabilities: A Survey,” IMF Working Paper No. 06/149 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Cihak, M. and N. Tamirisa, 2006, “Credit, Growth and Financial Stability,” Poland Selected Issues, IMF Country Report No. 06/392 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Cottarelli, C., G. Dell Ariccia and I. Vladkova-Hollar, 2003, “Early Birds, Late Risers, and Sleeping Beauties: Bank Credit Growth to the Private Sector in Central and Eastern Europe and the Balkans,” IMF Working Paper No. 03/213 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Hayward, P. and S. Byskov, 2005, “The Regulation and Supervision of Liquidity,” (Washington: International Monetary Fund). unpublished.

    • Search Google Scholar
    • Export Citation
  • Herderschee, J. and L. Ong 2006, “Bulgaria—The Implications of Bank Behavior and Credit Measures for Solvency Risk,” Bulgaria—Selected Issues, IMF Country Report No. 06/299 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Hilbers, P. and others, 2005, “Assessing and Managing Rapid Credit Growth and the Role of Supervisory and Prudential Policies,” IMF Working Paper No.05/151 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Igan, D. and N. Tamirisa, 2006, “Credit Growth and Bank Soundness in the New Member States,” Czech Republic, Republic of Estonia, Hungary, Republic of Latvia, Republic of Lithuania, Republic of Poland, Slovak Republic, and Republic of Slovenia—Export Structure and Credit Growth, IMF Country Report No. 06/414 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 2006a, “Bosnia and Herzegovina—Financial System Stability Assessment,” IMF Country Report No. 06/403 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 2006b, “Role and Risk Implications of Foreign Owned Banks,” Bosnia and Herzegovina FSAP Technical Note, unpublished.

    • Search Google Scholar
    • Export Citation
  • Jeanneau, S. and M. Micu, 2002, “Determinants of International bank Lending to Emerging Market Countries,” BIS Working Papers No. 112 (Basel: Bank of International Settlements).

    • Search Google Scholar
    • Export Citation
  • Mihaljek, D., 2006, “Rapid Cgrowth of Bank Credit in Central and Eastern Europe: the Role of Housing Markets and Foreign-Owned Banks,” the Twelfth Dubrovnik Economic Conference, June 28-July 1, 2006.

    • Search Google Scholar
    • Export Citation
  • Moreno-Badia, M., 2007, “Credit Growth and Bank Vulnerability in the Euro Area and Greece,” Greece—Selected Issues, IMF Country Report No. 7/27 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Schadler, S., and others, 2004, Adopting the Euro in Central Europe: Challenges of the Next Step in European Integration, IMF Occasional Paper No. 234 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
31

Prepared by Mali Chivakul.

32

See for example Chan-Lau (2006) for derivation and further explanation.

33

The original set of macroeconomic variables considered was real industrial production growth, FBiH real retail sales growth, real interest rates and real exchange rate depreciation. Quarterly GDP or GDP per capita is not available, and the official unemployment rate may be overestimated.

34

The original set of bank-specific variables considered was measures of profitability (net interest margin), efficiency (cost-to-income ratio), liquidity (loan to deposit ratio), risk (nonperforming asset ratio), ownership (state ownership, foreign ownership, and EU bank ownership dummies) and size (log (total assets)).

35

Only year-on-year credit growth is used here as quarter-on-quarter credit growth does not pass the test for AR(1), i.e. no serial correlation is observed in the first difference

Bosnia and Herzegovina: Selected Issues
Author: International Monetary Fund