Chapter 3. Financial Sector1: Deepening, Resilience, and Ongoing Challenges
- International Monetary Fund. European Dept.
- Published Date:
- March 2015
In the years since 2000, transition and transformation in the Western Balkans have been particularly significant in the banking sector. These banking systems have undergone significant financial deepening, more so than did those of the New Member States at the same stage of economic transition. In the run-up to the global financial crisis, the banking systems of the Western Balkans relied less on fast-moving wholesale funding than did the New Member States (with the exception of Montenegro), which suggests that a significant part of the precrisis credit expansion in these countries was perhaps part of a long-term trend of financial deepening. But financial development in the Western Balkans over this period has also been uneven. While banking sectors have developed rapidly, growth of nonbank financial services has been lackluster, with equity, pension, and insurance markets remaining shallow and corporate debt markets largely nonexistent, even today.
The period of sustained expansion of bank credit in the Western Balkans also coincided with the global financial boom that ended in 2008. The subsequent sharp fall in credit growth in the aftermath of the global financial crisis raises the question: was there a credit boom-bust cycle at play, and was the precrisis growth of the banking sector unsustainable? The analysis here suggests that while credit growth in the Western Balkans was procyclical prior to 2008, and appears to have exceeded somewhat what could have been sustained by these economies’ underlying structural characteristics, it was still milder, with the exception of Montenegro, than that in many of the New Member States, particularly the Baltic countries. The postcrisis credit correction in Western Balkan countries was not as severe as in the New Member States because the former relied more on expanding domestic deposit bases than did the latter.
But the fact that the credit cycle was milder has not meant that the Western Balkan countries were able to avoid the after-effects of the crisis. The fallout from the crisis and the subsequent economic sluggishness has resulted in a surge in nonperforming loans (NPLs) in most Emerging European countries. But tepid structural reforms in the Western Balkans to address these problems head on have left these countries with a legacy of higher level NPLs today than in their counterpart New Member States.
In the Western Balkans, households and firms do not have good access to credit and saving instruments; this is particularly true in lower-income countries in the region, such as Albania and Kosovo. The Western Balkan banking model has tended to rely on high interest margins, which is, among other things, an indicator of inefficiency and weak intermediation. While these margins have been on a declining trend and appear to be converging to New Member States levels, they remain wide. The extent to which the Western Balkan banking systems are able to support sustained growth over the medium term will depend not only on further financial deepening, but also on improvements in financial access and efficiency—that is, the ability to reach traditionally underserved segments in a cost-effective manner.
There is a sizable unfinished agenda of structural reforms in the financial sectors in the Western Balkans. While these countries have done relatively well in providing the infrastructure necessary for financial deepening, they have lagged their New Member State peers in strengthening the foundations of financial stability. Despite some attempts at reform, the regulatory and institutional environment in the Western Balkan countries, and the attendant enforcement, is not yet well suited for tackling what is arguably the most serious challenge today facing the region’s banking system—the NPL problem.
The global financial crisis of 2008 and the subsequent shocks emanating from Advanced EU economies have underscored the need for coordinated cross-border efforts to mitigate the particular financial risks to which Western Balkan countries are vulnerable. In this regard, the Vienna I and II Initiatives spearheaded by international financial institutions, including the IMF, have been useful, and continue to play an important role. IMF lending programs in the region have also helped.
A commonly-drawn lesson from the financial crisis is that to mitigate the risk of contagion from abroad, Emerging Europe’s banking systems should transition toward increased reliance on local-based funding and less reliance on foreign funding (IMF 2013). We note below that the share of foreign funding in Western Balkan banking systems—which also experienced a generally milder boom-bust cycle—was relatively small both before and after the global crisis. This, together with the relatively large investment needs that cannot be met by domestic financing alone, implies that the share of foreign financing, though not necessarily of banking systems, will actually need to increase in the Western Balkans. To attract the needed external financing, Western Balkan countries need to undertake significant structural reforms to improve access, reduce inefficiencies, and strengthen stability in their financial sectors. Foreign direct investment should form a large part of these external inflows, but there is also room for a sizable increase in foreign financing flows through the financial systems of these countries without making them overly reliant on foreign funding and the associated volatility.
A. Financial Deepening from a Low Base
Deep financial systems support sustained economic growth and macro-financial stability mainly through efficient allocation of resources between savers and borrowers, and by allowing economic agents to smooth consumption and overcome risks (Levine 2005; IMF 2012). Financial depth indicates the amount of financial services available in an economy, and is often proxied by the magnitude of assets or liabilities of financial institutions relative to the size of the economy. It tends to increase with economic growth and development, though it can also vary among countries at similar levels of income and market size, because of differences in macroeconomic stability, institutional strength, or the impact of past events, such as crises or wars.
Bank Assets and Income per Capita
Source: World Bank Global Financial Development Database (GFDD).
Over the course of economic development, financial deepening first occurs through the spread of banking services, and then involves increasing use and provision of nonbank financial services—capital, pension, and insurance markets. Nonbank deepening begins to takes place once challenges related to information, enforcement, and coordination are overcome and there is sufficient demand for sophisticated financial services (Pagnano 1993; De la Torre, Feyen, and Ize 2011).
The evolution of financial deepening in the Western Balkan countries so far appears to be following the standard pattern. These countries have advanced rapidly with bank deepening, perhaps more so than the New Member States at a similar stage of economic transition. However, the Western Balkan countries have lagged behind their peers in deepening their nonbank financial sector, as capital markets (equities and bonds), pensions, insurance, and other financial markets remain nonexistent or severely underdeveloped in many of the Western Balkan countries, except perhaps in Croatia. Box 3.1 discusses the state of nonbank financial deepening in the Western Balkan countries.
A Rocky Start to Bank Credit Deepening
The process of bank deepening in Emerging Europe in the 1990s was not smooth. Countries found it challenging to establish market-based financial systems because of the need to simultaneously undertake macroeconomic stabilization as well as financial and operational restructuring of banks and firms. New legal and institutional frameworks also had to be put in place (Barisitz 2009). As transition progressed, these countries began to open up their banking sectors through privatization, in many cases by applying liberal licensing criteria and insufficient supervision. Not surprisingly, the outcome was a wave of banking crises in a number of these countries (Roaf and others 2014; IMF 2013), with fewer but more costly crises in the Western Balkan countries than in the New Member States.2 Nonetheless, one positive outcome of the turbulent 1990s was a round of banking consolidation and increasing penetration of foreign banks, which brought professionalism, know-how, and arm’s length relations with borrowers. Not surprisingly, this set the stage for sustained financial deepening in the subsequent decade.
Banking Crisis Incidence
Toward Advanced Transition and Bank Credit Deepening
How does the pace of credit deepening in the Western Balkans compare with that in the New Member States? Such a comparison requires controlling for the stage of economic transition. As the two sets of countries began transitioning to a modern economic system almost a decade apart, a comparison of their relative progress necessitates putting them on an appropriate timeline. We use the EBRD’s transition index to determine when these countries reached a “fairly advanced stage of transition.” This is defined as a country having achieved macroeconomic stability and progressed with structural economic reforms—or reaching the EBRD transition level of 3 or above (see chart). Fairly advanced transition was broadly under way in the median New Member States by around 1994.3 The Western Balkan countries reached a similar level of transition much later with a sizable intra-group variation. While advanced transition began prior to the 2000s in Croatia and FYR Macedonia, it began only in 2002 in the rest of the Western Balkan countries, after conflicts in the region had been settled. Our threshold analysis of comparing the evolution of financial deepening in the two sets of countries thus uses 1994 and 2002, respectively, as the starting years of advanced transition for the two country groups (i.e., for setting time t =0).4 As can be seen in the text table, the median Western Balkan country had slightly lower domestic banking asset depth than the New Member States at the same stage of transition (excluding the Baltics), but also lower comparable GDP per capita (except Bulgaria and Romania).
State of Transition
1/ Excludes Czech Republic.
|Year reached fairly advanced transition 2/||2002||1994||1994||1994||1994|
|EBRD Transition Index3/||3.2||3.5||3.5||3.5||2.1|
|GDP per capita (constant 2000 U.S. dollars)||2,570||4,866||3,545||7,314||2,899|
|GDP per capita (current U.S. dollars)||1,827||2,410||1,903||3,891||1,236|
|Bank assets to GDP||28.2||33.0||16.5||36.5||96.0|
|Bank credit to the private sector to GDP||19.8||20.3||14.5||22.6||48.3|
|Bank deposits to GDP 4/||39.0||30.9||17.0||44.6||44.6|
Our transition threshold analysis suggests that the Western Balkan countries’ banking systems deepened more rapidly than did those of the New Member States in the respective period following the beginning of advanced transition. The ratio of bank credit to the private sector as a percent of GDP, a widely-used measure of credit depth, grew relatively slowly in the New Member States in the five years following the start of advanced transition (i.e., during 1994–99).5 On the other hand, Western Balkan countries’ private credit-to-GDP ratio grew briskly after these countries had reached advanced transition and increased by a cumulative 30 percent of GDP in 10 years.
Credit Deepening 1/
Source: World Bank FinStat Database.
1/ Change since average transition index >3 (1994 for NMS and 2002 for WBS).
Source: World Bank FinStat Database.
The rapid credit deepening in the Western Balkan countries also coincided with the global financial boom of the mid-2000s, which raises the question of whether credit growth was excessive and resulted in the buildup of vulnerabilities. A number of factors suggest that there was a lesser buildup of vulnerabilities and a milder boom-bust cycle in the Western Balkan countries than in the New Member States (the latter issue is discussed in fuller detail in the next section). First, much of the credit growth in the Western Balkans coincided with a rapid widening of the domestic deposit base—faster than what the New Member States experienced in the period following their advanced transition. Large inflows of remittances from abroad, as well as high real interest rates (except in Serbia) may partly explain the deposit growth.6 More importantly, the rapid entry of foreign banks and the introduction of deposit insurance schemes during the 2000s likely helped boost the confidence of depositors in the local banking sectors, attracting new deposits.
Foreign Bank Ownership
B. Credit Swings and Financial Stability
Sustained expansion of credit is a sign of financial deepening, but the speed of this expansion also matters. Excessively fast credit growth—a credit boom—can lead to resource misallocation and macroeconomic imbalances through looser lending standards, excessive leverage, greater credit risk, and/or asset price bubbles, which in turn can trigger or exacerbate financial crises (Loayza and Ranciere 2006; Rousseau and Wacthel 2011; Barajas and others 2013). Credit busts are painful. They expose vulnerabilities built up in boom years, damage the balance sheets of corporates and households and reduce their ability to repay loans, and, as a consequence, spill over to banks’ books. The fallout is more severe if it is accompanied by an asset price bust. Recovery from credit busts depends, among other things, on the quality of existing resolution frameworks and the forcefulness of the government’s crisis response. Not surprisingly, credit busts often lead to protracted weakening in economic activity, and in some instances even a balance sheet recession (Koo 2011).
Credit Expansion Prior to the Global Crisis
The existing evidence suggests that in the run-up to the global financial crisis in 2008, the Western Balkan countries’ credit cycle was milder than that in New Member States. The increase in credit in the Western Balkans during 2003–07 was notably smaller than in the Baltic States, Bulgaria, and Romania. But credit expansion was not mild in all the Western Balkan countries. Countries with initially shallower credit bases—Montenegro, and to a lesser extent, Albania and Kosovo—experienced a significant and prolonged credit expansion, as measured by the change in the credit-to-GDP ratio while countries with relatively deeper credit bases at the outset—Croatia and Bosnia and Herzegovina—saw more moderate growth in credit. In this context, a number of Western Balkan countries undertook prudential measures to contain rapid credit growth, though with mixed success (see Box 3.2 for the case of Croatia).
Credit to GDP Ratio
Source: IMF’s International Financial Statistics Database.
1/ Group medians.
Credit Cycle in Emerging Europe
Source: IMF, International Financial Statistics Database; and IMF staff calculations.
Note: The chart shows deviations in credit from its trend value in relation to the typical credit expansion over the business cycle. See Mendoza and Terrones (2008) for more deatails in the methodology.
We employ two approaches to assess the severity of the credit expansion. The first one defines a credit boom as an episode during which credit grows at a pace faster than what would be warranted by a cyclical economic expansion (Mendoza and Terrones 2008). It estimates deviations in credit from its trend value and compares it to where credit would typically have been over the business cycle. We find that the deviation of credit growth from its long-run trend was smaller in Western Balkan countries than in the New Member States7 at the peak of the credit boom (i.e., at t=0) and in the quarters leading up to the peak. Of course, the latter are a heterogeneous group. The boom-bust in the Baltics was particularly severe, but the credit cycle in the five Central European countries resembled that in the Western Balkan countries. Kosovo, Montenegro and FYR Macedonia saw the largest credit boom, while Croatia and Serbia experienced the mildest.
The second approach in judging if credit growth was excessive in this period is to assess if it outstripped economic fundamentals and structural factors. This entails following an approach developed by the World Bank, which benchmarks credit developments for each country against the level of its economic development (proxied by GDP per capita), size of the economy, and other structural factors such as size, density, and age profile of the population. The results (also see Annex 3.1) suggest that after 2005 both Western Balkan countries and the New Member States saw private credit growth that was above the respective norms—that is, beyond what could have been sustained by these economies’ underlying structural characteristics.8 Again, as in the cyclical analysis, the excess expansion at its peak was substantially milder in the Western Balkan countries than in the New Member States, particularly the Baltics. There was sizable variation among the Western Balkan countries—the increase in the credit-to-GDP ratio in Albania and FYR Macedonia was twice or more than what is predicted by the model, while Serbia’s credit growth appears to have been insufficient to meet the economy’s development and growth needs during the period.
Bank Credit to Private Sector in WBS
Source: World Bank FinStat Database.
1/ Does not include UVK or MNE; includes BIH and SRB only from 1997
Bank Credit to Private Sector in NMS
Source: World Bank FinStat Database.
The degree of reliance of the pre-2008 credit growth on domestic deposits and foreign financing provides insights into whether credit developments in this period were sustainable or transitory. As discussed earlier, loan growth in the Western Balkan countries remained more dependent on expansion of the domestic deposit base than did loan growth in the New Member States (Figure 3.11). This was particularly so in Albania and Kosovo, while at the other end of the spectrum, Montenegro’s credit expansion relied overwhelmingly on external sources. The rapidly growing presence of foreign banks in Western Balkan countries did allow for fast credit expansion that went beyond the available sources of domestic financing, as indicated by the increase in the loan-to-deposit ratio during the precrisis years, particularly so in the couple of years prior to the outbreak of the global financial crisis. But with the exception of Bosnia and Herzegovina, the external funding that financed credit expansion in the Western Balkans came largely from parent banks in the form of increasing capitalization of local banks in Croatia and Serbia it was the result of policy action that induced banks to do so (see Box 3.2). On the other hand, in the Baltics, the relatively greater sophistication of domestic financial markets, as well as strong connectedness with parent banks and sources of wholesale funding allowed swift and heavy capital inflows through the banking systems (IMF 2013).
Evolution of Deposits
(Percent of GDP) 1/
Sources: World Bank FinStat Database; and IMF, International Financial Statistics.
1/ Groupings calculated with medians.
Source: IMF, International Financial Statistics Database.
1/ Groupings were calculated with medians.
2/ Group exludes MNE.
Nature of Foreign Banks Non-deposit Funding
Source: BIS and IMF, World Economic Outlook Databases.
Exchange rate regimes appear to have mattered as well, both in the Western Balkan countries and in the New Member States. Economies with fixed exchange rates generally received larger bank flows than those with more flexible currencies—in particular Montenegro among the Western Balkan countries, and the Baltics and Bulgaria among the New Member States (Slavov 2009). The expectation of low currency risk attracted external flows into these economies. Albania saw an increase in external claims as well, but these were channeled through the banking system into its public debt market. In general, the more fixed the exchange rate regime, the bigger the credit expansion and buildup of financial vulnerabilities in the period immediately before 2007 (except in FYR Macedonia). Economies with flexible exchange rates were better able to withstand the capital inflow pressures, and thus avoided building up excessive vulnerabilities.
The extent of external exposure of the banking system is associated with the size of foreign currency lending in the economy. A positive cross-country association exists between foreign currency loans in relation to GDP extended by the local banks and the extent of exposure that the country’s financial system has to banks from Advanced EU economies, also in relation to GDP. The Western Balkan countries, with their relatively low external exposure, also rank lower in terms of the size of foreign currency loans, though Croatia is a notable exception. Among the New Member States, there is a wide dispersion: at one end is the group of Baltic states—particularly Latvia and Estonia—with a heavy exposure to foreign banks and a high share of foreign currency loans, and at the other are the remaining countries that do not appear to differ much from the Western Balkans.
Foreign Bank Financing and Private Sector Credit
Sources: BIS locational statistics and IMF, International Financial Statistics Database.
Note: Montenegro data start in 2006. Blue boxes represent countries with fixed or pegged exchange rate regimes, or those using the euro. Countries in red circles represent Western Balkan states.
Foreign Bank Exposure to Domestic Banks and Foreign Currency Lending, 2008
Source: IMF staff calculations.
Postcrisis Credit Slowdown
Credit growth slowed sharply across Emerging Europe in the aftermath of the 2008 global financial crisis and the subsequent economic troubles in Europe (Figure 3.16). However, the credit correction was generally milder in Western Balkan countries than that experienced by the New Member States. In the former, credit continued to grow in real terms at an annual average of about 6½ percent, while in the latter it largely stalled, pulled down by a sharp contraction in the Baltic States.9 The Central European economies experienced a moderate postcrisis correction, similar to the Western Balkans. Not all Western Balkan countries were identical, however. Montenegro, which underwent what was arguably the most severe credit boom in Emerging Europe, experienced the deepest fall in credit as well. In Croatia and Bosnia and Herzegovina credit growth stalled, while in Albania, Kosovo, and FYR Macedonia credit has continued to grow until recently.
Credit Boom and Credit Bust
Sources: BIS and IMF, International Financial Statistics Database.
Note: yoy = year-on-year percent change.
In general, the drop in credit growth after the global financial crisis was closely linked to the magnitude of the precrisis boom—the larger the credit boom, the more severe the credit correction. We employed monthly data to get a granular feel of the severity of the fallout, as annual information tends to mask sharp intra-year movements. Our calculation of peak-to-trough decline in real credit growth reveals that, with the exception of Montenegro (and to some extent Albania), the drop in credit growth from precrisis peak to the postcrisis trough was less severe in the Western Balkan countries than in the New Member States.10
The sources of precrisis credit boom also seem to be relevant in explaining the postcrisis credit slowdown. Heightened global risk aversion in the aftermath of the 2008 crisis triggered a sharp reduction in capital flows as parent banks deleveraged from the region. The funding shrinkage led to a large decline in credit growth in countries that had depended heavily on external funding inflows to finance their booms. Here again, the relatively mild after effects of the crisis in Western Balkan countries than in New Member States, particularly the Baltics, can be partly ascribed to the smaller precrisis exposure to fast-moving external financing in the former.
Not surprisingly the credit cycle also affected economic activity. The intensity of the postcrisis credit slowdown was closely associated with that in economic growth, largely because of demand effects. As funding dried up, risk aversion increased and credit conditions tightened, with the crunch in credit in some countries associated with severe demand contraction. Those countries that experienced a more intense boom-bust episode also endured sharper growth contractions. Western Balkan countries fared relatively well, as they saw relatively mild growth slowdown or contraction, with the exception of Montenegro and Croatia.
Credit Growth and Output Growth
Source: IMF’s International Financial Statistics Database.
Nonperforming Loans: a Difficult Legacy of the Credit Cycle
The post-2000 credit cycle in most of the Western Balkan countries was relatively moderate, but the aftermath of the global financial crisis and the subsequent economic slowdown has resulted in a high level of NPLs. Ample loanable funds, and relatively lax lending standards, particularly just prior to the outbreak of the 2008 global crisis, sowed the seeds of subsequent asset quality deterioration. NPLs have risen sharply in countries where the precrisis credit boom was intense and where the postcrisis economic slump was deep.
The postcrisis NPL surge was somewhat larger in the New Member States than in the Western Balkan countries—particularly in the Baltic States—but the former were also able to reduce NPLs substantially from the immediate postcrisis peak, particularly because of comprehensive reforms to address corporate NPLs. On the other hand, the Western Balkan countries have been largely unable to progress forcefully in this area. Where reforms have taken place—for example, recent improvements in the tax treatment of loan write-offs and steps to improve collateral execution in Albania—there has been little improvement on the ground to date because of weak enforcement. The use of factoring companies in Montenegro to offload bad loans from banks’ books has also met with little success, as loan workouts have been few and corporate balance sheets remain weak. Not surprisingly, the median level of NPLs in Western Balkan countries has persisted at the postcrisis peak (see Box 3.3; and Liu and Rosenberg, 2013).
Sources: IMF, Financial Soundness Indicators Database; and country authorities.
Note: Group medians. Definition of NPL may differ by country.
Bank Provisioning for NPLs
Sources: Financial Soundness Indicators (FSI); and IMF staff calculations.
1/ Total provisions (specific plus general)/NPLs. Definition of NPL and coverage ratio may differ across countries. Horizontal lines indicate averages.
Unresolved NPLs tend to constrain economic activity of overextended borrowers and discourage resources from being allocated to productive uses. If the NPL problem in the Western Balkan countries continues to fester, it runs the risk of keeping credit sluggish and putting a drag on near- and medium-term investment and economic growth. Banks in the Western Balkans that are saddled with high NPLs have been risk-averse given the trouble with corporate balance sheets, and have been tightening credit conditions with the aim to avoid further deterioration in their loan portfolio. A recent bank lending survey confirms that high NPL levels are the biggest constraint on credit supply in the region, with the exception of Serbia, more so than the low availability of parent funding.11 Experience from earlier financial crises suggests that a forceful clean-up of bank and corporate balance sheets and reduction in NPLs are needed for a lasting economic recovery.
Elevated levels of NPLs can also pose a threat to financial stability. Banks’ earnings can suffer if eventual recovery rates on NPLs disappoint relative to provisioning. And outright losses can weaken banks’ capital bases, potentially causing insolvency or illiquidity, and financial instability—if such problems become systemic. However, capital adequacy ratios in the Western Balkan countries are high—at the upper end of international spectrum (about 17 percent on average)—as is the provisioning for bad loans. Both factors provide buffers and mitigate some of the financial stability risks. Sensitivity analysis also suggests that elevated NPLs may not pose high financial stability risks in most Western Balkan countries (European Banking Coordination Vienna Initiative 2010). Low recovery rates on existing and recognized NPLs would generally be manageable.
C. Financial Sector Inclusion and Efficiency
The financial deepening discussed above, as typically proxied by the credit-to-GDP ratio, overlooks some of the impediments that hinder financial development and thus economic growth over the medium and long term. It is therefore important to take into consideration financial inclusion and banking sector efficiency as well. Inclusive financial systems provide a broad spectrum of individuals and firms with access to financial services, allowing them to take advantage of saving and business opportunities and insure against risks (Beck, Demirguc-Kunt, and Honohan, 2008 and 2009). Efficient banking systems are solvent and profitable, with low costs and diversified income sources. They intermediate efficiently—that is, they collect savings and allocate funds for productive uses—and provide instruments to insure against risks. National savings rates and credit creation are crucially dependent on the efficiency with which the financial system provides services to enterprises or households, particularly the traditionally underserved, such as small businesses, low-income strata, rural inhabitants, older generations, and women (Karlan and Morduch 2009; World Bank 2013).
The degree of financial inclusion is typically assessed through provider-based and user-based indicators. Provider-based indicators capture the existence and geographic dispersion of financial services infrastructure, measured by, for instance, the number of bank retail outlets and geographic density of ATM machines (Beck and others 2004; Beck, Demirguc-Kunt, and Martinez Peria 2006; CGAP 2012 and 2013). User-based indicators, typically constructed from surveys of households or enterprises, quantify difficulties in accessing financial services (Demirguc-Kunt and Klapper 2012).
Compared to the New Member States, Western Balkan countries appear under-endowed in financial services infrastructure, albeit with important variation across countries. An examination of provider-based indicators suggests that the Western Balkan countries have fewer ATMs relative to what would be expected given their financial depth (credit-to-GDP ratio), particularly in Bosnia and Herzegovina. On the other hand, the Western Balkan countries seem to be relatively well covered by bank branches, especially Montenegro, allowing a reasonable proportion of the population access to financial services.
ATM Coverage and Financial Depth, 2012
Source: IMF Financial Access Survey (FAS); World Bank World Development Indicators.
Note: 2008 data for Slovakia.
Bank Branches Coverage and Financial Depth, 2012
Sources: IMF Financial Access Survey (FAS); World Bank Development Indicators.
Note: 2010 data for Lithuania.
Surveys indicate that individuals and households in Western Balkan countries have particularly limited access to saving and credit instruments compared to their counterparts in the New Member States. According to the World Bank’s Global Financial Inclusion (Global Findex) database, the degree of individual financial penetration is relatively low in the Western Balkan countries, with only 57 percent of adults, on average, maintaining a bank account, against about 70 percent, on average, in New Member States (Demirguc-Kunt and Klapper 2013). While this partly reflects the lower savings rates in Western Balkan countries given their income level, it also points to a greater dearth of attractive formal instruments for long-term saving in these countries. Household access to bank credit in the Western Balkans appears in line with the average for the New Member States, except in Albania and Kosovo.
Firms’ access to formal finance is an even more serious challenge in the Western Balkans. Access to credit for financing investments is more severely limited in Western Balkan countries than in New Member States, with firms in Croatia, FYR Macedonia, and Serbia falling below the New Member State average, and firms in Albania at about one-third of the level. High collateral requirements seem to be a major reason—they are significantly higher in Western Balkan countries than in the New Member States, except in Serbia. Banks do not appear to lend on the basis of business models and cash flow projections, but rather based on physical assets that can be pledged. Seemingly, the Western Balkans’ oft-cited problems with accounting standards, judicial systems, and governance are to blame. Regardless, the share of working capital financed by bank loans is extremely low in the Western Balkan countries.
Sources: Global Financial Inclusion (Global Findex) database; and IMF staff calculations.
Sources: World Bank Enterprise Survey; and IMF staff calculations.
Banking Sector Efficiency
Two commonly used indicators of bank inefficiency are interest rate margins and noninterest costs (see Kalluci 2010 for Albania). The Western Balkan banking model has traditionally relied on high interest margins. This could be the outcome of a low degree of income diversification, which has also left Western Balkan banks vulnerable to upward pressures on funding costs, but also high reserve requirements (Gerard and Tieman 2013) and noninterest costs associated with NPLs. Over 2003–13, banks in the Western Balkans maintained wide lending-to-deposit spreads, averaging some 2.5 percentage points higher than in the New Member States, though these ratios have been on a downward trend in the Western Balkans. Western Balkan banks also have significantly higher noninterest costs of operation than do their counterparts in the rest of Emerging Europe.
Local Currency Lending to Deposit Spreads
Sources: IMF International Financial Statistics; and staff calculations.
Non-Interest Expense to Gross Income
Sources: IMF International Financial Statistics; and staff calculations.
D. The Unfinished Institutional Reform Agenda and International Cooperation
As this report has shown, the Western Balkan financial systems need to deepen further and broaden access to financial services while preserving and enhancing financial system stability. Reforms are needed to reduce market imperfections and information asymmetries, and to allow for efficient intermediation of credit to finance investment. While Western Balkan countries have done relatively well over the last 15 years in providing the infrastructure necessary for financial development more generally and credit deepening in particular, they have lagged their New Member State counterparts in strengthening the foundations of financial stability.
Improving Institutions for Financial Deepening
At the beginning of the 2000s, the Western Balkan countries trailed their Emerging Europe counterparts in reform of financial institutions by a large margin, as seen through the EBRD’s Banking Reform Index, which is compiled largely to assess progress in banking sector liberalization in Emerging Europe. However, since then reforms in the Western Balkan countries have progressed in a sustained manner, including through the period of the global financial crisis and its subsequent fallout.
Banking Sector Reform
Source: EBRD; author’s calculations.
Reforms have particularly helped with credit market development. Dissemination of information in the credit market is an area where the Western Balkan countries have done well, even better than their New Member State counterparts. The World Bank’s Credit Depth Information Index captures the rules affecting the scope, accessibility, and quality of credit information available through credit registries. Prior to 2009, many of the Western Balkan countries—namely, FYR Macedonia, Albania, Montenegro, Serbia, and Croatia—were undertaking effective reforms, as captured by the index. By 2009, the Western Balkans as a group stood ahead of many New Member States. Further, the Western Balkan countries have shown considerable improvement in the period following the global and European crises. But while Western Balkan countries have done well in strengthening public credit registries, they have lagged their peers in developing private ones.
Quality of Credit Information
Source: World Bank, Global Financial Development Report, 2013.
Institutions for Financial Stability
The performance of Western Balkan countries in developing and strengthening institutions that maintain financial stability has not kept pace with that of the New Member States. This is apparent in the areas of supervision of financial institutions, crisis resolution and financial safety nets, and the introduction of the Basel Framework.
For a comparison of how the two groups have fared on compliance with supervisory standards, we put them on a common timeline, and assess the findings of the first assessments of compliance with such standards (the Basel Core Banking Supervision Principles or BCBS Principles).12 The divergence of the Western Balkan countries with their New Member State counterparts is particularly stark when it comes to supervision associated with cross-border banking (particularly in Albania, Bosnia and Herzegovina, Croatia, and Kosovo), consolidated accounting (Albania, Bosnia and Herzegovina, Croatia, Kosovo, and FYR Macedonia) and to a lesser extent on information requirements. Governance and risk-management practices in Western Balkan countries, as well as legal protection of supervisors (Bosnia and Herzegovina, Croatia, Kosovo, and Montenegro) and enforcement power, are also weak. There is also considerable scope in both the Western Balkan countries and other Emerging European countries to enhance governance and supervisory infrastructure in the nonbank area, particularly in insurance, which is a growing sector.
Compliance with Basel Core Principles
(First assessment/1; Average rating; 1=compliant to 5= not assessed / not applicable)
Source: International Monetary Fund
1/ EST, HUN, CZE, SVN, POL (2000), BGR, LVA, LTU, HRV (2001), SVK (2002), ROU, MKD (2003), ALB, SRB, BIH (2005), MNE (2006), KOS (2013)
Note: BCP methodology has changed over time.
Actual and Expected Adoption of Basel II
Sources: Financial Stability Institute (2014); and IMF country reports.
Note: NMS (2004): CZE, HUN, POL, SVK, SLV, EST, LTV and LTU; and NMS (2007): ROM and BGR.
Reforms to improve the credit and operational risk environment in Western Balkan countries have also not kept pace with those in the New Member States. While most of the latter adopted Basel II Framework in the mid-2000s, most of the Western Balkan countries did so only during the past four years. Though Croatia was an early adopter, it was driven by the need to align with the EU Capital Requirements Directive. Notwithstanding the delays, the adoption of Basel II frameworks should aid corporate governance, risk management, capital management, and transparency in the financial sectors in these countries.
A granular look at the practices associated with regulation and supervision reveals a picture where Western Balkans countries do well on enacting regulations, but lag on risk based regulatory and supervisory practices. More specifically, explicitly tying the intensity and frequency of supervisory activities to the riskiness of banks, linking regulatory capital requirements to commercial banks’ market and credit risk, and requiring commercial banks to disclose information about transactions with related parties are generally weaker in the Western Balkans than in the New Member States.
Bank Regulation and Supervision1/
Source: World Bank, Bank Regulation and Supervision Survey (database), 2011
1/ Data are for 2011 except for that of onsite supervision which are 2006-10.
Note: The WBS group excludes Albania and Kosovo. The NMS group excludes Czech Republic and Poland.
Not all the relative weaknesses in the reform drive in the Western Balkan countries can be attributed to domestic factors. The catalytic role played by the EU accession process in accession countries—which are now New Member States, and, more recently, Croatia—should not be discounted. This is particularly the case when it comes to improvements in financial regulations and institutional frameworks. As most Western Balkan countries have only recently moved to EU candidate status, with the exception of FYR Macedonia, the EU accession process in the coming years should help strengthen financial regulations in the period ahead.
International Coordination and Support
In the immediate aftermath of the 2008 global financial crisis, the international financial institutions launched the Vienna I Initiative to ensure a coordinated response and prevent a sudden withdrawal of Advanced EU economies’ parent bank funding from Emerging Europe. This initiative brought together home and host country authorities, parent banking groups, and multilateral organizations, including the IMF. Home authorities provided assurances that any public support for parent banks would not discriminate between their domestic and foreign operations, while parent banking groups committed to maintaining their exposure in host countries and recapitalizing their subsidiaries, if needed in the five countries with IMF-supported programs (Bosnia and Herzegovina, Hungary, Latvia, Romania, and Serbia). In retrospect, the initiative proved successful in that it avoided a generalized or local shock being transmitted to the host country financial systems. This initiative lapsed in line with the expiration of IMF and EU programs.
The Vienna II Initiative of 2012, spearheaded again by international financial institutions including the IMF, put the Western Balkan countries under greater focus, largely because there was a risk that these countries would not be covered under the forthcoming Europe-wide financial stability umbrella. The initiative acknowledged that the standpoint of home and host authorities can differ when assessing systemic risk of financial institutions, especially when a subsidiary accounts for only a minor part of the parent group’s balance sheet but could be systemic in a host country. These concerns were even more material in non-EU countries where EU-based banks are important. Vienna II has had strong participation of all of the Western Balkan countries, and its salient contribution has been the initiation of home-host country cross-border banking forums that allow host country authorities to interact with the systemically important banks, their parents, and the parents’ regulators. Vienna II is also an appropriate forum to assess possible adverse spillovers from EU regulatory changes on host countries’ banking and macroeconomic environment.
Vienna II has also called for an increasing focus on addressing the lingering problem of NPLs—a particularly acute issue in the Western Balkan countries, as seen in this report—through market-based solutions. This has entailed establishing a legal framework conducive to debt resolution (such as strengthening insolvency laws and improving law enforcement), removing tax impediments and regulatory obstacles (valuation and control-taking of collateral, asset classification and provisioning rules), and improving out-of-court restructuring frameworks. The lack of progress on the ground largely reflects weak domestic buy-in for NPL reforms (see Box 3.3).
Host Country Cross-Border Banking Forums
Source: International Monetary Fund.
In addition to playing a key role in spearheading the Vienna Initiatives, the IMF has also provided financial support to Emerging European countries, including the Western Balkan countries, to withstand shocks emanating from global and regional events. The IMF-supported programs put in place since 2008 have given priority to strengthening financial sector regulation and supervision. However, the relatively simple nature of Western Balkan countries’ financial systems, and their lesser reliance on fast-moving financial capital compared to their New Member State peers, has meant that these countries faced fewer banking system shocks in the crisis and its aftermath. Subsequently, there was more limited need for direct government action or support from international financial institutions. Not surprisingly, financial sector reforms supported by IMF programs have not made up a large share of overall reforms supported by IMF programs in most of the Western Balkan countries. On the other hand, the relatively sophisticated and complex nature of financial systems in the other Emerging European countries, notwithstanding the more advanced state of financial supervision in these countries, meant that financial sector reforms figured more prominently in their IMF programs in the post-2008 period.
Strutural Conditionality in IMF-Supported Programs
Source: IMF MONA database.
E. Cross-Border Finance and the Future of Western Balkan Banking
As discussed earlier in this report, bank deepening in the Western Balkans since the early 2000s came about on the back of the increasing presence of foreign banks. Western Balkan countries also benefited from the positive effects of foreign bank ownership, including greater professionalism and efficiency. Not surprisingly, foreign banks appear to have brought stability to the region’s banking systems, which had suffered from weak public confidence prior to 2000. At the same time, the Western Balkan states were less affected by the foreign bank-related fallout than many of the New Member States, largely because foreign ownership in Western Balkan countries (except in Montenegro) did not result in as much fast-moving capital inflows in the boom years, as it did in the New Member States. As things stand today, financial development still has a long way to go in the Western Balkan countries, and there is a significant potential for expanding financial intermediation. Nonbank financial services and capital markets need to be built, from scratch in many cases. Without bridging the gap of financial development, Western Balkan countries would find it challenging to ensure economic convergence with the rest of Europe.
A new paradigm of cross-border banking in Emerging Europe has been gaining ground recently (IMF 2013). It draws lessons from the experience of Emerging Europe, particularly the Baltics, and argues that the sharp boom-bust cycle in these countries was magnified by the reliance of banks on foreign funding, and that the postcrisis financial and economic contraction in the host countries was exacerbated by the withdrawal of foreign bank funding. It concludes that home authorities and regulators, as well as banks themselves, should seek to gradually reduce reliance on foreign funding and expand the share of domestic deposits and funding. In essence, the new paradigm espouses a gradual transition to lesser parental-based and more local-based (host country) financing. In fact, since the fall of 2008, parent banks operating in Emerging Europe have done exactly that—that is, they have been shifting away from international banking to multinational banking by rebalancing local activities toward local sources.
The characteristics and recent experience of the Western Balkan countries, however, suggest a somewhat different role for foreign funding over the medium term. As elsewhere, continued reliance on domestic funding is essential. But as the income levels of these countries are far below those of Advanced EU economies, domestic savings and deposits alone are unlikely to be sufficient to finance the credit, investment, and GDP growth that will be needed to enable the Western Balkan countries to converge to EU income levels (see Chapter 2). Thus, there is scope for an increased share of foreign funding in the Western Balkans, both because it is much lower in these countries than in New Member states, and because the overall scope for financial deepening is greater in the Western Balkans than in the New Member states, where it is already high.
While access to foreign funding over the medium and long term would be critical for financial development in the Western Balkan countries, there should be little doubt that continued deepening, improvement in access, and maintenance of stability of these financial sectors will require significant structural reforms by host governments and greater forcefulness of government policies in responding to the crisis. The Western Balkan countries need to make progress in resolving their NPLs in order to improve the creditworthiness of their private sectors and boost demand for credit. Recovery from the recent credit slowdown will require improving the quality of resolution frameworks. Reforms should also aim to reduce market imperfections and information asymmetries, as well as allow efficient intermediation of credit to finance investment. Central banks and other nonbank public institutions have to adequately regulate and supervise increasingly sophisticated private financial firms. Moreover, these firms are operating in an uncertain environment subject to external shocks. In this respect, the development and application of macro-prudential policy frameworks, still at a nascent state in Western Balkan countries, will help limit financial stability risks in the region.
Box 3.1.Nonbank Financial Deepening in the Western Balkans
Nonbank financial intermediaries and capital markets in Western Balkan countries are shallow and sometimes nonexistent, reflecting the countries’ low income levels and lagging institutional development. The lower depth of nonbank intermediaries, equity, and bond markets compared to that of banking is generally consistent with the pattern observed in other countries, whereby these more sophisticated financial markets tend to take root at higher levels of economic development. This is largely because it takes time to develop the necessary institutions to overcome higher transactions costs, reflecting agency (information and enforcement) and collective action problems (De la Torre, Feyen, and Ize 2011). Nevertheless, progress has been slow in the Western Balkan countries because reforms of legal and supervisory frameworks—critically needed to support development of capital market and nonbank financial institutions—have tended to take longer. With the exception of Kosovo and FYR Macedonia, Western Balkan countries have also lagged in the pace of privatizing the nonbank financial sector and opening it to foreign investors.1
Stock Market Capitalization and level of economic development
Sources: EBRD Structural Change Indicators; World Bank Global Financial Development Database; and IMF staff calculations.
Progress on Non-Bank Reforms
Source: EBRD Transition Indicators.
Insurance Assets Gap
Source: World Bank Finstat database.
Note: A positive gap indicates actual insurance assets are below the expected median.
Insurance markets have deepened in the past decade in those Western Balkan countries that have strengthened their institutional frameworks, although most are still further away from their potential than the New Member States. Many Western Balkan countries have passed new insurance laws since the mid-2000s and are enhancing regulatory frameworks and supervision—for example, by transferring supervision from the Ministry of Finance to an independent supervisor. This has led to consolidation in some countries where insurance depth has been high, such as Bosnia and Herzegovina and FYR Macedonia. In Serbia, where high inflation and poor claims payment have tended to hamper insurance demand, there has been progress on the structural front, as enhancements in supervision and regulatory standards since 2004 have put insurance market development on an upward trend, although it remains below its expected benchmark. Overall, insurance depth varies from 1 percent (Albania) to 10 percent (Croatia), which is the New Member State average. Further, unlike in Advanced EU countries, where life insurance is a key source of long-term savings and insurance companies are important institutional investors, in the Western Balkan countries, as in the New Member States, the insurance industry remains dominated by non-life business. However, in the latter, insurance depth appears to be more in line with their level of economic development and market size.
The development of other nonbank financial institutions that intermediate long-term savings has also been below par.
Pensions: Only the pension industries in Croatia and FYR Macedonia have seen persistent deepening over time, largely because of significant pension reforms. Serbia passed a law on voluntary pensions in 2006, but further reform on mandatory pensions (Pillar I) is needed to spur development in this area. The recent passage of the Albanian pension law holds promise for the development of insurance industry in that country.
Mutual funds: These financial institutions have been prominent in Bosnia and Herzegovina and Croatia, spurred by the use of voucher privatizations. However, there has been little progress in developing these institutions in other countries. The recent rapid expansion of nonbank investment funds in the government bond market in Albania, particularly at the longer end of maturity, is a welcome development but needs an attendant strengthening of the liquidity and regulatory environment.
Equity and corporate bond markets: These institutions are much shallower in Western Balkan countries, than in the New Member States and other parts of Europe. Stock markets do exist in all of the Western Balkan countries, except Albania and Kosovo, but the value traded is minimal. Corporate bond markets are largely nonexistent—except for small ones in Croatia and Montenegro—largely because of a lack of liquid secondary government securities markets and weak bankruptcy frameworks.2 Insufficient investor protection (such as the ease of shareholder suits) and weak regulatory frameworks hamper both corporate debt and equity markets.3
Source: Doing Business Database, 2013
1/ EUF includes BGR, FRA, DEU, ITA, LUX, and NLD.
Ease of Shareholder Suits Index, 2014
Source: Doing Business Database.
1/ EUF includes BGR, FRA, DEU, ITA, LUX, and NLD.
Box 3.2.Croatia’s Defense against the Boom1
The Western Balkan credit boom was relatively mild partly because some countries actively employed macroprudential and other prudential and capital control measures to lean against the wind. A good example is Croatia, which actively introduced such measures while pursuing a path to accession to the European Union (eventually acceding in 2013).
During 2000–08 Croatia attracted large capital inflows, spurred by sustained economic growth, looming EU accession, and, of course, easy global financial conditions. The inflows were largely in the form of foreign direct investment (FDI), but banks, which had been acquired by European parents, also increased foreign borrowing. This was aided by a legacy of high structural liquidity in the system, and a stable exchange rate. Not surprisingly, unhedged foreign currency credit shot up during the period. Bank foreign exchange-linked liabilities and assets made up as much as three-fourths of balance sheets.
The measures implemented from 2004 onward, which aimed to restrain credit growth and the foreign borrowing that was financing it by raising the cost of foreign borrowing and local lending, included the following:
A marginal reserve requirement (MRR) was introduced, and gradually increased, on banks’ new foreign borrowing.
To close a loophole, a special reserve requirement (SRR) was introduced at the rate of 55 percent on increases in banks’ liabilities arising from issued debt securities in 2006.
Credit controls required banks to purchase low-yield central bank bills for half of the increase in credit growth exceeding the allowed limit, which was increased to 75 percent in 2008.
Banks were also required to comply with a monthly 1 percent sublimit on credit growth.
The liquidity ratio of 32 percent for assets maturing in three months was extended to foreign-exchange-indexed instruments,
The general reserve requirement was reduced in steps, but remained high at 17 percent until end-2008.
The measures appear to have achieved some success. Banks’ external borrowing started falling in 2006, credit growth decelerated, and the share of foreign exchange loans declined. Following the introduction of the MRR, the loans and advances Croatian banks owed to nonresident banks declined by 10 percent. Introduction of the SRR was followed by a close to 20 percent drop in capital inflows. The measures also led to some disintermediation. To avoid the reserve requirements, the corporate sector resorted to direct borrowing from abroad. The high MRR also encouraged parent banks to fund Croatian subsidiaries by beefing up their equity (FDI inflows), rather than by debt financing. This raised bank capital buffers, which paid off during the crisis, and also enabled the banks to continue lending.
Empirical analysis in Kraft and Galac (2011) also points to some success of the measures. Vector autoregression results suggest that the MRR and SRR dampened the overall volume of inflows and contributed to exchange rate depreciation for about two quarters. In addition, the prudential (including macroprudential) measures reduced capital inflows for one quarter and led to a short-lived minor depreciation. The prudential measures also increased monetary independence marginally, for about a year.1 Based on IMF (2010) and Kraft and Galac (2011).
Box 3.3.Nonperforming Loans in the Western Balkans: Why Has Progress Been So Limited?
The pace of resolution of nonperforming loans (NPLs) in the Western Balkan countries has been slow. Domestic factors are important in explaining the lack of progress on the ground despite considerable efforts by banks. The sale of problem loan portfolios and the outsourcing of collection remain relatively rare in the Western Balkan countries. A few countries such as Latvia, Romania, Moldova, Russia, Estonia, and Poland, along with Serbia in the Western Balkans, have strived to overhaul corporate or household insolvency regimes or encouraged out-of-court restructurings. External factors have also affected progress.
A long list of obstacles in the legal, judicial, tax, and regulatory areas is holding up NPL resolution. A survey by the European Banking Coordination Vienna Initiative (2012) of international institutions and banks operating throughout the region has identified the following issues, which, of course do not necessarily apply to every country:
Collateral enforcement takes a long time and relies heavily on cumbersome judicial processes, in the form of multiple auctions with prescribed minimum bidding prices and troubled property rights regime.
Underdeveloped frameworks for going-concern restructurings mean that potentially viable firms end up in lengthy liquidation, with low loan recovery.
Out-of-court restructuring as a speedy and cost-efficient tool to settle debt is underutilized in the Western Balkan countries. Latvia and Romania have reformed their systems since the outbreak of the crisis.
Corporate insolvency frameworks in the Western Balkan countries are particularly weak, whereas Latvia, Lithuania, Estonia, and Romania have reformed their systems under the umbrella of IMF programs and technical assistance.
Most Western Balkan countries and many countries in the rest of Emerging Europe lack an insolvency framework for natural persons, disallowing financially responsible individuals from getting a “fresh start” and allowing their debt to linger on banks’ books.
Weakness and inefficiencies in the legal institutional framework delay NPL resolution. Overloaded court systems, lengthy and costly judicial proceedings, and inconsistent and unpredictable court decisions are key obstacles.
Tax deductibility of loan loss-provisions and write-downs of loans are often limited, though some Western Balkan countries, such as Albania, have recently passed laws to overcome these weaknesses.
Finally, underdeveloped markets for distressed assets limit the scope for NPL resolution, despite recent attempts by some Western Balkan countries governments (e.g., Montenegro) to facilitate this market.
Prepared by Marc Gerard, Patrick Gitton, Ricardo Llaudes, and Pamela Madrid Angers, under the supervision of Nadeem Ilahi.
See the data compiled by Laeven and Valencia (2012), which reports lower liquidity support, but higher NPLs and fiscal costs in percent of GDP for the Western Balkan countries than for the New Member States.
Both Bulgaria and Romania were lagging, particularly the latter with a median of all transition indices below 3.
Exogenous global factors in the 1990s and the 2000s complicate the comparison somewhat. The emerging market crises of the 1990s are likely to have exogenously dampened the pace of financial deepening in the New Member States, while the global credit boom in the post-2002 Great Moderation period may exaggerate the extent of financial deepening in the Western Balkans.
The 1996 banking crisis in Bulgaria was particularly severe. There were also banking crises in Lithuania and Latvia (1995) and the Czech Republic (1997). Romania had a currency crisis in 1997, and growth in many countries was also affected by the 1998 Russian crisis (IMF 2013).
The higher real interest rates in some Western Balkan countries are related to declines in inflation, which serve as a proxy for greater macroeconomic stability.
The long-run credit growth trend is estimated with a Hodrick-Prescott (HP) filter using quarterly data and a smoothing parameter of 1600.
The approach does not entirely control for endogenous factors, such as policies and institutional factors, changes in financial market structure, and regulatory environment (see Annex 3.1).
Of course, the post-2008 slowdown in credit in the Western Balkans would have been sharper had these countries written off their NPLs as rapidly as the New Member States.
The trough quarter in real credit growth in the postcrisis period (2009–10) is defined as the quarter in these two years when real credit growth was at its lowest.
See the European Investment Bank CESEE Bank Lending Survey, H1-2014. Banks in Albania, Bosnia and Herzegovina, Croatia, and Serbia participate in the survey (EIB 2014).
First assessments of New Member States were done in the early 2000s, and of Western Balkan countries around mid-2000s.