WPS7436 Policy Research Working Paper 7436 Determinants of Long-Term versus Short-Term Bank Credit in EU Countries Haelim Park Claudia Ruiz Thierry Tressel Development Research Group Finance and Private Sector Development Team October 2015 Policy Research Working Paper 7436 Abstract This paper empirically examines the determinants of credit short-term credit in most countries in the sample, and con- at different maturities across European Union countries tracted less than short-term credit after 2008. The paper during the last decade. The paper documents the length- finds that domestic deposits and foreign liabilities were ening of maturities since the early 2000s, and whether more important sources of funding in emerging market these patterns were driven by similar factors in advanced countries than in advanced countries. Moreover, trade countries and in emerging market countries. Before openness and initial banking sector depth matter more for the 2008 crisis, long-term credit expanded faster than emerging market countries than for advanced countries. This paper is a product of the Finance and Private Sector Development Team, Development Research Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at haelim.park@treasury.gov, cruizortega@worldbank.org and ttressel@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Determinants of Long-Term versus Short-Term Bank Credit in EU Countries1 Haelim Park Office of Financial Research, U.S. Treasury Claudia Ruiz World Bank Thierry Tressel World Bank JEL Classification: G21, E44, E51 Keywords: bank credit, credit maturity, EU countries, advanced economies, emerging markets 1 Author email addresses: haelim.park@treasury.gov; cruizortega@worldbank.org; ttressel@worldbank.org. The views expressed in this paper are those of the authors and do not represent those of the U.S. Treasury, World Bank, or their policies. All errors are our own. 1. Introduction During the last decades, many advanced economies (AEs) and emerging market economies (EMEs) experienced “credit boom-bust cycles” (Terrones et al., 2011). Among EMEs, Central and Eastern European (CEE) developing countries experienced a significant growth of credit after the restructuring of their banking systems and in the run-up to the global financial crisis (Cottarelli et al., 2005; Bakker and Gulde, 2010). Interestingly, in the run-up to the crisis in the EU, the maturity composition of credit to enterprises was also changing. In particular, in CEE countries, short-term credit was growing at 21 percent on average annually, while long-term credit was growing at 34 percent. While the aggregate cyclical patterns of bank credit and the determinants of banking system deepening have been well studied in the literature (Levine, 2005; Djankov et al., 2007; Detragiache and Tressel, 2008; Gozgor 2014), there is little empirical evidence studying the composition of bank credit at different maturities and its determinants in the 2000s. Moreover, even though there are studies that consider the impact of institutional factors on the maturity of bank debt (Qian and Strahan, 2007; Bae and Goyal, 2009), few papers have examined the macroeconomic determinants of the evolution of the maturity composition of bank credit over time. On the other hand, the literature studying the maturity structure of firm financing choices has emphasized the impact of the institutional environment and of financial development on the capital structure of firms (Caprio and Demirguc-Kunt, 1998), and the impact of the financial crisis (Demirguc-Kunt, Martinez Peria and Tressel, 2015). Studying the determinants of the maturity composition of credit provides a new perspective on economic growth because short-term and long-term loans do not serve the same purpose and 2 thus may have different impacts on the productive capacity of an economy. For example, it is a well-established fact that firms strive to match the maturity of their assets and liabilities (Hart and Moore, 1995; Rajan and Zingales, 1995; Booth et al, 2001). Short-term credit, which includes commercial paper, lines of credit, and overdrafts, is typically used by firms to finance working capital for a limited period of time, usually less than a year. This type of credit allows firms to finance current expenditures and undertake investments that both take relatively little time to build and generate output relatively fast. In contrast, long-term credit is suitable for fixed investments in plants, machinery and equipment and in high-return projects that require a continuing commitment of capital for years. Those investments take time to complete but may contribute more to long-term productivity growth than short-term investments (Aghion et al., 2005). Hence, a growing importance of long-term credit may indicate that the structure of the banking system is becoming more geared toward financing long-term investments and sustaining economic growth. The goal of this paper is to determine the factors that drive the growth of short, medium and long-term bank credit to non-financial corporations in EU member countries during the 2003- 2014 period, both in advanced countries and emerging countries, and during the pre-crisis and post-crisis periods. We compare the experience of the advanced EU countries to that of emerging EU countries. By comparing EU countries at different income levels, we can mitigate the concerns that the evolution of credit growth was influenced by the evolution of institutional characteristics since institutional country characteristics have generally remained stable during the period studied. The region was nevertheless characterized during this period by the existence of a new currency, the euro, and achieved very high levels of economic and financial integration (Eichengreen and Park, 2003; Laeven and Tressel, 2013). The financial integration supported convergence forces as capital flew from the relatively rich countries to the relatively less developed countries (Abiad, 3 Leigh and Mody, 2007). However, since the global financial crisis, many countries have experienced reversals of capital flows, and commentators have raised questions on the sustainability of the pre-crisis convergence of income levels. This context provides us with a unique environment to study the evolution of the maturity composition of bank credit when the institutional environment is broadly stable and in the presence of income convergence associated with capital inflows. We show that, before the crisis, credit growth on average accelerated for all groups of credit maturity (less than one year, one year to five years, above five years) in advanced EU countries and in emerging EU countries. Although there are differences across countries, credit growth was in most countries stronger in the categories of credit with longer maturities, particularly at maturities exceeding 5 years, both in advanced EU countries and in emerging markets of Central and Eastern Europe. After the crisis, credit contracted significantly in both groups of countries. A notable pattern is that long-term credit contracted moderately in advanced economies, while it contracted sharply in emerging market countries. We find that the cyclical determinants of the growth of credit at different maturities differed significantly between emerging markets and advanced economies in the EU. In the emerging markets of the EU, both domestic funding and foreign funding were significant drivers of bank credit to enterprises at all maturities, and foreign funding was more important to the growth of long-term (above five years) bank credit than that of short-term credit. Countries with less deep banking systems also experienced faster growth of credit with longer-maturities. These patterns of the data suggest that a self-sustaining process of financial deepening and increased use of long- term credit for fixed investments was at play. Credit growth at short and long maturities was also significantly positively associated with the inflation rate and with trade openness, suggesting that 4 countries with booming demand conditions and with growing links to world trade experienced rapid growth of credit. As the crisis struck, some of these forces reversed. The growth of credit in emerging economies decoupled from the growth of domestic deposits and of foreign funding. It also decoupled from the evolution of trade openness while countries with deeper banking systems experienced a faster decline of short-term bank credit. In advanced EU countries, the relationship between credit growth and its cyclical determinants was less clear. The growth of deposits and of foreign liabilities was less strongly associated with credit to enterprises at various maturities. There is some indication that the relationship between inflation and initial banking sector depth on the one hand and credit growth on the other hand seemed to remain. There is also evidence that demand conditions – captured by real GDP growth – were a driver of credit growth before the crisis. We next restrict the sample to the so called “periphery” countries of the euro area, which shared with the emerging markets the experience of credit booms and inflows of foreign capital.2 We find that in these countries, the relationship between credit growth on the one hand and credit funding conditions or initial banking sector depth on the other hand was significant, particularly at shorter maturities. The impact of the global financial crisis was that demand conditions (real GDP growth, inflation) became less significant drivers of credit growth. The paper is organized as follows: section 2 presents elements of the literature on the determinants of bank credit and its maturity composition. Section 3 describes the data we use in our empirical strategy. Sections 4 and 5 present our empirical specification and results respectively. Finally, section 6 concludes. 2 The “periphery” countries that we consider are Greece, Ireland, Portugal, Italy and Spain. 5 2. Literature Review Our paper is related to two strands of literature. First, it speaks to a growing literature on the determinants of bank credit growth to the private non-financial sector, particularly in EMEs. Guo and Stepayan (2011) find that domestic deposit growth and non-resident liability growth contribute to credit growth. They also find that stronger economic growth, higher inflation, and loose domestic and global monetary policies tend to increase domestic credit availability in the EMEs. Aisen and Franken (2010) show that credit booms prior to the crisis are an important determinant of the credit contraction observed in the aftermath of the crisis. They also find that the degree of financial integration and the type of monetary policy implemented in each country play an important role in explaining bank credit growth. Elekdag and Wu (2011) show that both domestic factors (including domestic monetary policy stances) and external factors (for instance, international interest rates which proxies global liquidity conditions) are the main drivers of credit booms across both advanced and emerging market countries. Mendoza and Terrones (2012) find that most financial crises and domestic credit booms in EMEs have occurred as a result of excessive capital inflows. Lastly, Gozgor (2014) shows that external balances and perceptions of global tail risk affect credit growth. Our paper is related to the literature studying the determinants of firms’ capital structure also provides evidence on the determinants of debt maturity from the perspective of firms’ balance sheets. This literature has argued that better institutional frameworks, more developed financial systems and more stable macroeconomic and political conditions are associated with greater use of long-term debt financing by firms (see among others, Caprio and Demirguc-Kunt, 1998; Boone et al., 2001; Demirguc-Kunt and Maksimovic, 1999; Fan, Titman and Twite; 2010). 6 Our study also speaks to the literature on the effect of laws and institutions on bank loan contracts. Qian and Strahan (2007) find that contracting costs, as measured by legal formalism, shape the terms of bank loans. In countries where legal formalism is high, domestic banks are the main lender to unrated firms. In these environments, loan maturity tends to be longer and the use of collateral tends to be more common. Later, Haselmann, Pistol, and Vig (2010) also report that improving collateral law has a large effect on increasing the supply of credit. In addition, Bae and Goyal (2009) report that banks respond to port enforceability of contracts by reducing loan amounts, shortening loan maturities, and increasing loan spreads. Sufi (2007) finds that third-party certification by rating agencies increases the availability of debt financing for firms since it reduces information asymmetries. Love, Martinez Peria, and Singh (2013) investigate the impact of introducing credit information sharing systems on firm’s use of bank loans and their maturity using firm level survey data for more than 75,000 firms in 73 countries over the period 2002–13. They find that the introduction of credit bureaus is associated with increased maturity of bank loans. Gopalan, Mukherjee, and Singh (2014) show that, in India, the establishment of new specialized courts, called debt recovery tribunals (DRTs), to process debt recovery cases, led to financial deepening and increased maturity of bank lending. Tasic and Valev (2008) show that credit maturity is longer in countries with strong institutions, low inflation, large financial markets and better information sharing among banks. Tasic and Valev (2010) explain the share of credit with maturity above one year or above five years by the political and institutional environment, the rate of inflation, the levels of economic and financial development and the establishment of credit information sharing institutions. 7 Our paper differs from the literature in the sense that the contracting right and property right environment is broadly stable in the sample. All our specifications include country fixed effects that absorb time invariant determinants of bank credit and allow to exclusively exploit variation over time within a country, and the impact of time varying factors on the maturity of bank credit. 3. Data and Summary Statistics This paper empirically examines the determinants of domestic credit levels across 26 EU countries: Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, and Sweden. Our definition of advanced economies and emerging market economies is based on the classification of the IMF. Table 1 presents the list of the main variables we use as well as the sources of information from which we obtained each variable. We construct quarterly data series for bank credit variables and macroeconomic variables for EU member countries. Data comes from three major sources: ECB Statistical Data Warehouse, the International Financial Statistics (IFS) and Global Financial Stability Report (GFSR). Additional data on institutional characteristics described in Table A2 are from various editions of the World Bank Doing Business database and from the International Country Risk Guide database. The information on bank credit of different maturities for EU member countries is from the ECB Statistical Data Warehouse, which contains: the total outstanding amount of bank credit to non-financial corporations with maturity of i) less than 1 year; ii) over 1 year and less than 5 years; and iii) over 5 years. We define these three types of loans as short-term, medium-term, and long-term loans, respectively. Credit flows are corrected for break in the series and other 8 adjustments and represent transactions.3 The information on other financial and macroeconomic variables was obtained from the International Financial Statistics (IFS) and Global Financial Stability Report (GFSR). Our analysis focuses on the following variables discussed in section 4: banking sector domestic deposits, banking sector foreign liabilities, real GDP, inflation, deposit rate, EONIA rate, trade openness, and credit-to-GDP ratio. Since information on bank credit is obtained from the ECB, we only observe countries starting from the beginning of the euro, or from the moment they joined the monetary union, turning our panel unbalanced. Table A1 lists the year in which each country is first observed in our data set. For advanced EU countries, the sample period starts in 2003:Q1, but for most emerging European countries, the sample period starts in 2004:Q1. For some of these countries (Croatia, Estonia, Latvia, and Slovenia), bank credit data availability is a greater constraint and the sample only starts in the late-2000s. The sample ends in 2014:Q1.4 For convenience in exposition, we define the pre-crisis period as the period 2003:Q1-2007:Q4 and the post-crisis period as the period 2008:Q1-2014:Q1. Figure 1 plots the growth of credit of different maturities in advanced countries and emerging countries. Credit expanded in both advanced countries and emerging countries, but it grew much faster in emerging countries than in advanced countries. An interesting feature is that all three maturity groups expanded at roughly the same rate in advanced countries, but long-term credit expended much faster than short-term credit in emerging countries. Credit stopped growing at the end of 2008, and declined in 2009-2010. While credit was on average growing faster in 3 Specifically, growth rates are corrected for reclassifications of instruments and other breaks in the series, for valuation changes, and for loan write-offs and write-downs. See ECB (2012) for more details. 4 For euro area countries, information by maturity on the total amount of loans to non-financial institutions only covers loans within the euro area, and for other euro area countries, it covers domestic loans. In the case of euro area countries, most of loans to non-financial firms are domestic loans. 9 EMEs, the contraction of credit turned out to be broadly similar in advanced countries and in emerging countries. The contraction of credit was driven by its short-term component (less than one year maturity) and medium-term component (one to five years maturity). A rebound took place in 2011 but it was soon followed by a renewed decline coinciding with the euro area crisis. In contrast, both advanced and emerging EU countries experienced a protracted and mild contraction in the long-term component of credit. The stylized facts suggest that banks first cut short-term credit lines at the onset of the 2008 crisis, followed by medium-term credit. The smoother evolution of the long-term component of credit seems more consistent with an evolution of bank loans aligned with lingering weaknesses in potential output and productivity. Table 2 provides summary statistics. Panel A provides information on credit growth variables. During the pre-crisis period, the growth of credit was on average broad based, across all three maturity groups. In advanced EU countries, credit to enterprises grew on average at 11 percent, and in emerging EU countries, it grew at 25 percent. As discussed in Bakker and Gröss (2004), some of the former socialist emerging market countries experienced important reforms to their previously state-owned banking sectors during the 1990s, and, after periods of banking crisis, restructured their banking systems and allowed entry of foreign banks. The removal of restrictions on foreign capital in the context of the EU, expectation of these countries joining the Euro Area, and economic convergence contributed in fueling capital inflows, which were associated with a rapid growth of bank credit (Abiad, Leigh and Mody, 2007). As shown in Figure 1, while short, medium and long-term credit were expanding at broadly similar annual rates of 10 to 13 percent in the EU advanced countries on average, longer-term credit in emerging countries of Central and Eastern Europe experienced significantly faster growth than shorter term credit. Specifically, short-term, medium-term and long-term credit grew annually on average at 21 percent, 24 percent 10 and 35 percent before the crisis. Since the global financial crisis, the average growth of credit to enterprises has been 0.2 percent and 0.4 percent annually in the advanced EU and in the emerging EU countries, respectively. By maturity, the average post-crisis growth of credit was, respectively for advanced EU countries and emerging EU countries, -2.4 percent and -2 percent for short-term credit, 0.2 percent and 1.6 percent for medium-term credit, and 1 and 2.1 percent for long-term credit. During the post-crisis period, the share of long-term credit in total credit remained higher than in the pre-crisis period. In advanced EU countries, it was 49 percent post-crisis and 44 percent pre-crisis on average, and in emerging EU countries, it was 42.6 percent post-crisis and 34.5 percent pre-crisis on average. Figure 2 shows the pre-crisis and post-crisis credit growth numbers country by country. The stylized fact that longer-term credit grew faster than shorter-term credit is remarkably confirmed at the country level. If we exclude Slovenia, all emerging EU countries but Bulgaria experienced a faster growth of medium-term credit than of short-term credit, and all countries but Slovakia experienced a faster growth of long-term credit than of short-term credit. Turning to the advanced EU economies, about two thirds of the countries also experienced a faster growth of medium-term credit and of long-term credit relative to short-term credit. Thus, a look at the data at the country level shows that the faster growth of longer maturity credit was in fact quite widespread both among advanced EU countries and among emerging countries of Central and Eastern Europe. Panel B of Table 2 presents summary statistics of our main explanatory variables. Before the crisis, domestic deposits were on average growing at 9 percent and 15 percent in advanced EU countries and in emerging EU countries, while foreign liabilities of banks were growing 11 respectively at 14 percent and 29 percent.5 Emerging EU countries were on average more opened to trade (and had an average ratio of trade to GDP of 123 percent) than advanced EU countries (which had an average trade to GDP ratio of 108 percent). They also experienced faster GDP growth (10 percent compared with 5 percent) and higher inflation rates (4 percent compared with 2 percent). Since the crisis, deposit growth has fallen to 2.3 percent and 3 percent respectively in the advanced EU countries and in the emerging EU countries. Foreign liabilities have contracted at 2.5 percent and 2.9 percent respectively in the advanced EU countries and in the emerging EU countries. Trade openness has increased in both groups of countries, suggesting that international trade has recovered faster than domestic activity. The overnight interbank interest rate of the euro area (EONIA), our proxy for global liquidity conditions, has fallen to about 0.4 percent on average since the crisis, compared with about 3 percent during the pre-crisis period, as a consequence of accommodative monetary policy.6 Lastly, the private sector credit-to-GDP ratio, which proxies the initial banking sector depth, has increased in both groups of countries. The higher ratio after the financial crisis may be because GDP contracted faster than credit, raising the level of the ratio. 4. Methodology The baseline specification follows recent studies on the determinants of bank credit growth.7 , = , ∙ _ , , ∙ _ , , ∆ , _ , , , , , 5 Note that, for euro area countries, the foreign liabilities of banks on national residency basis also include liabilities vis-à-vis the European Central Bank. 6 We use the EONIA rate instead of the federal funds rate since it is the overnight rate in the European market. However, the EONIA rate and the federal funds rate move closely together. 7 See for instance Guo and Stepanyan (2011). 12 In this specification, y , refers to bank credit growth (either total, short-term, medium-term or long-term) to non-financial corporations of country c during period (quarter-year) t. To examine the importance that domestic and foreign funding have on the growth of credit, we include as covariates the deposits growth rate _ , ) and the growth of non-resident liabilities ( _ , ). We interact the deposit growth rate of a country in a given period by the four-period- lagged share of deposits of the private sector total credit ( , ). Likewise, we interact the growth of non-resident liabilities by the four-period-lagged share of non-resident liabilities to private sector total credit ( , ). As explained by Guo and Stepanyan (2011), weighting the domestic and foreign funding variables by their respective shares helps control for the overall importance of these types of funding within a country. Since inflation and GDP growth are two direct macroeconomic indicators of aggregate demand and of economic activity affecting the credit growth of a country, we also include the one- period-lagged inflation ( , ) and GDP growth rate (∆ , ) in our specification. We include in our analysis the one-period lagged deposit rate to account for the impact of monetary conditions through the banking channel. In addition, given that not all EU countries use the EURO as their currency, we include the EONIA rate to adjust for liquidity conditions in Europe. Given that imports and exports of goods and services can also affect credit demand, and thus, domestic credit growth, we further control for openness to trade. In addition, economies that are more open to trade are more vulnerable to the transmission of financial shocks originating in their main trading countries (Aisen and Franken, 2010). Lastly, to control for the overall depth of the banking system, we include the lagged ratio of private sector credit to GDP, which is a standard measure of financial development. We would expect emerging market countries to have faster credit growth since they 13 might be experiencing financial deepening. To attenuate concerns of reverse causality, we use lagged values for our macroeconomic, trade, financial development and monetary policy variables. Finally, our specification includes country level fixed effects ( ) to account for other country-specific time unvarying variables that may be related to bank credit growth. For example, the country fixed effects will absorb the impact of slow-moving institutional characteristics on the deepening of the credit markets at different maturities. In general, we would expect that both domestic and foreign funding (measured by the deposit growth and non-resident liabilities rate) have a positive effect on bank credit, since more domestic deposits or more foreign liabilities may translate in more funds that banks can use to lend to firms. A positive coefficient of inflation may indicate that higher aggregate demand is associated with faster credit growth and will imply that inflation increases nominal bank credit growth. However, a coefficient less than one will indicate that while lagged inflation increases nominal credit growth, it hinders real credit growth (Boyd, Levine, and Smith, 2001). In particular, if the coefficient is less than 1, then inflation will in fact decrease real private credit. Regarding the other macroeconomic and monetary policy variables, we would expect to be positive since an economy that is growing will in general demands more credit. As with monetary policy (deposit rate), tighter policies should result in lower credit growth.8 For the EONIA rate, tighter global liquidity conditions should also result in lower credit growth. We expect a positive impact of trade openness on credit growth, as firms will require more external financing when they engage in international trade. This finding would also be consistent with the hypothesis of Rajan and Zingales (2003) that financial development tends to occur when economies are opened up to foreign competition. We expect a negative impact of the initial banking sector depth on credit growth to 8 As mentioned earlier, we use the deposit rate to proxy for the overall monetary policy stance because not all members of the EU countries use Euro. 14 the extent that banking sector depth converges to a long-term level consistent with countries’ structural socioeconomic characteristics (Čihák, Demirgüç-Kunt, Feyen, and Levine, 2013). As discussed earlier, our baseline specification includes fixed effects that are expected to absorb the impact of institutional characteristics. This hypothesis is vindicated to the extent that the institutional characteristics that are important for credit growth did not change significantly during the period we study. Table A2 reports various indicators of institutional characteristics that are relevant for bank credit, for 2003 and 2013. Indicators such as rule of law and corporate governance are very stable over time. However, credit bureaus coverage improved significantly (by more than 40 percent) in 8 countries; the indicator of contract enforcement (the number of days to enforce contracts) deteriorated significantly in 8 countries; and the recovery rate improved by more than 20 percent in five countries. To account for these institutional changes, we conduct a series of robustness that we discuss in the next section. 5. Empirical Results Tables 3-6 present the estimation results from our baseline specification for the period 2003-2014.9 Table 3 presents the results for all EU countries. The growth of deposits and of foreign liabilities adjusted for their shares in total credit to firms contributes significantly to credit growth at all maturities. These effects are economically very significant. A one standard deviation increase in the deposit variable is associated with an increase in total credit, short-term credit, medium-term credit and long-term credit growth of 1.5, 1.2, 1.1 and 1.3 percentage points respectively. A one standard deviation increase in the foreign liabilities variable is associated with an increase in total credit, short-term credit, medium-term credit and long-term credit growth of 0.8, 0.7, 1.7 and 0.5 percentage points, respectively Higher inflation, which may indicate higher aggregate demand, is 9 Tables 3-6 present the results without time fixed effects. However, our results are robust to their inclusion. 15 associated with faster credit growth, particularly for short and long maturities. In contrast, higher deposit rates are associated with slower growth of credit at all maturities, as a higher cost of retail funds translates in higher interest rates on loans to protect lending margins. In contrast, higher interbank rates are associated with faster credit growth of all maturities. This result may indicate reverse causality, as monetary policy rates would normally be higher in periods of stronger aggregate demand and credit booms. Increased openness to trade is associated with faster growth of short-term and medium-term credit, and slower growth of long-term credit, perhaps because international trade tends to require trade finance which is generally geared toward shorter-term maturities (Aisen and Franken, 2010). Last, the growth of credit is slower in countries with deeper banking systems. Tables 4 and 5 present the empirical estimates for advanced EU countries and emerging EU countries, respectively. In the EU emerging markets, the contributions of the growth of domestic deposits and foreign liabilities to credit growth are much larger than in the advanced EU countries. In particular, the impact of the growth of foreign liabilities is particularly large for long- term credit. The estimated coefficient implies that a one standard deviation increase in the explanatory variable is associated with a 21.6 percentage point increase in long-term credit.10 This finding suggests that the capital inflows in Central and Eastern Europe played a major role in financing more long-term economic activity. This beneficial effect contrasts with the recent literature that has argued that the capital flows in these countries financed mortgage borrowing by households (Vandenbussche, Vogel, and Detragiache, 2015). It is also worth noticing that the significant positive impact of trade openness on credit growth appears to be concentrated on short- 10 Studies examining more granular data of bank lending in developing countries also find that expansionary foreign monetary policy translates in higher liabilities among foreign banks, and in turn, in larger and lengthier loans to firms (see Morais, Peydro and Ruiz, 2015). 16 term credit in emerging EU countries, in line with the hypothesis that it could reflect an increase in trade finance. Last the impact of the initial banking sector depth is stronger in emerging EU countries where the banking systems are generally less deep, and these impacts are stronger for longer maturities. This finding suggests that a “catch-up” process of financial deepening was at play, and was reflected in an increased maturity of bank credit. These effects are economically large. The estimated coefficients imply that a one standard deviation increase in banking sector depth resulted in a faster growth of short-term, medium-term and long-term credit by respectively 1.8, 5.3 and 4 percentage points. This pattern of stronger financial deepening at longer maturities does not appear clearly among the advanced EU countries where both short-term and long-term credit are statistically associated with the initial depth of the banking sector. Table 6 shows the regressions on the subsample of advanced EU countries that experienced large increases of capital inflows during the pre-crisis period (these countries include Greece, Ireland, Portugal and Spain, to which we add Italy which had large pre-crisis gross external liabilities). At the creation of the euro, these countries had relatively lower GDP per capita than other countries in the monetary union, and thus were expected to experience a neoclassical convergence and capital inflows as exchange rate risk was being removed. We find that for these countries, domestic deposit growth contributed significantly to the growth of credit at all maturities. However, foreign liabilities were significant contributors to the growth of short-term credit to enterprises, in contrast to the experience of the emerging EU countries where foreign funding contributed more to long-term credit. Initial banking sector depth contributes negatively to the growth of short-term and long-term credit, suggesting there is a catch-up process of financial development, while trade openness had, surprisingly, a negative impact on credit growth in these countries. 17 Tables 7 and 8 present our empirical analysis by differentiating the pre-crisis and post- crisis periods. In these tables, we interact each of the explanatory variables with the post crisis dummy. We present the estimation results separately for advanced EU countries, and for emerging EU countries. Among emerging EU countries, the impact of deposit and foreign liabilities growth on the growth of credit turned out to be much larger before the crisis than previously estimated, and this impact dropped very significantly after the start of the crisis. Hence, while funding conditions of banks seemed to play a crucial role in driving the growth of credit at various maturities before the crisis, and especially for long-term credit, they became economically insignificant once the crisis started. Moreover, aggregate demand pressures – proxied by the CPI inflation rate – were also a strong driver of credit to firms before the crisis, especially at long maturities, but became barely relevant once the crisis started. GDP growth was insignificant in both periods, suggesting that demand factors captured by this variable may be dominated by other forces. Credit growth also became less significantly correlated with trade openness once the crisis started. In advanced EU countries, we observe the same evolution of the correlation between credit growth and the growth of foreign liabilities, while deposit growth became a stronger correlate of credit growth after the crisis, reflecting a stronger association at short maturities. After the crisis, credit growth became more systematically associated with the initial banking depth and with trade openness. But while a clear positive association between real GDP growth and credit growth was observed before the crisis, it became significantly smaller once the crisis started. Thus, our evidence suggests that, since the global financial crisis, the relationship between credit growth on the one hand – especially at long-maturities – and on the other hand, domestic and foreign funding conditions and aggregate demand pressures, has significantly weakened. Given that many countries in our sample experienced significant adjustments of capital flows, this evidence 18 suggests that these adjustments were not systematically transmitted to credit to enterprises, especially at longer maturities. The post-crisis weakening of aggregate demand – as measured by the inflation rate – was also not clearly associated with the evolution of credit to firms. A conjecture consistent with these findings is that, since the start of the global financial crisis, other types of credit – such as credit to households – may have adjusted more systematically with the evolution of funding conditions and of aggregate demand, while credit to enterprises at various maturities may have been affected in a less systematic way. Several robustness tests were conducted. First, we removed from our baseline specification Latvia and Estonia because the domestic credit cycles in these countries have a volatile structure compared to other emerging countries (Gozgor, 2014). Hence we examine if our main results do not vary when excluding such outlier countries and report our results in Table 9. The exclusion of Latvia and Estonia makes the coefficient on the deposit variable become insignificant. However, all the other variables remain with the same sign and are statistically significant, suggesting that the results are largely robust to excluding these two economies. Second, in Table 10, we include the real USD GDP (in logs) to control for convergence factors. Other than some minor differences, the main results hold fairly well. Third, we have shown in the descriptive analysis that, while the institutional environment has remained broadly stable in our sample, some countries have experienced some institutional changes during the period of analysis. We focus on the evolution of two indicators that are known to matter for bank credit and its maturity: the coverage of credit bureaus, as an indicator of transparency and information sharing; the recovery rate in bankruptcy, as an indicator of the efficiency of insolvency regimes. 11 Since our analysis includes country fixed effects, the impact of a country specific 11 Other indicators – such as the time it takes to enforce contracts, may be endogenous to the economic conditions. For example, the sharp increase in Greece or in Latvia of the number of days to enforce contracts could be an outcome, rather than a causal factor, of the sharp economic crises experienced by these countries. 19 institutional change that occurred during this period is indeed absorbed by the fixed effects. However, to assess whether specific and large institutional changes impacted credit growth on average, we re- run our baseline regressions by removing the country fixed effects and introducing a dummy variable equal to one if a country experienced a significant institutional change. While this approach is subject to potential omitted variable problems, it allows us to estimate whether the growth of credit at different maturities was systematically different in countries with significant changes in the institutional environment. Results are presented in Tables 11 and 12. While results show that the coefficient for domestic deposits change, our results remain broadly intact. Lastly, we analyze the specification where the dependent variables are the share of each maturity of credit in total credit to enterprises instead of the credit growth rates in Table 13. This approach allows us to provide evidence on the determinants of the credit shares of each maturity of credit to enterprises. Interestingly, we find that, in the emerging EU countries, the foreign capital inflows contributed in expanding the proportion of credit with long maturities, and in reducing the proportion of credit with short or medium maturities. This finding tends to support the hypothesis that capital inflows were partly driven by economic convergence and helped finance long-term productive investments. The coefficient on the initial banking sector depth variable suggests that long-term credit (respectively short-term credit) tends to expand relatively faster (respectively relatively slower) in countries with deeper banking systems initially. This tends to support the view that financial development and increasing maturities of credit are intrinsically linked. As for the role of monetary policy, a tighter monetary policy seems to impact the share of long-term credit negatively, at the benefit of short-term credit. 6. Conclusion Policy makers and economists have examined the factors that determine the growth of private sector credit, which plays an important role in economic growth. While the aggregate 20 cyclical patterns of bank credit and the determinants of banking system deepening have been well studied, there is little empirical evidence studying the composition of bank credit at different maturities and its determinants in the 2000s. Given that short-term and long-term credit play different roles in the financing of economic activity, it is important to understand how the maturity composition of credit to enterprises changes, and to identify the factors that determine the credit of different maturities. We examine the determinants of the growth of short, medium and long-term bank credit to non-financial corporations in EU member countries during the 2003-2014 period, both in advanced economies and emerging markets, and in the pre-crisis and post-crisis periods. We find that credit expanded for all groups of credit maturity (less than one year, one year to five years, above five years) in EU advanced economies and in EU emerging markets before the crisis. In addition, credit contracted for all groups of credit maturity after the crisis. While the boom-bust cycle appears to be driven by long-term credit in emerging market countries, it seems to have been driven by short-term credit in advanced economies. We find that the cyclical determinants of the growth of credit at different maturities differed significantly between emerging markets and advanced economies in the EU. In the emerging markets of the EU, foreign funding was one of the most important drives of bank credit to enterprises at all maturities, especially for the growth of long-term (above five years) bank credit. In advanced EU countries, the relationship between credit growth and its cyclical determinants was less clear. The growth of deposits and of foreign liabilities was less associated with credit to enterprises at various maturities. Our study indicates that foreign capital can be a mixed blessing for many countries. While it can promote financial deepening, in particular longer credit maturities that benefit investment 21 and help sustain economic growth in recipient countries, it can also be a destabilizing factor if capital flows are too volatile (see also World Bank Group, 2015). As seen from the last decade, the volatility of capital flows can harm their financial systems and real economies by creating short-term booms that eventually lead to crises. Therefore, macro-prudential policies should be implemented to monitor foreign-capital fueled credit booms, which can reverse course very quickly. 22 References Abiad, Abdul, Leigh, Daniel, and Ashoka Mody. 2007. “International Finance and Income Convergence: Europe is Different”, IMF Working Paper 07/64. Aisen, Ari and Michael Franken. 2010. “Bank Credit during the 2008 Financial Crisis: A Cross- Country Comparison”, IMF Working Paper 10/47. Bae, Kee‐Hong, and Vidhan K. Goyal. 2009. "Creditor rights, enforcement, and bank loans." The Journal of Finance, 64(2): 823-860. Bakker, Marie-Renée, and Alexandra Gröss. 2004. Development of non-bank financial institutions and capital markets in European union accession countries. World Bank Publications. Booth, Laurence, Varouj Aivazian, Asli Demirguc-Kunt, and Vojislav Maksimovic. 2001. "Capital structures in developing countries." Journal of finance, 87-130. Boyd, John H., Ross Levine, and Bruce D. Smith. 2001. “The Impact of Inflation on Financial Sector Performance,” Journal of Monetary Economics, Vol. 47, pp. 221-48. Caprio, Gerard, and Asli Demirguc-Kunt. 1998. "The role of long-term finance: theory and evidence." The World Bank Research Observer, vol. 13(2): 171-189. Čihák, Martin, Asli Demirgüč-Kunt, Erik Feyen, and Ross Levine. 2013. “Financial development in 205 economies, 1960 to 2010,” National Bureau of Economic Research. Working Paper 18946. Cottarelli, Carlo, Dell'Ariccia, Giovanni and Ivanna Vladkova-Hollar. 2005. "Early birds, late risers, and sleeping beauties: Bank credit growth to the private sector in Central and Eastern Europe and in the Balkans," Journal of Banking & Finance, Elsevier, 29(1): 83-104. Claessens Stijn, M. Ayhan Kose and Marco E. Terrones, 2011. "Financial Cycles: What? How? When?" NBER International Seminar on Macroeconomics, University of Chicago Press, 7(1): 303 - 344. Demirgüč-Kunt, Asli, Martinez-Peria, Maria Soledad, and Thierry Tressel. 2015. “The Impact of the Global Financial Crisis on Firms’ Capital Structures: The Role of Financial Markets and Institutions.” World Bank, Washington, DC. Djankov, Simeon, McLiesh, Caralee, and Shleifer, Andrei. 2007. "Private credit in 129 countries," Journal of Financial Economics, 84(2): 299-329. Eichengreen, Barry, and Yung Chul Park. 2003. "Why has there been less financial integration in Asia than in Europe?." Institute of European Studies. Elekdag, Selim Ali, and Yiqun Wu. "Rapid credit growth: boon or boom-bust?”. 2011. IMF Working Papers: 1-42. 23 European Central Bank. 2012. Manual on MFI Balance Sheet Statistics, European Central Bank, Eurosystem, April 2012. Fan, Joseph PH, Sheridan Titman, and Garry Twite. 2012. "An international comparison of capital structure and debt maturity choices." Journal of Financial and quantitative Analysis, vol. 47(1): 23-56. Gopalan, Radhakrishnan, Abhiroop Mukherjee, and Manpreet Singh. 2014. “Do Debt Contract Enforcement Costs Affect Financing and Asset Structure?” Olin School of Business, Washington Business School in Saint Louis. Gozgor, Giray. 2014. "Determinants of domestic credit levels in emerging markets: The role of external factors." Emerging Markets Review, 18: 1-18. Hart, Oliver, and John Moore. 1995. “Debt and seniority: An analysis of the role of hard claims in constraining management.” The American Economic Review, 85(3): 567-585 Haselmann, Rainer, Katharina Pistor, and Vikrant Vig. 2010. "How law affects lending." Review of Financial Studies 23.2: 549-580. Laeven, Luc and Thierry Tressel. 2013. “The European Union: Financial Integration and Fragmentation in the European Union”, European Union Financial Sector Assessment Program Technical Note, IMF Country Report No. 13/71, March 2013. Levine, Ross. 2005. "Finance and Growth: Theory and Evidence," Handbook of Economic Growth, in: Philippe Aghion & Steven Durlauf (ed.), Handbook of Economic Growth, edition 1, volume 1, chapter 12, pages 865-934 Elsevier. Love, Inessa, María Soledad Martínez Pería, and Sandeep Singh. 2013. "Collateral Registries for Movable Assets: Does Their Introduction Spur Firms’ Access to Bank Financing?." Journal of Financial Services Research: 1-37. Mendoza, Enrique G., and Marco E. Terrones. 2012. An anatomy of credit booms and their demise. No. w18379. National Bureau of Economic Research. Morais, Bernardo, José-Luis Peydró, and Claudia Ruiz. 2015. "The International Bank Lending Channel of Monetary Policy Rates and Quantitative Easing." World Bank. Qian, Jun, and Philip E. Strahan. 2007, "How laws and institutions shape financial contracts: The case of bank loans." The Journal of Finance, 62(6): 2803-2834. Stepanyan, Vahram, and Kai Guo. 2011. Determinants of bank credit in emerging market economies. No. 11-51. International Monetary Fund. 24 Sufi, Amir. 2007. "Information asymmetry and financing arrangements: Evidence from syndicated loans." The Journal of Finance, 62(2): 629-668. Rajan, Raghuram, and Luigi Zingales. 2003. “The Great Reversals: The politics of Financial Development in the Twentieth Century,” Journal of Financial Economics, 69: 5–50. Tasic, Nikola, and Neven Valev. 2008. “The Maturity Structure of Bank Credit: Determinants and Effects on Economic Growth.” Andrew Young School of Policy Studies Research Paper Series 08-12, Georgia State University, Atlanta. Tasic, Nikola, and Neven Valev. 2010. “The Provision of Long-Term Financing in the Transition Economies.” Journal of Comparative Economics, 38(2): 160–72. Tressel, Thierry and Enrica Detragiache. 2008. “Do Financial Reforms Lead to Financial Development? Evidence from a New Dataset”, International Monetary Fund, Working Paper No. 08/265. Vandenbussche, Jérôme, Ursula Vogel, and Enrica Detragiache. 2015. "Macroprudential Policies and Housing Prices: A New Database and Empirical Evidence for Central, Eastern, and Southeastern Europe." Journal of Money, Credit and Banking, 47(1): 343-377. World Bank Group. 2015. Global Financial Development Report 2015 – Long-Term Finance, World Bank, Washington DC. 25 Table 1. Variables and Sources Variable Description Database/Source Private credit Credit to enterprises IFTS, ECB Short-term credit Credit to enterprises with maturity up ECB to 1 year Medium-term credit Credit to enterprises with maturity 1-5 ECB years Long-term credit Credit to enterprises with maturity over ECB 5 years Foreign liabilities Liabilities to non-residents/Foreign IFTS Liabilities, based on domestic residence for euro area countries Domestic deposits Transferable/demand deposits; IFTS Other/Time and savings deposits, based on domestic residence for euro area countries Inflation CPI inflation IFTS Real GDP GDP Index, constant local currency IFTS EONIA rate EONIA rate ECB Deposit rate Interest rate on deposits ECB Trade openness Exports plus imports to GDP ratio IFTS Private credit to GDP Credit to the private sector by IFTS depository credit institutions to GDP 26 Table 2. Summary Statistics. Advanced EU countries Emerging EU countries Panel A: Dependent Variables # Obs Mean SD # Obs Mean SD All periods Growth of total credit 600 0.054 0.095 335 0.096 0.157 Growth of short-term credit 600 0.034 0.124 335 0.068 0.167 Growth of medium-term credit 600 0.064 0.152 335 0.101 0.169 Growth of long-term credit 600 0.061 0.100 335 0.145 0.252 Share of short-term credit 600 0.332 0.147 335 0.325 0.089 Share of medium-term credit 600 0.201 0.061 335 0.280 0.057 Share of long-term credit 600 0.468 0.159 335 0.395 0.079 Pre-crisis period Growth of total credit 270 0.110 0.095 121 0.248 0.156 Growth of short-term credit 270 0.096 0.127 121 0.210 0.168 Growth of medium-term credit 270 0.128 0.150 121 0.240 0.165 Growth of long-term credit 270 0.118 0.101 121 0.353 0.304 Share of short-term credit 270 0.363 0.139 121 0.374 0.071 Share of medium-term credit 270 0.193 0.063 121 0.281 0.058 Share of long-term credit 270 0.443 0.160 121 0.345 0.067 Post-crisis period Growth of total credit 315 0.002 0.057 206 0.004 0.057 Growth of short-term credit 315 -0.024 0.087 206 -0.020 0.093 Growth of medium-term credit 315 0.002 0.120 206 0.016 0.104 Growth of long-term credit 315 0.010 0.067 206 0.021 0.087 Share of short-term credit 315 0.304 0.148 206 0.295 0.086 Share of medium-term credit 315 0.206 0.058 206 0.279 0.057 Share of long-term credit 315 0.489 0.156 206 0.426 0.071 27 Panel B: Independent Variables # Obs Mean SD # Obs Mean SD All periods Deposit growth 588 0.055 0.075 318 0.076 0.109 Non-resident liability growth 588 0.050 0.166 318 0.094 0.288 inflation 588 0.019 0.014 318 0.035 0.025 GDP Growth 588 0.026 0.040 318 0.054 0.068 Deposit rate 561 0.891 0.741 300 0.961 0.944 EONIA rate 592 1.626 1.401 322 1.467 1.437 Trade openness 592 1.123 0.722 322 1.293 0.338 Credit-to-GDP ratio 592 5.369 1.980 322 2.253 0.750 Pre-crisis period Deposit growth 260 0.093 0.058 113 0.152 0.076 Non-resident liability growth 260 0.140 0.098 113 0.294 0.208 Inflation 260 0.024 0.011 113 0.048 0.028 GDP Growth 260 0.052 0.022 113 0.108 0.047 Deposit rate 234 1.433 0.795 95 1.382 0.939 EONIA rate 264 2.948 0.861 117 3.111 0.852 Trade openness 264 1.082 0.693 117 1.233 0.292 Credit-to-GDP ratio 264 4.630 1.419 117 1.743 0.517 Post-crisis period Deposit growth 315 0.023 0.074 197 0.031 0.101 Non-resident liability growth 315 -0.025 0.173 197 -0.029 0.261 Inflation 315 0.015 0.015 197 0.027 0.019 GDP Growth 315 0.006 0.040 197 0.023 0.058 Deposit rate 315 0.462 0.277 197 0.726 0.814 EONIA rate 315 0.454 0.357 197 0.422 0.349 Trade openness 315 1.160 0.745 197 1.335 0.357 Credit-to-GDP ratio 315 5.984 2.171 197 2.557 0.706 Source: Authors’ Calculations. 28 Table 3. Determinants of Bank Credit by Loan Maturity, Full Sample. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits 0.200*** 0.159*** 0.144*** 0.167*** [0.027] [0.035] [0.047] [0.037] Non-res liab growth * shr non-res liab 0.046*** 0.041*** 0.100*** 0.028* [0.011] [0.014] [0.019] [0.015] Lagged Inflation 0.909*** 1.810*** 0.242 0.665*** [0.156] [0.202] [0.275] [0.214] Lagged GDP Growth 0.005 -0.020* 0.003 0.021 [0.009] [0.012] [0.017] [0.013] Lagged deposit rate -0.061*** -0.044*** -0.049*** -0.088*** [0.006] [0.008] [0.010] [0.008] Lagged EONIA rate 0.061*** 0.052*** 0.072*** 0.074*** [0.003] [0.004] [0.005] [0.004] Lagged trade openness 0.010 0.062** 0.065** -0.062** [0.019] [0.024] [0.033] [0.026] Lagged credit-GDP -0.013*** -0.027*** 0.001 -0.013** [0.004] [0.005] [0.007] [0.005] Country fixed effect Yes Yes Yes Yes Time fixed effect No No No No Observations 836 836 836 836 R-squared 0.656 0.561 0.443 0.527 Number of countries 26 26 26 26 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 29 Table 4. Determinants of Bank Credit by Loan Maturity, Advanced EU countries. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits 0.119*** 0.067 0.076 0.030 [0.029] [0.042] [0.059] [0.029] Non-res liab growth * shr non-res liab 0.048*** 0.043*** 0.102*** 0.039*** [0.010] [0.015] [0.021] [0.010] Lagged Inflation 0.968*** 2.490*** 1.047** 0.028 [0.215] [0.304] [0.429] [0.212] Lagged GDP Growth 0.026 0.016 0.090 0.028 [0.052] [0.073] [0.103] [0.051] Lagged deposit rate -0.039*** -0.008 -0.037** -0.057*** [0.009] [0.012] [0.017] [0.008] Lagged EONIA rate 0.046*** 0.029*** 0.065*** 0.054*** [0.004] [0.006] [0.008] [0.004] Lagged trade openness 0.010 0.046 0.023 0.006 [0.025] [0.036] [0.051] [0.025] Lagged credit-GDP -0.014*** -0.033*** 0.013 -0.021*** [0.004] [0.006] [0.008] [0.004] Country fixed effect Yes Yes Yes Yes Time fixed effect No No No No Observations 547 547 547 547 R-squared 0.607 0.506 0.416 0.560 Number of countries 15 15 15 15 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 30 Table 5. Determinants of Bank Credit by Loan Maturity, Emerging EU countries. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits 0.294*** 0.256*** 0.295*** 0.320*** [0.049] [0.063] [0.085] [0.088] Non-res liab growth * shr non-res liab 0.524*** 0.385*** 0.351*** 0.751*** [0.053] [0.067] [0.090] [0.094] Lagged Inflation 0.743*** 0.938*** -0.016 0.910** [0.220] [0.283] [0.379] [0.392] Lagged GDP Growth 0.018* -0.012 0.005 0.041** [0.009] [0.012] [0.016] [0.016] Lagged deposit rate -0.038*** -0.033*** -0.038*** -0.057*** [0.008] [0.011] [0.015] [0.015] Lagged EONIA rate 0.042*** 0.052*** 0.050*** 0.045*** [0.005] [0.006] [0.008] [0.009] Lagged trade openness 0.064** 0.126*** 0.081* -0.024 [0.026] [0.033] [0.045] [0.046] Lagged credit-GDP -0.049*** -0.025** -0.071*** -0.054*** [0.010] [0.013] [0.017] [0.018] Country fixed effect Yes Yes Yes Yes Time fixed effect No No No No Observations 289 289 289 289 R-squared 0.814 0.715 0.575 0.668 Number of countries 11 11 11 11 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 31 Table 6. Determinants of Bank Credit by Loan Maturity, Periphery of Advanced EU countries. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits 0.557*** 0.542*** 0.432*** 0.558*** [0.100] [0.157] [0.159] [0.095] Non-res liab growth * shr non-res liab 0.169*** 0.244*** 0.084 0.095** [0.041] [0.064] [0.065] [0.039] Lagged Inflation 0.763** 1.858*** -0.371 0.153 [0.335] [0.528] [0.533] [0.320] Lagged GDP Growth 0.083 -0.412 0.649** 0.068 [0.171] [0.270] [0.273] [0.163] Lagged deposit rate 0.008 0.027 -0.053 -0.006 [0.031] [0.050] [0.050] [0.030] Lagged EONIA rate 0.017 -0.003 0.052*** 0.031*** [0.011] [0.017] [0.017] [0.010] Lagged trade openness -0.202*** -0.116 -0.469*** -0.103 [0.070] [0.110] [0.111] [0.066] Lagged credit-GDP -0.019*** -0.047*** -0.007 -0.028*** [0.007] [0.011] [0.011] [0.007] Country fixed effect Yes Yes Yes Yes Time fixed effect No No No No Observations 160 160 160 160 R-squared 0.833 0.685 0.753 0.842 Number of countries 4 4 4 4 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 32 Table 7. Determinants of Bank Credit by Loan Maturity before and after the Crisis, Advanced EU Countries. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits -0.191*** -0.118 -0.405*** -0.090 [0.060] [0.087] [0.118] [0.060] Non-res liab growth*shr non-res liab 0.124*** 0.065** 0.232*** 0.108*** [0.023] [0.033] [0.045] [0.023] Lagged Inflation 2.471*** 3.509*** 5.155*** 1.473*** [0.441] [0.641] [0.876] [0.442] Lagged GDP Growth 1.125*** 1.841*** 2.103*** 0.604*** [0.192] [0.279] [0.381] [0.193] Lagged deposit rate -0.029*** -0.015 -0.014 -0.040*** [0.009] [0.014] [0.019] [0.009] Lagged EONIA rate 0.012* 0.027*** -0.001 0.025*** [0.007] [0.010] [0.013] [0.007] Lagged trade openness 0.023 -0.006 0.045 -0.023 [0.033] [0.048] [0.066] [0.033] Lagged credit-GDP 0.010 -0.020** 0.049*** -0.003 [0.006] [0.009] [0.012] [0.006] Postcrisis*Dep gth*shr dep 0.367*** 0.298*** 0.542*** 0.109 [0.071] [0.103] [0.141] [0.071] Postcrisis*Non-res liab gth*shr non-res liab -0.098*** 0.000 -0.143*** -0.085*** [0.027] [0.040] [0.055] [0.028] Postcrisis*Lagged Inflation -1.948*** -0.840 -5.611*** -2.080*** [0.501] [0.728] [0.995] [0.503] Postcrisis*Lagged GDP Growth -1.100*** -1.890*** -2.021*** -0.549*** [0.196] [0.285] [0.390] [0.197] Postcrisis*Lagged deposit rate 0.026* 0.027 0.091*** 0.041*** [0.014] [0.020] [0.027] [0.014] Postcrisis*Lagged EONIA rate 0.024*** -0.022* 0.050*** 0.014 [0.009] [0.013] [0.018] [0.009] Postcrisis*Lagged trade openness 0.011 0.015 0.046** 0.045*** [0.012] [0.017] [0.023] [0.012] Postcrisis*Lagged credit-GDP -0.011*** 0.010* -0.019*** -0.014*** [0.004] [0.005] [0.007] [0.004] Country fixed effect Yes Yes Yes Yes Time fixed effect No No No No Observations 547 547 547 547 R-squared 0.667 0.560 0.511 0.615 Number of countries 15 15 15 15 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 33 Table 8. Determinants of Bank Credit by Loan Maturity before and after the Crisis, Emerging EU Countries. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits 0.586*** 0.549*** 0.334** 0.727*** [0.082] [0.122] [0.167] [0.151] Non-res liab growth*shr non-res liab 0.941*** 0.487*** 0.907*** 1.620*** [0.090] [0.134] [0.182] [0.165] Lagged Inflation 1.275*** 0.754* 0.992* 2.450*** [0.267] [0.396] [0.539] [0.488] Lagged GDP Growth 0.000 -0.006 -0.009 0.013 [0.010] [0.014] [0.019] [0.018] Lagged deposit rate -0.041*** -0.048*** -0.043** -0.043*** [0.009] [0.013] [0.018] [0.016] Lagged EONIA rate 0.015* 0.022* 0.051*** 0.013 [0.008] [0.012] [0.017] [0.015] Lagged trade openness 0.150*** 0.127** 0.204*** 0.165** [0.036] [0.054] [0.073] [0.066] Lagged credit-GDP -0.059** 0.064* -0.186*** -0.169*** [0.023] [0.034] [0.047] [0.042] Postcrisis*Depgth*shr dep -0.481*** -0.429*** -0.158 -0.722*** [0.095] [0.141] [0.192] [0.174] Postcrisis*Non-res liab gth*shr non-res liab -0.774*** -0.407** -0.844*** -1.387*** [0.112] [0.167] [0.227] [0.205] Postcrisis*Lagged Inflation -1.096*** -0.359 -1.398* -2.342*** [0.352] [0.523] [0.712] [0.644] Postcrisis*Lagged GDP Growth 0.019 -0.017 0.002 0.014 [0.015] [0.023] [0.031] [0.028] Postcrisis*Lagged deposit rate 0.024*** 0.040*** 0.020 0.006 [0.008] [0.011] [0.015] [0.014] Postcrisis*Lagged EONIA rate 0.027** 0.032* -0.012 0.040* [0.012] [0.017] [0.024] [0.021] Postcrisis*Lagged trade openness -0.069*** 0.009 -0.123** -0.147*** [0.026] [0.038] [0.052] [0.047] Postcrisis*Lagged credit-GDP 0.039* -0.065** 0.133*** 0.154*** [0.021] [0.031] [0.042] [0.038] Country fixed effect Yes Yes Yes Yes Time fixed effect No No No No Observations 289 289 289 289 R-squared 0.883 0.760 0.629 0.780 Number of countries 11 11 11 11 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 34 Table 9. Determinants of Bank Credit by Loan Maturity, Full Sample without Latvia and Estonia. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits 0.023 0.018 -0.010 0.012 [0.014] [0.017] [0.024] [0.018] Non-res liab growth * shr non-res liab 0.027*** 0.037*** 0.068*** 0.016 [0.009] [0.011] [0.015] [0.012] Lagged Inflation 0.898*** 1.857*** 0.374 0.557** [0.169] [0.207] [0.287] [0.223] Lagged GDP Growth -0.033*** -0.033*** -0.032* -0.014 [0.010] [0.012] [0.017] [0.013] Lagged deposit rate -0.053*** -0.038*** -0.041*** -0.082*** [0.006] [0.008] [0.011] [0.008] Lagged EONIA rate 0.063*** 0.053*** 0.074*** 0.076*** [0.003] [0.004] [0.005] [0.004] Lagged trade openness 0.022 0.089*** 0.056 -0.059** [0.020] [0.025] [0.034] [0.027] Lagged credit-GDP -0.024*** -0.034*** -0.010 -0.021*** [0.004] [0.005] [0.007] [0.005] Country fixed effect Yes Yes Yes Yes Time fixed effect No No No No Observations 823 823 823 823 R-squared 0.628 0.566 0.439 0.519 Number of countries 24 24 24 24 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 35 Table 10. Determinants of Bank Credit by Loan Maturity, Full Sample, Adding Lagged Real GDP. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits 0.012 0.034** -0.021 0.001 [0.013] [0.017] [0.023] [0.018] Non-res liab growth * shr non-res liab 0.030*** 0.031*** 0.071*** 0.019* [0.009] [0.011] [0.015] [0.011] Lagged Inflation 0.835*** 1.799*** 0.151 0.649*** [0.167] [0.210] [0.285] [0.219] Lagged GDP Growth -0.022** -0.052*** -0.020 -0.002 [0.009] [0.011] [0.016] [0.012] Lagged deposit rate -0.054*** -0.038*** -0.041*** -0.082*** [0.006] [0.008] [0.011] [0.008] Lagged EONIA rate 0.063*** 0.054*** 0.074*** 0.076*** [0.003] [0.004] [0.005] [0.004] Lagged trade openness 0.021 0.085*** 0.055 -0.047* [0.020] [0.025] [0.034] [0.026] Lagged credit-GDP -0.024*** -0.034*** -0.012* -0.022*** [0.004] [0.005] [0.007] [0.005] Lagged real GDP -0.022 0.003 0.074 -0.094** [0.034] [0.043] [0.058] [0.045] Country fixed effect Yes Yes Yes Yes Time fixed effect No No No No Observations 851 851 851 851 R-squared 0.622 0.554 0.429 0.516 Number of countries 26 26 26 26 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 36 Table 11. Determinants of Bank Credit by Loan Maturity, Full Sample, Dummy for Credit Bureaus. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits 0.022 0.039** -0.013 0.014 [0.015] [0.019] [0.023] [0.020] Non-res liab growth * shr non-res liab 0.024** 0.020* 0.060*** 0.015 [0.009] [0.012] [0.014] [0.013] Lagged Inflation 1.265*** 1.789*** 0.786*** 1.757*** [0.162] [0.205] [0.253] [0.222] Lagged GDP Growth -0.018* -0.045*** -0.017 0.000 [0.010] [0.013] [0.016] [0.014] Lagged deposit rate -0.031*** -0.029*** -0.026*** -0.044*** [0.004] [0.005] [0.007] [0.006] Lagged EONIA rate 0.058*** 0.056*** 0.066*** 0.061*** [0.003] [0.003] [0.004] [0.004] Lagged trade openness 0.009* -0.002 0.013* -0.019*** [0.005] [0.006] [0.007] [0.006] Lagged credit-GDP -0.003** -0.003* -0.003 -0.004** [0.001] [0.002] [0.002] [0.002] Dummy Credit bureaus 0.006 -0.008 0.010 0.019** [0.007] [0.009] [0.011] [0.010] Country fixed effect No No No No Time fixed effect No No No No Observations 851 851 851 851 R-squared 0.558 0.469 0.402 0.439 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 37 Table 12. Determinants of Bank Credit by Loan Maturity, Full Sample, Dummy for Recovery Rates in Bankruptcy. (1) (2) (3) (4) Total Short-term Medium-term Long-term Deposit growth*shr deposits 0.020 0.039** -0.016 0.011 [0.015] [0.019] [0.023] [0.020] Non-res liab growth * shr non-res liab 0.025*** 0.021* 0.064*** 0.015 [0.009] [0.012] [0.015] [0.013] Lagged Inflation 1.272*** 1.716*** 0.757*** 1.840*** [0.160] [0.203] [0.250] [0.221] Lagged GDP Growth -0.018* -0.044*** -0.015 -0.000 [0.010] [0.013] [0.016] [0.014] Lagged deposit rate -0.031*** -0.031*** -0.026*** -0.041*** [0.004] [0.005] [0.007] [0.006] Lagged EONIA rate 0.058*** 0.057*** 0.066*** 0.059*** [0.003] [0.003] [0.004] [0.003] Lagged trade openness 0.008* -0.001 0.011 -0.020*** [0.004] [0.006] [0.007] [0.006] Lagged credit-GDP -0.004*** -0.003* -0.005** -0.006*** [0.001] [0.002] [0.002] [0.002] Dummy Recovery Rates -0.007 -0.012 -0.025** -0.002 [0.008] [0.010] [0.012] [0.011] Country fixed effect No No No No Time fixed effect No No No No Observations 851 851 851 851 R-squared 0.558 0.470 0.404 0.437 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 38 Table 13. Determinants of Bank Credit Shares by Maturity of Loans. (1) (2) (3) (4) (5) (6) (7) (8) (9) All EU Countries Advanced EU Countries Emerging EU Countries Share of Share of Share of Share of Share of Share of Share of Share of medium- Share of short-term medium- long-term short-term medium- long-term short-term term long-term credit term credit credit credit term credit credit credit credit credit Deposit growth*shr deposits 0.080*** -0.059*** -0.021 0.114*** -0.075*** -0.038** 0.010 -0.016 0.006 [0.012] [0.010] [0.014] [0.016] [0.011] [0.017] [0.016] [0.020] [0.027] Non-res liab growth * shr non- res liab -0.007 -0.005 0.012** -0.009* 0.004 0.006 -0.091*** -0.044** 0.135*** [0.005] [0.004] [0.006] [0.006] [0.004] [0.006] [0.017] [0.021] [0.029] Lagged Inflation -0.012 -0.108* 0.120 -0.118 -0.005 0.122 -0.001 -0.102 0.103 [0.068] [0.057] [0.082] [0.115] [0.079] [0.122] [0.071] [0.089] [0.121] Lagged GDP Growth -0.003 0.005 -0.002 -0.010 -0.007 0.017 -0.004 0.002 0.002 [0.004] [0.003] [0.005] [0.028] [0.019] [0.029] [0.003] [0.004] [0.005] Lagged deposit rate 0.003 -0.014*** 0.011*** -0.005 -0.029*** 0.035*** 0.007** -0.001 -0.006 [0.003] [0.002] [0.003] [0.005] [0.003] [0.005] [0.003] [0.003] [0.005] Lagged EONIA rate 0.009*** 0.008*** -0.016*** 0.013*** 0.014*** -0.027*** 0.014*** 0.002 -0.016*** [0.001] [0.001] [0.002] [0.002] [0.002] [0.002] [0.002] [0.002] [0.003] Lagged trade openness -0.024*** 0.019*** 0.005 -0.036*** 0.028*** 0.008 -0.013 0.012 0.001 [0.008] [0.007] [0.010] [0.014] [0.009] [0.014] [0.008] [0.010] [0.014] Lagged credit-GDP -0.017*** 0.005*** 0.012*** -0.015*** 0.010*** 0.006** -0.015*** -0.019*** 0.034*** [0.002] [0.001] [0.002] [0.002] [0.002] [0.002] [0.003] [0.004] [0.005] Country fixed effect Yes Yes Yes Yes Yes Yes Yes Yes Yes Time fixed effect No No No No No No No No No Observations 836 836 836 547 547 547 289 289 289 R-squared 0.458 0.161 0.293 0.479 0.349 0.350 0.520 0.115 0.359 Number of countries 26 26 26 15 15 15 11 11 11 Note: Dependent variables are growth rates of total, short-term, medium-term, and long-term credit in EU countries. Standard errors are clustered at the country level. ***, **, * denote significance at 1, 5 and 10 percent levels respectively. 39 Figure 1. Evolution of Credit in EU Countries. Total Credit -.15 -.1-.05 0 .05 .1 .15 .2 .25 .3 .35 .4 .45 .5 Percentage 01jul2003 01jan2006 01jul2008 01jan2011 01jul2013 Date Advanced Countries Emerging Countries Short-term Credit -.15-.1-.05 0 .05 .1 .15 .2 .25 .3 .35 .4 .45 .5 Percentage 01jul2003 01jan2006 01jul2008 01jan2011 01jul2013 Date Advanced Countries Emerging Countries               40 Medium-term Credit -.15 -.1-.05 0 .05 .1 .15 .2 .25 .3 .35 .4 .45 .5 Percentage 01jul2003 01jan2006 01jul2008 01jan2011 01jul2013 Date Advanced Countries Emerging Countries Long-term Credit -.15 -.1-.05 0 .05 .1 .15 .2 .25 .3 .35 .4 .45 .5 Percentage 01jul2003 01jan2006 01jul2008 01jan2011 01jul2013 Date Advanced Countries Emerging Countries Source: ECB Statistical Data Warehouse. 41 Figure 2. Average Growth of Loans to Non-Financial Corporations in EU Member Countries. Advanced EU Countries Emerging EU Countries Total Credit Total Credit 2003Q1‐2008Q3 2009Q1‐2014Q1 50% 2003Q1‐2008Q3 2009Q1‐2014Q1 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% ‐10% ‐10% Under 1 Year Under 1 Year 50% 2003Q1‐2008Q3 2009Q1‐2014Q1 50% 2003Q1‐2008Q3 2009Q1‐2014Q1 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% ‐10% ‐10% Over 1 Year and Under 5 Years Over 1 Year and Under 5 Years 50% 2003Q1‐2008Q3 2009Q1‐2014Q1 50% 2003Q1‐2008Q3 2009Q1‐2014Q1 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% ‐10% ‐10% 42   Over 5 Years Over 5 Years   2003Q1‐2008Q3 2009Q1‐2014Q1 2003Q1‐2008Q3 2009Q1‐2014Q1 50% 100%   40% 80% 30% 60%   20% 40%   10% 20% 0% 0%   ‐10% ‐20%         Source: ECB Statistical Data Warehouse. 43 Appendix Table A1. Sample of EU Countries. Non-ECA Countries ECA Countries country year country year Austria 2003 Bulgaria 2007 Belgium 2003 Croatia 2013 Cyprus 2003 Czech Republic 2004 Denmark* 2003 Estonia 2004 Finland 2003 Hungary 2004 France 2003 Latvia 2004 Germany 2003 Lithuania 2004 Greece 2003 Poland 2004 Ireland 2003 Romania 2007 Italy 2003 Slovakia 2004 Luxembourg 2003 Slovenia 2004 Malta 2003 Netherlands 2003 Portugal 2003 Spain 2003 Sweden 2003 United Kingdom* 2003 Note: * indicates that data from credit to non-financial corporations by maturity missing for this country. 44 Table A2. Institutional Characteristics. Country Rule of law Contract enforcement (days) Credit bureaus coverage Anti-director rights Recovery rates 2003 2013 2003 2013 2005 2013 2003 2013 2003 2013 Austria 6 6 397 397 45 53 5 5 73 82 Belgium 5 5 505 505 0 0 6 6 86 89 Bulgaria 4 2.5 564 564 0 0 2 2 33.5 33.2 Croatia 5 4.5 561 572 0 100 6 6 28.5 30.5 Cyprus 5 5 . 735 0 6.8 . 4 . 70 Czech Republic 5 5 653 511 38 77 5 5 17.8 65.6 Denmark 6 6 380 410 7.7 7.8 5 5 67.2 87.5 Estonia 4 4 425 425 12 34 3 3 39 39.3 Finland 6 6 235 375 14 19 4 4 89 90.2 France 4.5 5 390 395 0 0 3 3 47.5 77.2 Germany 5 5 394 394 88 100 5 5 81.3 83.4 Greece 3 4.5 819 1580 17 82 3 4 45.9 34.3 Hungary 4 4 335 395 4 74 4 4 35.7 40.2 Ireland 6 6 515 650 100 100 6 6 88 87.7 Italy 3 4 1390 1185 60 100 4 4 63.6 62.8 Latvia 5 5 279 469 0 0 4 4 33.9 48.2 Lithuania 4 4 210 300 12 98 4 4 49.8 43.6 Luxembourg 6 6 . 321 0 0 . 4 . 44 Malta 5 5 . 505 0 0 . 6 39.6 Netherlands 6 6 514 514 69 79 4 4 88.4 88.9 Poland 4 4.5 980 685 38 84 2 2 32.1 57 Portugal 5 5 577 547 10 23 5 5 74.7 72.2 Romania 4 3.5 537 512 1 47 5 5 17.5 30.7 Slovakia 4 4 565 545 18 68 4 4 38.6 54.4 Slovenia 4.5 4.5 1350 1270 0 100 8 9 44 50.1 Spain 4.5 5 515 510 6 15 6 6 74.1 71.3 Sweden 6 6 508 321 100 100 4 4 74.9 76.1 United Kingdom 6 5 404 437 76 100 7 7 85.3 88.6 Source: World Bank Development Indicators.  45