Abstract
To reduce the negative effects of a bank-lending crunch on economic activity, adequate credit provision should be ensured. Further bank recapitalization, restructuring and consolidation of the banking sector, and regulatory reform decisions will reduce uncertainty. A long-lasting configuration of the euro-area’s fiscal architecture can be achieved by tightly coordinated reforms of national fiscal frameworks. Substantial benefits will emanate from deepening further structural reforms. Financial sector reform in the EU is proceeding at a rapid pace, and poses challenges and opportunities for the EU.
I. Euro Area: A Bank Credit Crunch?1
A. Introduction
1. Recent bank losses could induce significant loan supply restrictions which would weigh heavily on the recovery of the euro-area economy. Since the eruption of the global financial crisis, European banks have suffered heavy losses. These developments could result in a “bank credit crunch” as argued e.g., by Holmstrom and Tirole (1997). Shocks that affect banks’ capital positions force especially the weaker banks to adjust loan supply to meet targeted leverage and capital ratios, as well as regulatory requirements.2 Underlying this bank capital channel is the assumption that the market for bank equity is imperfect and that capital may be very costly to raise for banks during economic downswings.
2. The divergence in external debt financing patterns of corporations suggests that bank lending constraints are binding. To assess whether a bank lending crunch is mounting in the euro area, section II analyzes the external financing patterns of non-financial corporations as a means to disentangle loan supply from demand effects. Since early 2009, corporate credit costs have declined and net issuance of corporate debt has soared. At the same time, bank lending growth to non-financial companies has fallen sharply. Survey results and disaggregated bank loan data indicate that bank loan supply constraints are driving this divergence that rather than differences in firms’ risk profiles. Section III uses a simple, standard VAR model that features GDP, the interest rate, bank loans, the growth gap between bonds and bank loans and controls for external developments, to identify loan supply shocks. The model indicates that such recent shocks could subtract up to two percentage point from GDP growth over the next two years. Given that the analysis suggests that loan supply conditions could constitute a significant drag on the euro area recovery, we conclude that fixing the banking sector is crucial.
B. Is there a Bank Credit Crunch in the Euro Area?
3. There is little doubt that negative feedback loops between the real and financial sectors occurred in the euro area through a broad credit channel during the downturn. But disentangling loan supply effects, caused by large bank losses, from demand effects is a difficult task, especially if the analysis is focused on aggregate output and bank loans. Aggregate data show that the credit-to-GDP ratio is non-stationary and does not exhibit any obvious deterministic trend but do not reveal if bank lending supply constraints have had a negative impact on output. GDP led credit through most of the euro-area recession that lasted through 2008Q2–2009Q2 (Figure 1). Therefore, it is difficult to argue that aggregate credit or bank lending constraints caused the recession, though there have likely been pockets of credit rationing in some market segments.
Credit in Relation to Output
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Data on bank loans to non-financial corporations and securities other than shares issued by non-financial corporations are from the ECB and nominal GDP data are from Eurostat.Credit in Relation to Output
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Data on bank loans to non-financial corporations and securities other than shares issued by non-financial corporations are from the ECB and nominal GDP data are from Eurostat.Credit in Relation to Output
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Data on bank loans to non-financial corporations and securities other than shares issued by non-financial corporations are from the ECB and nominal GDP data are from Eurostat.4. Demand factors are seen to have played a key role this far. Bank lending growth to nonfinancial corporations has fallen steeply and recently turned negative. Both, the ECB in its various publications and policy statements and the EC in its 2010 Spring forecast explain weak lending growth primarily with subdued demand for corporate loans. They point to the fact that bank lending to households seems to have stabilized and the low level of bank lending to nonfinancial corporations could simply reflect weak demand. Indeed, along with value added, corporate investment activity slumped and working capital and inventories fell to very low levels. Moreover, bank lending surveys have pointed to the weak economic outlook as the main reason for tightening credit standards, while capital costs and banks’ access to market financing seem to have mattered considerably less in this regard (Figure 2).
ECB Bank Lending Surveys (2003Q1–2010Q1)
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: The net percentage is calculated as the difference between reported “increases” and “decreases.”ECB Bank Lending Surveys (2003Q1–2010Q1)
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: The net percentage is calculated as the difference between reported “increases” and “decreases.”ECB Bank Lending Surveys (2003Q1–2010Q1)
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: The net percentage is calculated as the difference between reported “increases” and “decreases.”5. The recent surge in corporate debt issuance stands in stark contrast to bank lending developments (Figure 3). The share of bank loans in total credit to non-financial corporations has fluctuated over the past decade (Figure 4). For the short period for which data are available, however, it is difficult to formally distinguish between regular variation over the business cycle and structural events, also because the beginning of the period was marked by a significant structural shift: the introduction of the euro and the 2002/03 banking crisis in Germany. Nonetheless, the recent divergence in bond issuance and bank lending is unprecedented.
ECB Bank Lending Survey-The Effects of Capital Costs on Lending
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
ECB Bank Lending Survey-The Effects of Capital Costs on Lending
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
ECB Bank Lending Survey-The Effects of Capital Costs on Lending
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Bond Issuance versus Bank Loans of Non-Financial Corporations
(Annual growth rates)
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Data on bank loans to non-financial corporations and securities other than shares issued by non-financial corporations are from the ECB. BBB (corporate euro bonds)-German Bund yields are from DataStream.Bond Issuance versus Bank Loans of Non-Financial Corporations
(Annual growth rates)
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Data on bank loans to non-financial corporations and securities other than shares issued by non-financial corporations are from the ECB. BBB (corporate euro bonds)-German Bund yields are from DataStream.Bond Issuance versus Bank Loans of Non-Financial Corporations
(Annual growth rates)
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Data on bank loans to non-financial corporations and securities other than shares issued by non-financial corporations are from the ECB. BBB (corporate euro bonds)-German Bund yields are from DataStream.6. The substitution of debt securities for bank loans by non-financial corporations could suggest binding bank loan supply constraints. Kashyap, Stein and Wilcox (1993) argued that changes in loan demand should affect all types of credit and other external financing sources in broadly the same way. A shift in the composition of corporate debt from bank debt to bonds would therefore imply a supply constraint. Oliner and Rudebusch (1996), however, suggested that such a shift at the aggregate level could also be caused by a shift in financing within the corporate sector from riskier SMEs to safer, large corporations. Credit may be redirected away from SMEs to large corporations because informational asymmetries magnify the premium charged for external debt for riskier SMEs during a downturn, a mechanism consistent with the broad view of the credit channel. According to this view, banks should also shift lending from SMEs to large companies as they lend to both. At the individual firm level where the level of risk is given there should not be any substitution according to this view and the bank and capital market debt mix should not change.
7. The evolution of credit spreads, disaggregated bank lending data and survey results corroborate the view that there are binding bank loan supply constraints in the euro area. Against the Oliner and Rudebusch thesis, credit spreads reflecting the premium for external debt fell while the shift in the composition of corporate debt from bank debt to bonds occurred (Figure 4). Also, disaggregated bank loan data published by the ECB indicate that the debt mix of large corporations, with similar risk profiles, shifted significantly in favor of capital market debt. The data show a relatively stable ratio of large to total bank loans during the crisis and over the past years (Figure 5). Assuming that large loans are primarily extended to large companies, bank loans to large corporations must have fallen with a similar rate as total loans and large corporations have increasingly relied on capital market debt for external financing. Moreover, new surveys on the access to finance of SMEs in 2009 conducted by the ECB showed that both large corporations and SMEs had experienced a significant deterioration in the availability of bank loans with the net percentage (28 percent) of large firms reporting a deterioration similar to that of SMEs’ (32 percent, see Figure 6). These findings suggest that the euro area is facing a bank lending crunch.
Ratio of Large over Total New Bank Loans to Non-Financial Corporations
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Large loans are defined as corporate loans exceeding 1 million euro. Data are from the ECB.Ratio of Large over Total New Bank Loans to Non-Financial Corporations
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Large loans are defined as corporate loans exceeding 1 million euro. Data are from the ECB.Ratio of Large over Total New Bank Loans to Non-Financial Corporations
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Large loans are defined as corporate loans exceeding 1 million euro. Data are from the ECB.
ECB SME Survey-Changes in the Availability of External Financing
(Net percentages of responders; over July-December 2009)
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: The net percentage is calculated as the difference between reported “increases” and “decreases.”ECB SME Survey-Changes in the Availability of External Financing
(Net percentages of responders; over July-December 2009)
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: The net percentage is calculated as the difference between reported “increases” and “decreases.”ECB SME Survey-Changes in the Availability of External Financing
(Net percentages of responders; over July-December 2009)
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: The net percentage is calculated as the difference between reported “increases” and “decreases.”C. What Could be the impact of a Bank Lending Crunch on the Recovery?
8. Constrained bank loan supply could weigh heavily on the recovery of the euro-area economy. Bank lending remains the predominant external financing source of most corporations. Large corporations could escape a bank lending crunch by issuing their own bonds in capital markets, but other bank-dependent corporations, mostly SMEs would face binding credit constraints.3 The effect on growth should be significant as SMEs account for 60 percent of value added and 70 percent of employment in the euro.
9. A simple, standard VAR model suggests that the loans supply shock experienced in 2009 could lower GDP by about 2 percentage points during 2010–11.4 The model features GDP, the interest rate, bank loans, the growth gap between bonds and bank loans and controls for external developments. Other studies, e.g., Cara and Morgan (2006), use bank credit standards published by the Senior Loan Officer Opinion Survey to identify U.S. loan supply shocks. But if surveyed bank lending conditions, as in the euro area case, are primarily determined by overall economic conditions and the outlook, they may not help much in identifying loan supply shocks. Instead, our model uses the difference between the growth of non-financial corporate bonds and bank loans to identify loan supply shocks. A shortcoming of this identification procedure is that a rise in this indicator may well reflect benign developments such as capital market innovation or financial deepening. In particular, euro adoption may well have provided a bigger boost to corporate bond markets than corporate bank lending. Adding a dummy for this period, 1999–2001, to control for euro adoption does not have much of an impact on the results, however. Moreover, to control for external and global financial credit conditions, US GDP growth and the Moody’s Baa-Aaa spread are added as exogenous variables. Figure 7 plots the response of GDP to these variables. As expected, GDP growth is relatively persistent and falls significantly in response to interest rate shocks; a 100 basis point increase lowers GDP by about 1.2 percent after two years. The estimated response of GDP growth to a shock in bank loan growth is very small and only significant after several years. A shock to the bond/bank loan growth difference triggers a negative and significant effect on GDP after about two years; the recent 20 percentage point increase in the growth difference would thus reduce GDP by about 2 percent over the next two years
GDP’s Impulse Responses (VAR with 3 lags) to one Standard Deviation Shocks in the Endogenous Variables
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Staff calculations.GDP’s Impulse Responses (VAR with 3 lags) to one Standard Deviation Shocks in the Endogenous Variables
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Staff calculations.GDP’s Impulse Responses (VAR with 3 lags) to one Standard Deviation Shocks in the Endogenous Variables
Citation: IMF Staff Country Reports 2010, 222; 10.5089/9781455205806.002.A001
Note: Staff calculations.10. While results of this model should be taken cautiously, they are mostly confirmed by cross-country and analytical studies. For most countries, banking crisis are not a common feature of the business cycle and are likely to entail other important features than variations in bank capital, for example a strong regulatory response. Cross-country studies can better analyze how recessions caused by financial crisis differ from others. The 2009 October WEO found that the negative repercussions of financial crisis induced recessions are deeper and longer-lasting and feature lower and roughly equal growth contributions from all major production inputs: capital/investment, labor and TFP. Abiad, Dell’Ariccia, and Li (2010) analyze “creditless” recoveries and find that they do occur but, growth is on average about two percentage points lower than otherwise. Moreover, many of these “creditless” recoveries were driven by exports. In an analytical study, Meh and Moran (2009) find that exogenous bank capital shocks create sizeable declines in output and investment using a DSGE model in which bank capital mitigates an agency problem between banks and their creditors.5
D. Conclusion
11. As many European banks have suffered heavy losses in the wake of the global financial crisis, the bank capital channel is likely to constrain loan supply and therefore weigh on the euro-area recovery. Recent financial sector developments suggest that a bank lending crunch is mounting in the euro area and, according to a simple VAR analysis, could slow growth by about 1 percentage point in both 2010/11.
12. One way to reduce the negative effects of a bank lending crunch on economic activity could be to ensure adequate credit provision to sound SMEs through other channels than commercial banks. In the context of the EU’s response to the crisis, many countries put measures in place to facilitate SME’s access to financing but the uptake has been rather limited. The cost of assessing SMEs’ business situation and risks to ensure that loans are made to sound companies could be relatively high. Especially, where bank-SME relationships already exist, banks could do such assessments much more cost efficient and supervisors should ensure that their actions do not unduly limit the availability of bank credit to sound SMEs.
13. More important though are further bank recapitalization, restructuring and consolidation of the banking sector and that regulatory reform decisions are made soon to reduce uncertainty. Especially weaker banks need to raise additional capital, clean up their balance sheets and put forward a convincing business model embedded in a sound governance structure, or face restructuring, divesture or takeover. The desirability of long-term public ownership stakes is questionable as they may rather hamper than accelerate restructuring and consolidation of the European banking sector, especially across national borders.
References
Abiad, Abdul, Giovanni Dell’Ariccia, and Bin Li, 2010, “Creditless Recoveries,” forthcoming IMF Working Paper (Washington: International Monetary Fund).
Bernanke, Ben, and Cara Lown, 1991, “The Credit Crunch,” Brookings Papers of Econmic Activity, No. 2, pp. 205 –240.
Cara, Lown and Donald Morgan, 2006, “The Credit Cycle and the Business Cycle: New Findings Using the Loan Officer Opinion Survey,” Journal of Money, Credit and Banking 38(6), pp. 1575–1597.
Fagan, Gabriel, Jérôme Henry and Ricardo Mestre, 2001, “An Area-wide Model (AWM) for the euro area,” ECB working paper No. 42 (Frankfurt: ECB).
ECB, 2009a, “Monetary Policy and Loan Supply in the Euro Area,” Article in October Monthly Bulletin (Frankfurt: ECB).
ECB, 2009b, “Survey on the access to finance of small and medium-sized enterprises in the euro area,” September (Frankfurt: ECB).
Holmstrom, Bengt and Jean Tirole, 1997, “Financial intermediation, loanable funds, and the real sector,” Quartely Journal of Economics 112, pp. 663–691.
Kashyap, Anil, Stein, Jeremy and David W. Wilcox, 1993, “Monetary Policy and Credit Conditions: Evidence from the Composition of External Finance,” American Economic Review Vol. 83 (March) pp. 78–98.
Kashyap, Anil, Stein, Jeremy and David W. Wilcox, 1996, “Monetary Policy and Credit Conditions: Evidence from the Composition of External Finance: Reply,” American Economic Review Vol. 86 (March) pp.310–314.
Oliner, Stephen and Glenn Rudebusch, 1996, “Monetary Policy and Credit Conditions: Evidence from the Composition of External Finance: Comment,” American Economic Review Vol. 86 (March), pp.300–309.
Notes
Prepared by Thomas Harjes.
See, ECB (2009a) for a comprehensive overview of the main monetary policy transmission channels involving banks.
Very large and highly-rated euro-area corporations usually have access to international capital markets but others may be constrained by relatively underdeveloped national capital markets, such as Germany’s.
The quarterly data are real GDP (q-o-q growth), the overnight interest rate, real bank loans (q-o-q growth), and the difference in annual growth rates of the stock of bank loans to non-financial corporations and debt securities issued by non-financial corporations over 1990:1–2009:4. Real GDP and GDP deflator data for 1995:1–2009:4 are from Eurostat, for 1990:1–1994:1, these data are from the Area-wide Model (AWM) database for the euro area, see Fagan, Henry and Mestre (2001). US GDP data and Moody’s Baa-Aaa spread are from by the Federal Reserve Bank of Saint Louis and all other data are from the ECB.
Especially the recent euro-area bank losses due to U.S. mortgage market exposure reflect an exogenous shock not directly related to the euro-area business cycle that led to a decline in bank capital.
