On the Existence of a Credit Channel of Monetary Policy in Germany

Guender, Alfred; Moersch, Mathias Article On the Existence of a Credit Channel of Monetary Policy in Germany Kredit und Kapital Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Guender, Alfred; Moersch, Mathias (1997) : On the Existence of a Credit Channel of Monetary Policy in Germany, Kredit und Kapital, ISSN 0023-4591, Duncker & Humblot, Berlin, Vol. 30, Iss. 2, pp. 173-185, https://doi.org/10.3790/ccm.30.2.173 This Version is available at: https://hdl.handle.net/10419/293348 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ On the Existence of a Credit Channel of Monetary Policy in Germany* By Alfred Guender and Mathias Moersch, Christ church and Frankfurt/Main I. Introduction Since the mid 1980s, a new wave of research has explored the special role accorded to bank loans in the transmission of monetary policy from the financial to the real sector of the economy.1 This credit view argues that a bank's choice to allocate assets away from loans as a reaction to tight monetary policy has potentially large effects on real output. While the theoretical arguments are well established, debate still continues about the empirical strength of the credit channel. In recent surveys both Kashyap and Stein (1994) and Cecchetti (1995) conclude, albeit tentatively, that the credit channel is responsible for significant movements in real output. Neumann (1995) and Eichenbaum (1994) on the other hand, doubt its empirical relevance. While almost all studies focus on the U.S. experience, we argue in this paper that Germany, due to its institutional environment, is a particularly interesting case for the study of the credit channel. Relations between banks and firms, which lie at the core of the transmission mechanism of the credit view are particularly strong in Germany. These strong ties between German banks and firms have at least two important implications. First, German firms are more dependent on bank financing than their U.S. counterparts. Second, due to close and long-lasting relationships between banks and firms, the so-called Hausbank relationship, banks may be more reluctant to cut credit supply to firms. Relying on a number of tests based on aggregate data, we find very little evidence for the existence of a separate credit channel of monetary * We would like to thank seminar participants at the University of Birmingham and the Southern Economic Association meetings in New Orleans and an anonymous referee for valuable suggestions. The usual disclaimer applies. 1 The lending view of monetary policy is not a new idea. For early expositions see for example Rosa (1951) and Wojnilower (1980). 12 Kredit und Kapital 2/1997 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.30.2.173 | Generated on 2023-01-16 13:09:05 174 Alfred Guender and Mathias Moersch policy in Germany. Our results are, however, consistent with the traditional money-view. The remainder of the paper is organized as follows. In Section II we review the credit channel of monetary policy and in Section III the institutional background in Germany. In Section IV we conduct the empirical investigation. The main results are summarized and put in perspective in Section V. II. The Credit Channel of Monetary Policy The credit view of monetary policy complements the traditional money view by focusing on two channels of transmission of monetary policy that are not addressed in the money view. They are a balance sheet effect and a portfolio effect, which are discussed in detail below. Both effects tend to increase the potency of monetary policy and ultimately work through the availability of bank credit. The credit view differs from the money view by incorporating a more detailed picture of the process of financial intermediation. In the money view there is neither a special role for banks as providers of assessments about the probability of repayment of an investment as stressed in the balance sheet effect, nor is there any modeling of the asset side of banks, as detailed in the portfolio effect. Instead, monetary policy has effects on the economy only by changing the money supply and interest rates. Changes in interest rates affect the profitability of the marginal investment project and thus output. The balance sheet effect emphasizes the role that monetary policy has on the net-worth of borrowers. Due to informational asymmetries in evaluating an investment project, a firm's balance sheet is an important factor in determining its ability to obtain external funding. Changes in interest rates lead to a change in a firm's net worth by changing the value of debt and future sales. More specifically, contractionary policy will lead to a lower net worth, which makes firms less creditworthy. Hence some firms lose access to credit, because monetary policy tightens. Due to these worsening credit market conditions, the output decline at the aggregate level is much more pronounced than it would be in the absence of a balance sheet effect. This effect, where small changes in interest rates have large effects on output via credit market conditions has also been called the financial accelerator by Bernanke, Gertler and Gilchrist (1996). OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.30.2.173 | Generated on 2023-01-16 13:09:05 On the Existence of a Credit Channel of Monetary Policy in Germany 175 The portfolio effect stresses the asset allocation of banks and the fact that some firms - especially small ones - depend on a particular bank asset, namely loans, for funding. In addition to the size of banks' balance sheets, asset allocation also matters. In particular loans are only one possible asset in banks' portfolios. Asset allocation away from loans will have a negative effect on output, whenever firms exist whose only source of finance are bank loans. In sum, two institutional aspects must be in place for the credit channel to be operational. First, there must be firms that have no alternative to bank loans as a source of financing. Second, banks must reduce their supply of loans in reaction to a negative policy shock. If they were able to offset the policy shock by either increasing other liabilities or by reducing bond holdings disproportionately the credit channel would not be operational. In the next section we turn to the institutional environment. We argue that the way financial intermediation is organized in Germany has, at least according to the traditional view, strong implications for these two conditions. III. The Provision of External Finance in Germany The German case is usually cited as a prime example of a bank-based system of financial intermediation. This contrasts with the market-based systems of the United States and the United Kingdom. In bank-based systems, as the name implies, banks play a much larger role both in the channelling of wealth from savers to investors and in the governance of corporations. Market-based systems leave both functions mostly to financial markets. Empirically, one large difference between the two systems lies in the financial structure of non-financial enterprises. First, debt financing is more important in bank-based systems than in market-based systems. Second, among all debt, bank loans play a much larger role than commercial paper and corporate bonds in bank-based systems. Finally, market-based systems have better developed and more liquid money and capital markets than their bank-based counterparts.2 This strong dependence on bank financing implies that the credit channel of monetary policy ought to be particularly strong in a bank-based 2 For surveys on the differences in national financial markets see Bockelmann (1996) and Steinherr and Huveneers (1994). 12 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.30.2.173 | Generated on 2023-01-16 13:09:05 176 Alfred Guender and Mathias Moersch economy like Germany. Not only is debt more important than in marketbased systems, but most debt is bank financed and alternative forms of financing are not as readily available. However, there are also a number of institutional features, subsumed under the notion of corporate governance, that may weaken the credit channel in Germany A notable aspect of the German system is the Hausbank relationship. This term characterizes the long-term relationship between a firm and a bank. One bank will serve as the main provider of financial services to a firm. This role is facilitated by the fact that in Germany universal banking allows for the provision of a large spectrum of financial services by one institution. The reliance on a Hausbank , through the reduction of competition, is seen as a way to promote commitment and a longer term focus of the bank. The close ties between banks and firms in Germany are also manifest in German banks' representation on the supervisory board of firms. This representation provides banks with several advantages. First, they obtain confidential information, thus lowering information asymmetries and second, they can protect the suppliers of debt finance in times of financial distress. Both the Hausbank relationship and supervisory board representation give German banks an incentive to provide firms with long-term financing. Put differently, they are probably more reluctant than their American counterparts to cut off the supply of loans to firms. Consequently, the particularities of the German banking systems may reduce or even eliminate the existence of a credit channel of monetary policy. With respect to the two necessary conditions for the existence of the credit view of monetary policy mentioned above, the following implications arise. On the one hand, German firms rely on external finance, and in particular bank finance, to a larger extent than their counterparts in market-based systems. The absence of non-bank sources of funding would imply a particularly strong role for the credit view. On the other hand, the institutional features of the bank-based system make it likely that German banks provide more continuity with respect to long-term financing than banks in market-based systems. The strong relationship between banks and firms makes it less likely that financing will not be rolled over. As a result, the balance sheet effect may not be operational at all or its effect greatly diminished in Germany. Since the Hausbank relationship and board supervision reduce the problems associated with asymmetric information, net worth may lose its importance as a determinant of creditworthiness.3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.30.2.173 | Generated on 2023-01-16 13:09:05 On the Existence of a Credit Channel of Monetary Policy in Germany 177 IV. Empirical Method and Results The empirical investigation relies on impulse response functions. First, monetary policy is identified and the transmission of policy from money market interest rates to the real economy is traced out. Second, the reaction to a monetary policy shock of the most important portfolio shares of banks, namely loans and bonds on the asset side and deposits on the liability side are analyzed. Neither test finds any evidence for the existence of a credit channel. 1. The Transmission of Monetary Policy In this section we capture the interaction between monetary policy, moneyand credit-aggregates, output and prices, and trace out the transmission mechanism of monetary policy. To do so, we employ a five variable vector autoregressive model (VAR) that contains the following variables: IIP (the log of the index of industrial production), Price (the log of the consumer price index), Spread (the difference between a bond rate, measured as the average maturity of long bonds and the money market rate), Ml (the log of real Ml) and Crsht (the log of real short term credit to private enterprise and individuals). We estimate the system which contains 12 lags of all five variables and a constant, using monthly data from 69:1 until 94:12. It is identified via the following assumptions. First, the Cholesky decomposition is used to obtain orthogonalization. Second, it is assumed that innovations in the spread capture monetary policy surprises.4 The spread, rather than the short term interest rate alone is used to identify monetary policy to avoid the price 3 The importance of these institutional differences is a matter of some debate. Edwards and Fischer (1994) have recently argued that the differences between German and Anglo-Saxon banking systems are greatly exaggerated. Comparing the banking system in Germany and the United Kingdom, they find no support for the claim that external finance is more readily available in Germany than in the United Kingdom. They point out that supervisory board representation is limited to large firms and large banks. The majority of firms do not have boards and thus a large portion of bank lending is not connected to supervisory board representation. Similarly they doubt the importance of the Hausbank. If this view is right and the institutional differences are of little empirical importance when it comes to the availability of credit, the German case should not be very different from the Anglo-Saxon model. 4 This assumption is based on the work of Bernanke and Blinder (1992). It does, of course, assume that the supply function for reserves is perfectly elastic. See Gordon and Leeper (1994) for a criticism of this approach and an attempt to estimate interest elasticities of reserve supply and demand. OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.30.2.173 | Generated on 2023-01-16 13:09:05 178 Alfred Guender and Mathias Moersch puzzle.5 Third, the ordering of the variables is IIP, Price, Spread, Ml, Crsht. This implies that monetary policy does not affect output and prices contemporaneously, but does affect money and credit within the period.6 Figure 1 shows the set of 25 impulse responses and their two-standarddeviation error bands over a horizon of 24 months. The reactions to a positive shock in the spread, a monetary policy easing, are depicted in the third row. We find that the easing leads to an increase in real output. This effect becomes statistically significant after about one year. The monetary aggregate rises. This effect is significant on impact and remains so for the next 12 months. The volume of credit also rises after the policy shock, but the rise is never statistically different from zero. As shown in the last panel of the first row, credit reacts positively and significantly during the entire 24 months to a positive output shock. When combined with the finding that the volume of credit does not react to a change in monetary policy, the observed positive comovement between real output and credit constitutes strong evidence against the credit view. Credit appears to be driven more by developments in the real economy than by monetary policy actions. A consistent explanation for these impulse responses is that credit demand goes up as the economy expands. However, the impulse responses are not consistent with the view that tight policy leads banks to cut the supply of credit. 5 We use the spread, rather than a short term interest rate to identify monetary policy, because the price puzzle, which has been noted in previous research, for example Sims (1992), is also observed in our work. In the wake of a monetary tightening, measured by an increase in the short term interest rate, the price level is found to be increasing, a rather paradoxical result by conventional wisdom. Sims (1992) argues that the prize puzzle arises from the omission from the VAR of a variable measuring inflationary expectations. He shows that the inclusion of such a variable, commodity prices in his case, reduces the price puzzle. The spread measures monetary policy relative to the long bond, and therefore also incorporates inflationary expectations. This specification indeed makes the price puzzle disappear, while the results with respect to the other variables are similar to the case where the short term rate alone is used to identify policy. We therefore argue that our scheme of identifying monetary policy by innovations to the interest rate spread is sensible. 6 The nature of VARs and their identification assumptions are discussed, for example, in Sims (1980). OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.30.2.173 | Generated on 2023-01-16 13:09:05 On the Existence of a Credit Channel of Monetary Policy in Germany 179 2. Bank Balance Sheets The main focus in this section is on analyzing the changes in the composition of the structure of assets and liabilities of the aggregate banking sector in response to a monetary policy tightening. Since the price puzzle is not an issue here, we identify monetary policy as innovations to the money market interest rate. Following the work of Bernanke and Blinder (1992), in addition to short term credit to non-banks, we also analyze bonds on the asset side and deposits from non-banks on the liability side.7 Figure 2 traces the responses of the balance sheet items to a monetary tightening. The impulse responses are derived from a VAR system with the following ordering: Short term interest rate (Rate), real short term credit (Crsht), real bonds (Rbond) and real deposits (Rdeposit). We find that bond holdings fall significantly on impact and remain below the base value for the entire 24 months. This effect is statistically significant for the first twelve months. Deposits also fall over the entire period and the fall becomes significant after about ten months. Short credit rises throughout, but the effect is never significantly different from zero. In Figure 3 we enter all three balance sheet items as fractions of the balance sheet total. Scrsht refers to the share of short term credit as a percentage of the balance sheet total, similarly for bonds and deposits we use Sbond and Sdeposit. Here we see that the portion of loans actually increases significantly for about 5 months and then falls back towards the old value. Bonds are below the average for the entire 24 months, but the decline is not statistically significant. Deposit shares rise initially and then fall, but the effect is also never statistically significant. Bernanke and Blinder (1992) and Kashyap and Stein (1994) have pointed out that the fact that loans do not immediately fall as a reaction to tight policy is in itself no evidence against the credit view. If loans are quasi-contractual arrangements that are hard to change in the short run, the necessary initial portfolio adjustment is instead undertaken by shedding the more liquid bonds. However, we find no decrease in loans, either in real terms or as shares over a two-year period, which seems sufficiently long to undertake portfolio adjustments. Loans as a percentage of the overall portfolio even rise significantly on impact and then ? On the asset side other important positions are loans to other banks and on the liability side these are deposits of other banks and securities issued. OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.30.2.173 | Generated on 2023-01-16 13:09:05 180 Alfred Guender and Mathias Moersch >poqs OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.30.2.173 | Generated on 2023-01-16 13:09:05