Introduction

So far, empirical evidence of a credit channel in Germany is inconclusive, irrespective of the methodology or the type of data used. For example, while Favero, Giavazzi and Flabbi (2001), Guender and Moersch (1997), Stöss (1996) and Tsatsaronis (1995), do not find a credit channel, de Bondt (1999, 2000), Kakes and Sturm (2001), Küppers (2001), Hülsewig, Winker and Worms (2001) and Worms (1998), find evidence in support of it. This ambiguity in the results reflects the fundamental problem of identifying monetary policy-induced shifts in loan supply. In order to tackle this problem, most of the recent empirical literature on this issue has turned to micro data.

Along these lines, this study uses individual bank balance sheets in order to test for the existence of a credit channel in Germany. In contrast to the above listed studies, it covers the entire German banking population on an individual basis and it explicitly takes into account bank–individual seasonal patterns. While the results for Germany that are presented in Ehrmann et al. (chapter 14 in this volume) are also based on a data set characterised by these two features, this chapter extends the analysis in two directions. First, it uses bank-specific income and risk variables to improve the control for differential movements in loan demand. Second, it takes into account the network structures of the German banking system by explicitly looking at the role short-term interbank deposits play for banks' loan reaction to monetary policy.