سیاست های پولی و وام دهی بانک : شواهدی از گروه بانکی آلمان

This paper analyses the impact of monetary shocks on bank lending in Germany. We follow a cross-sectoral approach by looking at six different banking groups. In general, smaller banks hold a larger buffer of liquid assets which they can use to offset monetary shocks. In addition, the response of bank lending after a monetary contraction is very different across banking sectors. Lending by the credit co-operatives, which are on average the smallest banks, declines most, whereas big banks are able to shield their loans portfolio against monetary shocks. Overall, our results provide support for the existence of a bank lending channel.

The purpose of this paper is to provide empirical evidence on the role of banks in the monetary transmission process. The implications of the German institutional setting for the impact of monetary policy on bank lending are a priori ambiguous. On the one hand, the mere fact that banks play an important role suggests that the scope for an effective bank lending channel is potentially large. On the other hand, banks may try to shield their loans portfolio from monetary disturbances which may weaken, rather than strengthen, the impact of monetary policy. The latter may be particularly relevant for Germany, given the importance of long-run relationships between banks and clients in this country. In this paper, we look whether evidence can be found for a ‘bank lending channel’ of monetary policy, by considering the response of bank lending to monetary shocks. It is well known that this kind of research is complicated by a serious identification problem: is the fall in bank lending after a monetary tightening induced by supply or by demand? Several recent studies based on US data have addressed this problem by analysing disaggregated data, either for borrowers (e.g. Gertler and Gilchrist, 1993a and Gertler and Gilchrist, 1994; Gilchrist and Zakrajšek, 1998) or for lenders (e.g. Kashyap and Stein, 1995 and Kashyap and Stein, 2000). We follow the latter approach, by considering different banking groups, as defined in the Bundesbank’s Banking Statistics. In this way, we capture one key element of the bank lending channel, namely that some types of banks (particularly the smaller ones) face more information problems and find it more difficult to neutralise monetary shocks than other types of banks (typically large ones). Is it still sensible to look at Germany separately after the start of EMU? After all, monetary policy is first and foremost based on euro-wide aggregated data now. However, credit markets are still likely to exhibit specific national characteristics (see e.g. De Bondt, 2000). Hence, as cross-national differences in monetary transmission may complicate the implementation of a common monetary policy, it is still useful to consider individual EMU economies. The organisation of this paper is as follows. Section 2 discusses recent literature on monetary transmission and bank lending. Section 3 gives an overview of the German banking system and discusses key characteristics of banking groups. In Section 4, we present our empirical results. We look at cross-sectoral differences in balance sheet structure and present dynamic simulations of the response of bank lending to monetary shocks. Section 5 concludes.

In this paper, we have analysed the response of bank lending to monetary shocks, focusing on differences between German banking groups. We discussed the main features of these groups, presented the most important differences in balance sheet structure, and carried out impulse–response simulations to show the main differences in lending behaviour. We focused on two key predictions: (i) for small banks it is more important to invest in a buffer of liquid assets than for large banks, and (ii) small banks find it more difficult to shield their loans portfolio after a monetary contraction than large banks. If these hypotheses can be established empirically, this would support the existence of a ‘bank lending channel’. Especially for the largest and the smallest banks, our results are consistent with both predictions. Big banks have relatively little liquid assets and are nevertheless able to insulate their lending activity from monetary disturbances, whereas the credit co-operatives have a relatively large amount of liquid assets but still have to reduce their loans portfolio after a monetary contraction. The outcomes for two banking groups, private banks and co-operative institutions, are hard to interpret along the lines of the bank lending channel. However, as these two banking groups have a very small market share, they have little consequences for our overall conclusion. Although our results offer some support for the existence of a bank lending channel, one should be cautious, as the limitations of our approach do not allow strong conclusions. We only focused on the ‘first stage’ of this transmission channel, i.e. the impact of monetary policy on bank lending. In addition, we analysed banks at a sectoral level, whereas a further disaggregation would make it possible to perform more precise tests of bank behaviour. It would be useful to carry out a similar analysis with bank level data, as Kashyap and Stein (2000) have done for the United States, in order to obtain more rigorous conclusions.