محدود کردن جعبه سیاه : گروه بانکی آلمان در انتقال سیاست های پولی

The paper examines the role of portfolio decisions of German banking groups in the transmission of monetary policy. A small econometric model is augmented with segregated bank balance-sheet variables in order to study potentially different effects of tight money on these variables. The results provide evidence for banks playing an active role in the transmission process and are beneficial for two fields of research: The findings indicate sectoral differences of monetary policy in line with transmission models based on informational frictions. In addition, the results lend econometric support to the existence of close bank–customer ties for certain banking groups in Germany: Whereas large banks adjust their credit position markedly following a monetary shock, the smaller savings-banks and credit cooperatives seem to shield their customers from the brunt of a policy tightening.

Models studying monetary policy made extensive use of a black box ( Bernanke and Gertler, 1995) to circumvent the analysis of the tedious interactions between borrowers and lenders in capital markets. However, by their very nature, black boxes are transitory research tools at best, designated to facilitate step-by-step incorporation of variables not yet explicitly modelled. Nevertheless, for the time Arrow–Debreu contract wonderland being the reference point, monetary economics made no substantial efforts to trim the box in size. Renewed interest in monetary transmission corresponds to the resurgence of small econometric models. They assembled stylised facts on the macro-dynamics following a monetary shock that were hard to explain by traditional transmission theories (e.g. Bernanke and Gertler, 1995). Timing and magnitude of the responses could, however, be replicated by models who left the orbit of perfect capital markets and incorporated real world phenomena such as imperfect and asymmetric information. Whereas this spurred research foremost in the U.S., European interest has kicked in just a few years ago (BIS, 1995), and still, it is far from being a lively one. This is all the more startling, since the question whether a single monetary policy translates symmetrically throughout the EMU-11 seems far from being settled. An answer in the affirmative is, however, hard to defend even on a priori grounds. Institutional responses to informational frictions in capital markets have evolved historically such that there is no single European finance and banking system (Europäische Wirtschaft, 1997). Thus, it seems worthwhile to dismiss the idea of the financial system as a veil. Instead, a thorough reference to institutional answers concerning the information problem might augment our understanding of monetary transmission substantially. Existing cross-country time-series studies come up with rather inconclusive evidence on potential asymmetries in the transmission of monetary policy.1 Most of them submit some descriptive material on idiosyncrasies in national finance designs, but their econometric model largely abstracts from transmission theories that incorporate external finance premia. Instead, a huge black-box degenerates the financial system to a machine capable of frictionless money-bond swaps. Dornbusch et al. (1998) conclude that the potency of a monetary tightening in terms of real activity is weaker in countries with a bank-centred financial system. They quote France and Germany as examples. Experience in the U.K., generally cited as the paradigmatic example for a market-centred system, shows a stronger susceptibility to monetary shocks. Giovannetti and Marimon (1998), on the contrary, motivate their study with descriptive evidence of two inherently different financial systems in France and Germany. The existence of such disparate views is most likely due to the fact, that the widely held distinction between archetypal market-oriented and bank-oriented financial systems is a very stylised one ( Allen and Gale, 1995; Scholtens, 1997). This approach has its merits in broadly based international comparisons. But, it requires simplifications to a degree that rather obstructs deeper comprehension of financial institutions’ impact on monetary transmission. This paper is thus confined to a national scope, which gives way to model an institutional blueprint that does not rely on rather general features of a financial system. Germany, in particular, provides an interesting case study: Within an inherently universal banking system, there are distinctive banking groups serving their specific clientele. This raises the issue of the significance of close bank–customer ties in the propagation of monetary shocks. By endogenising portfolio decisions of the three dominant banking groups, the paper submits evidence for differential responses of German banks in the transmission of monetary policy.2 The remainder of the paper is organised as follows: Section 2 comments on the strategy to overcome the identification problem pertinent to credit-based transmission models. The innovative part goes to the usage of the Bundesbank's Bankenstatistik to extract information on disaggregated credit flows. Section 3 describes the model specification, Section 4 presents the results, and Section 5 concludes.

The paper studies the response of bank balance-sheet variables to a monetary shock in a small macroeconometric model. To this end some institutional features of the German banking system were put to use, namely the close links between certain banking groups and a specific clientele. The impact of monetary shocks is broadly in line with theories on monetary transmission based on asymmetric information. With some cautions regarding unpleasantly large confidence intervals there is evidence for disperse effects of monetary policy due to bank behaviour which does not fit into a Modigliani–Miller story. The results do not refute the hypothesis that German banks exploit a certain degree of their portfolio flexibility to isolate credit volumes from a monetary tightening. Analysis of disaggregated bank data, however, suggests that this aspect of a housebank relationship is only pertinent to a certain sector of the German banking system: Namely savings banks and credit co-operatives. The real impact of this ‘stylised material’ concerning the role of German banks is hard to assess. Bank credits are undoubtedly the single most important external source of finance in Germany. But, the bulk is raised from internal means (Worms, 1997). It is, however, important to note that this composition is not primarily a matter of the German financial system. Predominant industrial structures typically imply high capital–labor ratios which bring about lofty depreciation figures. It is essentially this depreciation that nourishes internal funds to these levels. But, this does not necessarily devaluate the overall impact of banks credit decisions in Germany: Bank loans are the most likely candidate for funds at the margin and presumably it is this moment that matters.