This paper shows that the substantial disparity in German bank lending towards industrial (IC) and non-industrial (Non-IC) countries is largely explained by differences in countries' endowments and only to a minor extent by German banks' different treatment of these country groups. This is demonstrated by applying a decomposition technique to an augmented gravity model that is estimated for German foreign lending using a new micro panel data-set on individual claims from the Deutsche Bundesbank covering the period from 1996 to 2002.
Keywords:German bank lending, gravity models, Oaxaca decomposition analysis JEL classification: F30; F34;G21
Non technical summaryIn theory, capital flows provide risk sharing opportunities for both debtors and creditors and can enhance economic efficiency. In practice, however, capital inflows can rapidly turn into capital outflows thereby creating economic instability or even financial crises, especially in developing countries.Broadly speaking, capital flows can be divided into foreign direct investment, portfolio investment, and bank lending. While bank lending has lost in importance as a source of finance in the 1990s as compared to the 1970s or 1980s, it still amounts to about 20 percent of capital inflows. As German banks are among the major creditors to industrial as well as developing countries, their lending behaviour is of particular interest. This paper analyses German banks' lending behaviour towards industrial and non-industrial countries. In particular, the discrepancy in bank lending to these two country groups is explained as about 85 percent is directed to industrial and only about 15 percent to non-industrial countries.The analysis is conduced in two steps. First, the determinants of German bank lending are estimated using a gravity model positing that claims depend positively on recipient countries' economic size and negatively on the distance to Germany as well as other factors. Second, the lending gap between IC and Non-IC countries is decomposed into one part stemming from differences in countries' endowments and one part that results from German banks' different treatment of these country groups with respect to their fundamentals.The findings are that market size proxies, distance from Germany, foreign direct investment and country risk explain most of the variation in German foreign claims. With respect to the lending gap between IC and Non-IC countries, it turns out that differences in countries' factor endowments explain more than two thirds of this gap, while less than one third is unexplained and hence due to different treatment of countries' fundamentals and a residual of unobserved factors.
Nichttechnische Zusammenfassung
Aus theoretischer Sicht bieten internationale