Credit risk assessment and relationship lending:: An empirical analysis of German small and medium-sized enterprises

We estimate a logit scoring model for the prediction of the probability of default by German small and medium-sized enterprises (SMEs) using a unique data set on SME loans in Germany. Our scoring model helps SMEs to gain knowledge about their default risk, which can be used to approximate their risk adequate cost of debt. This knowledge is likely to lead to a detection of hold-up problems that German SMEs might be confronted with in their bank relationships. Furthermore, it allows them to monitor their bank's pricing behavior and it reduces information asymmetries between lenders and borrowers. Finally, it can influence their future financing decisions toward capital market-based financing.