In this paper we suggest a new approach to risk assessment for banks. Rather than looking at them individually we try to undertake an analysis at the level of the banking system. Such a perspective is necessary because the complicated network of mutual credit obligations can make the actual risk exposure of banks invisible at the level of individual institutions. We apply our framework to a cross section of individual bank data as they are usually collected at the central bank. Using standard risk management techniques in combination with a network model of interbank exposures we analyze the consequences of macroeconomic shocks for bank insolvency risk. In particular we consider interest rate shocks, exchange rate and stock market movements as well as shocks related to the business cycle. The feedback between individual banks and potential domino effects from bank defaults are taken explicitly into account. The model determines endogenously probabilities of bank insolvencies, recovery rates and a decomposition of insolvency cases into defaults that directly result from movements in risk factors and defaults that arise indirectly as a consequence of contagion. JEL classification: G21, C15, C81, E44
