Systemic Risk Modeling : How Theory Can Meet Statistics

We propose a framework to link empirical models of systemic risk to theoretical network/general equilibrium models used to understand the channels of transmission of systemicrisk. The theoretical model allows for systemic risk due to interbank counterpartyrisk, common asset exposures/fire sales, and a 'Minsky" cycle of optimism. The empiricalmodel uses stock market and CDS spreads data to estimate a multivariate density of equityreturns and to compute the expected equity return for each bank, conditional on a badmacro-outcome. Theses 'cross-sectional" moments are used to re-calibrate the theoreticalmodel and estimate the importance of the Minsky cycle of optimism in driving systemicrisk