Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This work proposes a credit risk model for large panels of financial institutions in which default intensity interdependence is induced by exposure to common factors as well as dependence between entity specific idiosyncratic shocks. In particular, the idiosyncratic shocks have a sparse partial correlation structure that we call the bank credit risk network. A lasso estimation procedure is introduced to recover the network from CDS data. The methodology is used to study credit risk interdependence among European financial institutions. The analysis shows that the network captures a substantial amount of interconnectedness in addition to what is explained by common factors.