A Credit Risk Model Incorporating Microstructural Dependencies and Stochastic Recovery
by Matthias P. Jüttnery of the University of Zürich & Swiss Finance Institute
Abstract: A credit risk model for determining aggregated portfolio losses is suggested. Beside the common macrostructural dependencies between asset and recovery value, we incorporate possible inter-firm relations among the obligors of the portfolio. Through this channel we also establish related default probabilities and correlation between probability of default and loss given default. Proposing a recursive approximation we obtain explicit representations for the asset value. We compare the model with the IRB approach of Basel II with respect to the implied asset correlation and comparative statics of the conditional expected loss. Due to this analysis of the dynamics we realize that the regulatory framework neglects many important relations, e.g., among obligors and also between risk components. An adequate risk assessment is consequently not possible.