Description

The devastating impacts of the recent global financial crisis underscore the need for both financial institutions and banking supervision to develop more appropriate credit risk models to ensure the stability of the financial system. This work contributes to quantitative credit portfolio risk modeling in three ways. First, it introduces a general credit portfolio modeling concept that comprises specific credit risk management models as special cases. Second, analytical techniques are presented for specifying asset correlations in a credit portfolio through systematic factors. Finally, a new approach for clustering of obligors in a credit portfolio is proposed using threshold accepting, a stochastic optimization technique. In particular, a computationally tractable technique to validate ex-post the precision of the clustering system is suggested and applied to a real world retail credit portfolio. The contributions of this book should provide benefit to practitioners, academics and graduate students in the field of financial risk management.