This dissertation analyses one-factor copula models for credit portfolios and applies them to pricing as well as to risk measurement and management of synthetic CDOs. First, the correlation smile problem of the Gaussian copula model is improved by application of the NIG distribution. Further, other model extensions are introduced that provide higher model flexibility and have better properties. The final Crash-NIG model allows several correlation regimes and represents a suitable model for simulation based risk management applications.
«This dissertation analyses one-factor copula models for credit portfolios and applies them to pricing as well as to risk measurement and management of synthetic CDOs. First, the correlation smile problem of the Gaussian copula model is improved by application of the NIG distribution. Further, other model extensions are introduced that provide higher model flexibility and have better properties. The final Crash-NIG model allows several correlation regimes and represents a suitable model for simul...
»