![IconGems-Print](images/IconGems-Print.png)
504 – Risk, Prediction, and Financial Econometrics
Asymptotic Expansion of One-Factor Merton Models with Non-Gaussian and Serially Correlated Innovations
Takayuki Shiohama
Tokyo University of Science
The one-factor Merton model in the context of CreditMetrics is specialized by a single factor common to all counterparties. We extend the structural credit risk model to a model that includes underlying single risk factor and issuer-specific process have non-Gaussian and serially correlated asset returns. By using a standard Edgeworth expansion, we arrive at the closed-form analytic expressions for the default rate distribution. We also provide estimators of the parameters of the asset value process. Our empirical results illustrate the non-negligible effects of the skewness and kurtosis of the distributions on the systematic risk of credit portfolio risk evaluations.