A comparison of estimated default probabilities: Merton model vs. stable Paretian model

This paper is devoted to an empirical comparison of probabilities of default obtained from two structural type credit risk models: the Merton model and the stable Paretian model. First, we prove on our data that log-normal distributional assumptions of the classical Merton model are generally rejected. Therefore, we propose a more general structural model based on stable non-Gaussian processes as a representative of subordinated models. In the application of the models we use Moody's KMV methodology and a new extended KMV methodology respectively. Finally, our results suggest that default probabilities obtained from the classical Merton model are underestimated.