A unified credit risk model

The aim of this thesis is to bridge the gap between the two main approaches in credit risk modelling: structural and reduced form models. There are many attempts to do this using special assumptions. For instance, Duffie and Lando(1999) makes the connection between the two approaches by using incomplete accounting information. We propose here a unified approach that attains this objective in full generality. We use a model where the credit default event is defined as the minimum of the two default times, one from the structural default and the other from the events with an exogenous default intensity.; In particular, we look at the effect of having stochastic volatility in the structural approach. We study the effect of multiple time scales on the credit spread yield curves both for stochastic volatility and for stochastic intensity. We use perturbation analysis to derive closed-form approximations for the credit spreads that allow simpler and more efficient calibration of parameters.; The main quantities that we analyze are not only related to fixed-income market data, such as bond spreads, but also to equity market data such as variance and leverage ratio. We consider the calibration of the model and its stability with respect to parameters for particular names with different credit ratings. We discuss in detail the differences of the behavior of credit spreads between high and low rated names, and we interpret these differences in the context of our model. Our results show emphasize the importance of equity market data, such as variance and leverage ratio and also of the stochastic default intensity.