Abstract

Having a precise idea of how information is used is a key element in studying credit risk models. This concept plays an important role in structural and reduced form models and most recently in information based models. In this thesis the relationship between these models and the idea of information, more specifically through filtration expansions, is studied in depth. Special attention is given to the study of intensity processes under different types of filtration expansions.

Credit derivatives are path dependent financial products. Therefore their analysis is based on the history of the underlying risky process. If the underlying process is allowed to have jumps, then this analysis is more challenging. This explains why, normally, risk management techniques for these products assume that the underlying process is continuous, the derivative is path independent, or the probability measure is risk neutral. In our model, in the context of a locally risk-minimization approach, the problems of pricing and hedging of defaultable claims are discussed without imposing any of the above assumptions.

The impact of risk measures in financial markets can no longer be ignored. Considering this, a methodological procedure based on risk measures is developed to gauge the credit quality of defaultable bonds in real bond markets. Through this process a new type of indicator is introduced that can be useful to detect inconsistencies in bond markets. This can be helpful in market integration applications.