Credit Derivative Pricing Under Levy Models

Financial institutions need credit derivative instruments to protect portfolios against failure events. By mitigating risk exposures, credit derivatives are helpful for relaxing Basel II capital requirements. This has lead to an exponential growth of the trades of such instruments over the last decade. The financial community is now demanding models for pricing credit derivatives. Modelling credit risk supposes to define a default event and to estimate its associated default probability. Two main approaches are available. Structural models link a default event to the economic and financial condition of the company. A default occurs when the company asset value falls below a predetermined barrier. Intensity based models describe the credit quality of the obligor as a stochastic process. A default happens at the first jump of a Poisson process with stochastic intensity. Regardless the creativity of the modelers, the industry remains unwilling to adopt any standardized approach. Almost all models proposed so far fail to meet some important requirements. Namely, a model needs to be clearly interpretable, it must replicate market features, and it should be suited to multivariate settings. Finally run-time delay must be reasonable. In this work, we analyze the value-added of Levy-processes for modelling credit-risk. We will argue that such processes are indeed able to meet the above-mentioned requirements. Applying Levy processes to both structural and intensity models, we will evaluate their performance through pricing and calibration to real market data of credit default swaps. Each model will be extended to a multivariate setting.

CARIBONI Jessica; 

2007-06-25

JRC37459