Valuing Credit Default Swaps with a Structural Approach

Abstract: Valuing single-name Credit Default Swaps (CDS) is a dicult task since in order to make a fair valuation, one needs to assess the credit risk of the corresponding company. Many dierent models exist when it comes to modelling the credit risk, this report specically focuses on the branch of models named structural models. The aim of this thesis is to, for a number of companies, model the CDSspreads given by the market by rst modelling the credit risk of the company in the aforementioned models, and then using key metrics calculated in the models to value the corresponding CDS-contract. A couple of dierent models are tested with dierent sets of key parameters, and the results show that a certain implementation of the Black-Cox model produces best results, which also happens to be the model with the most realworld like features. The model manages to follow both major changes in the CDS-spreads, as well as minor changes. The corresponding residuals show some stationarity-features and are slightly improved when adjusting for the current level of volatility, however, they do not appear to be white noise as corresponding SACF-plots show clear correlation for many lags. The Black-Cox model also proves to be better than a simple regression model, used as a benchmark.