In the aftermath of the credit crisis 2007-2008, credit derivatives and counterparty risk have been pointed out as being major drivers of the crisis. Before the crisis many monoline insurance companies and large investment banks, had been considered as being risk-free institutions. When the crisis hit, many of these institutions turned out to be not-so-risk-free, and decreasing credit quality along with rating downgrades, led to huge CVA losses for their counterparties. In this master thesis, I present a model for calculating bilateral counterparty risk for credit default swaps under collateralization and correlated default events. The thesis consists of two main parts. In part I, I set up an intensity based model for the valuation of credit default swaps, and calibrate this model to observed market spreads. In part II, I set up a formula for pricing bilateral counterparty risk for a credit default swap under collateralization. Using the CDS model from part I, I present a numerical method for calculating the counterparty risk under correlated default events through simulations. The numerical calculations turn out to be very cumbersome, and to ensure stable pricing of counterparty risk one need a lot of computational power This is one of the main drawbacks of the model. However, through a number of simplifications, one is able to analyze the effects of collateralization and correlation. Using a specific example, I show how both CVA and DVA is heavily affected by correlation. It is also found that collateral has a risk reducing effect for low correlations, but when correlation increases, this effect vanishes. Therefore, under high correlations one is not only faced with high counterparty risk, but is also lacking the possibility of reducing this risk through a collateral agreement.
|Educations||MSc in Mathematics , (Graduate Programme) Final Thesis|
|Number of pages||102|