Credit valuation adjustment

Specialeforsvar ved Henrik Reinau Lang

Titel: Credit Valuation Adjustment

Abstract: This thesis examines Credit Valuation Adjustment; counterparty credit risk calculated in terms of market value. The aim of this study is to investigate the sensitivity of CVA towards uncertainty in interest rates and the sensitivity towards correlation between interest rates and default intensities. An experiment of the impact of including netting agreements is performed as well. An introduction to counterparty credit risk is given, followed by a specification of the stochastic models needed in the mathematical framework. Financial concepts of zero coupon bonds, interest rate swaps and credit default swaps are introduced and used to calibrate model parameters to real-world data and calculate CVA. The sensitivity analyses are carried out by changing the parameters in the short rate model and in the model of correlation between short rates and default intensities, respectively. The results reflect financial intuition and show that CVA is increasing in uncertainty in interest rates. Furthermore, CVA is increasing in correlation levels implying Wrong Way Risk, whereas it is decreasing in correlation levels implying Right Way Risk. As expected, the inclusion of netting agreements is empirically proven to reduce CVA. Trying to estimate the correlation parameter is not a straightforward task and is suggested for further research of the subject. Another proposed extension is to include more financial instruments in larger volumes compared to this thesis. Even though it makes the analyses more cumbersome, it would increase the topicality of the study 

Vejleder: Rolf Poulsen
Censor:   Bjarne Astrup Jensen, CBS