| Credit risk modeling has grown significantly over the past few years; driven by explosive growth in the credit derivatives market and the outlook of more quantitatively sophisticated bank capital regulations under the upcoming Basel II Accord which makes credit risk capital requirement more risk-sensitive. Thus credit risk is attracting strong interest from all market participants, financial institutions (commercial banks, investment banks, hedge funds), regulators (US Federal Reserve, FDIC, Bank of International Settlements) and investors.; Credit risk modeling relies mainly on three parameters, namely the probability of default (PD), the recovery rate (RR) and the correlation structure among borrowers (when computed on portfolios of debt instruments).; This study intends first to explain the implied "correlation skew" using the term structure of the interest rates, particularly the liquidity premium theory, (the "correlation skew is observed because the market implied correlation does not correspond to the single constant correlation assumed by the Single Factor Gaussian Copula, as suggested by David Li1). I show that liquidity has some explanatory power on the correlation that can explain the clustering at the lowest and highest tranche. Within Merton2 model, and given the scarcity in defaulted debt instruments data, the market generally uses equity returns as a proxy for assets returns. The second section analyses the relationship between credit default swap index spread and stock market returns and shows that effectively there is a relationship between both markets.; The third section provides a comprehensive analysis on the cyclicality of default rates, recovery rates and their dependence using financial data provided by Bank Call Reports3 from 1991 to 2005 for all US commercial banks with total assets greater than {dollar}300 millions. It shows that indeed, default rates4 and recovery rates are cyclical and inversely related. These findings have important implications in credit risk modeling for both the credit derivatives market and the new Basel II capital requirement proposed rule because not only current market practice assumes constant recovery rate5 and independence between probability of default and recovery rate, but also regulators are still debating on how banks should include a downturn LGD6 (loss given default).; 1Li, David (2000),"On default Correlation: A Copula Function Approach", The Journal of Fixed Income, March 2000. 2Merton, Robert (1974), "On the pricing of corporate debt: the risk structure of interest rates", Journal of Finance, VOL 29, 449-470. 3The Federal Financial Institutions Examination Council provides Banks Call Reports to the public on its website www.ffiec.gov. 4This dissertation applies a similar approach to that of Georges French by using the charge-off rate as a proxy for default rates. French, George (2003), "Estimating the Capital Impact of Basel II in the United States", Federal Deposit Insurance Corporation, December 2003. 5Market quotes given by Bloomberg which usually come from several suppliers such as Morgan Stanley are usually based on the assumption of a constant recovery rate of 40%. 6see Basel II: International Convergence of Capital Measurement and Capital Standards: a Revised Framework, www.bis.orq and also the Quantitative Impact Study at www.ffiec.gov/qis4. |
