The Default Dependencies Analysis Of Credit Risk Models

In Credit Risk Management, choose risk measurement indicators and deal with dependence problems is the two key issues, reseach on correlation structure of financial theory has always been an important issue. There are two main dependence problems in Portfolio management of credit risk: first, refers to the mutual relations of default between different customers in loan portfolios, or breach of a customer default arising from the possibility of another customer, which is default correlation coefficient. Second, refers to the assets relevance in loan portfolios, that is the dependence of assets value between one borrower and another borrower, namely assets correlation coefficient. The identification and modeling with correlation structure is a key skill to risk management researchers. If a model can reflect the dependence of the structure accurately, then we can conduct a more reasonable financial instruments, more accurate pricing.We commence with an overview of the three most widely usedcredit risk models developed by KMV, we J.P. Morgan (CreditMetrics) and Credit Suisse First Boston (CreditRisk+). The mathematical essentialsof each model lie in the way the joint distribution of the so-called 'defaultindicators' is modeled, a vector of Bernoulli random variables. With the focus on these vectors we will investigate two general frameworksfor modelling such binary random events. We will also show how the KMV and CreditMetrics methodology can be translated into the framework of CreditRisk+.The credit risk models are then compared for ' homogeneous' portfolios using Monte Carlo simulation. As two of the three models use the multivariate normal distribution for their 'latent variables', we investigate the impact when proceeding to the broader class of elliptical distributions. A so-called t-model, incorporating a t-copula for the latent vector, shall be used to show the consequences of a possible generalisation. In this context we introduce the notion of tail dependence. Comparison of the extended t-model with the 'normal' two credit risk models will again be performed for the same types of portfolios used for the previous comparison. we can couculd out that t-model can produce more extreme events than KMV/CM model,which appears closer to reality.