A Study Of CDO Pricing Methods And Their Applications

Collateralized Debt Obligation (CDO) is a very innovative financial derivative, andhas made great success and rapid development in financial market before the financialcrisis in2008. CDO has many advantages. For sponsor, it can remove loans and otherdebts from banks’ balance sheet and lower the regulatory capital requirements; or earnspread between the collateral asset and tranches. For investors, tranche with different riskcan meet their risk appetites, also, hundreds of assets in collateral pool can scatterinvestors’ investments.When pricing the spreads of CDO tranches, the biggest problem is how to model thedefault times of reference entities—because the defaults are extreme and related events.Copula function is a popular method to solve this problem. In this article, we willintroduce the definition, property and classification of copula, and also procedures ofsimulating multivariate correlated variables under different copula types. In the empiricalpart, we assume a collateral asset pool and a tranches structure, then compare the pricingresults under different copula types. We analyze the spreads’ sensitivity of parameters ingauss copula. At last, take xinyin2008for example, we analyze the information ofcollateral asset pool and tranches of ABS product of China, and different pricing resultsunder different copula types.In the empirical part, we can find that doubt t copula model with both five degrees offreedom has the best performance, Clayton-Copula and t-Copula model perform worst andGauss-Copula between the two. But not always doubt t copula model has the best result, inparticular circumstances, we can use other pricing models to obtain optimal results.