擔保債權憑證(Collateralized Debt Obligation:CDO)為近年來發展迅速的資產擔保證券(Asset Backed Security)之一。其架構主要有一固定收益債權之債券或貸款組合,經由特殊目的公司(Special Purpose Vehicle, SPV)將此債權或貸款群組加以重組證券化包裝後,再依不同信用品質來區分各種等級之分券(Tranche)銷售給投資人,債權產生的現金流量就依照證券發行所設定之條件付息給一般投資人。 在CDO整個架構上,信用風險為整個商品架構上影響最大的部分,因此首先需建構單一資產之違約模式,與債權群組之違約相關性,進而計算證券化之後所應該給予投資人之分券溢酬。其中,本文主要探討資產群組的聯合違約相關性與建構聯合違約機率分配,在文獻上以copula方法最廣為大眾接受,而copula方法在模型上的選擇,會影響評價之準確性與操作上的效率,且由文獻指出選擇方式上並無一固定之準則。本文利用factor copula來建構資產群組織聯合違約機率分配,再予以評價,並比較factor copula在模式的調整上是否有顯著差異,進而尋求較佳的評價方式。
In recently years, credit derivatives become more and more popular. Collateralized Debt Obligation is one of the credit derivatives and the trading volumes are growing fast. CDO is backed by a pool of portfolio and then tranched. When pricing CDO, it is an important thing that gets the correlation amount the portfolio that consists lots kind of assets. Copula method is one of the most efficient way to solve this problem. In this paper, we provide factor copula to price the premium of CDO. By comparing many types of factor copula, we want to find out which types of factor copula are useful and efficient.