在過去的文獻中,信用違約交換價值訂價決定於違約機率以及回復率。很多實證研究證實違約機率以及回復率無法完全解釋信用違約交換的市場價格,亦即,這兩個解釋因子對信用違約交換市價的解釋能力有限。故許多實證研究嘗試用不同的解釋因子〈如:債券流動性、總體因子〉來解釋殘差的部分,然而,實證的結果並不顯著。 信用違約交換可以被包裝成合成式的擔保債務憑證或者做為規避信用風險之用途。近年來,許多投資者以及交易者會以信用違約交換或者擔保債務憑證分層並根據自己對於各別信用違約交換相關性的投資看法做交易。所以,我們認為相關性可能是影響信用違約交換的一個重要因子。故在本研究中,我們利用公司債券導出理論的信用違約交換的價格,並且我們也發現理論價與實際市場價格的確存在顯著性的差距。我們進一步利用相關性做為解釋因子來解釋理論以及實際市場價格的差距,以證實我們的假設〈信用違約交換除了受違約機率以及回復率影響外,也受相關性影響,所以相關性可以解釋價差〉。最後,我們的結果支持最初的假設,結果顯著。 In theoretical pricing model, credit default swap value is dominated by two factors: default probability and recovery rate. In many empirical researches, these two factors can not determine credit spread changes. These variables have rather limited explanatory power. There are lots studies try to use other variables (i.e. bond illiquidity, macroeconomic factors) to explain the residuals; however, the results do not significant. Credit default swaps can used to pack a synthetic CDO (Collateralized Debt Obligation). Recently, many investors and traders will trade CDS or CDO tranches by their view on correlation. Therefore, we consider correlation can be a factor to affect CDS spread. In this paper, we use corporate bond to derive model CDS spread. Compared with market CDS spread, model CDS spread is significant different from market CDS spread. Our result suggests that the pricing difference comes from correlations among CDSs.