企業信用組合訂價之實證分析：現金流量基礎法

Abstract

本研究試圖利用Liao & Chen (2006) 所發展的內部評價法，對市場信用違約指數 (CDX) 進行評價。模型以現金流量基礎法衡量公司資產價值，並運用因子關聯結構法 (factor copula) 建立公司資產價值與景氣因子之連動性，達到投資組合損失模型考慮動態風險之目的。實證結果顯示此模型低估5年期CDX組合之信用風險。本文並提出可改善此低估情形之方法，如︰不同形式之違約邊界設定，考慮隱含現金流量波動性，或以跳躍擴散隨機過程描述現金流量。另外，在模型中考慮具預測性之總體經濟因子，可進一步捕捉稀有事件引起之信用價差跳躍。

We employ CDX market data to empirically examine the effectiveness of the credit portfolio model developed by Liao and Chen (2006), which combines a cash flow based model with a conditional independent default approach. Our empirical results show the credit portfolio model underestimate 5 year credit spread of a Dow Jones CDX credit portfolio. There are several ways to modify the model, including alternative default threshold settings, employing market implied cash flow volatility, or superimposing jumps on the cash flow processes. Moreover, incorporating forward looking macroeconomic factors enables the model to catch rare events induced jumps in the credit spread.