信用違約交換 (credit default swap, CDS) 在金融海嘯之後成為備受囑目的信用衍生性商品之一,其可用以衡量公司信用風險情況。本文以Easley et al. (2002) 提出的資訊交易機率 (probability of informed trading, PIN) 作為資訊交易行為的代理變數,並利用縱橫資料方法 (panel data methods) 探討2004年至2009年間資訊交易行為與CDS的關係。實證結果顯示,資訊交易行為顯著影響CDS價格,平均而言,資訊交易行為與CDS價格存在負向關係,證明資訊交易行為可當作CDS價格的決定因素。本文進一步利用分量迴歸模型 (quantile regression model) 提出資訊交易行為與CDS價格存在顯著的不對稱關係,說明負向關係主要發生在CDS較高的情況,且在經濟景氣處衰退時,不對稱關係呈現反轉的現象。最後,本文以門檻迴歸模型 (threshold regression model) 證實在高、低不同狀態的資訊交易行為下對信用風險有顯著不對稱的影響,其中,當60天與120天期的資訊交易行為處於高狀態時,資訊交易機率與CDS價格有顯著的正向關係;反之,240天期的資訊交易行為處於低狀態時,資訊交易機率與CDS價格存在顯著的負向關係,而當經濟景氣處於衰退的情況下,僅在60天與120天期的資訊交易行為基礎下,其高、低狀態的資訊交易行為對CDS價格有顯著不同程度的影響。
The credit default swap (CDS) has become the most popular credit derivative and a key economic indicator of credit condition. This paper investigates the relationship between the dynamics of information-based trading and CDS spread, and contrary to prior research, it shows that the probability of informed trading (PIN) should be a key determinant of CDS spread. In addition, through the application of the recently developed method of using the quantile regression model, this paper finds that the effects of information-based trading activity on CDS spread would be conditionally observed with different directions and magnitudes across various price regimes of CDSs. Finally, this paper also made use of the threshold regression (TR) model as evidence of the asymmetry effect of the information-based trading activity on CDS spread during its higher and lower state. The results of the TR suggest that based on the 60-day and 120-day PINs, there exist a positive relationship where a high PIN will cause the CDS spread to increase in higher PIN state, whereas in the lower PIN state based on 240-day PIN, it has an obvious negative relationship. Furthermore, it is only when it is based on 60-day and 120-day PIN that information-based trading activity has regime effect on CDS spread when the market suffers recession impact.