基本因子能否解釋信用傳染效果?

Abstract

本篇的主旨在於觀察信用傳染效果是否存在以及它是否能被基本面因子所解釋，我們採用Duan, Sun and Wang (2010) 的簡約式遠期違約強度模型，並利用最大概似估計法進行估計，以及進行樣本外的違約預測分析，模型中同時檢驗了代表信用傳染的變數、總體經濟變數以及公司特有的變數，我們發現破產事件確實與其他存活公司之違約機率有顯著正向關聯，而在控制總體與公司變數下，產業內仍存在傳染效果，但並沒有不分產業的傳染效果。

The purpose of this paper is to examine whether the contagion effect exists and if the contagion effect can be explained by firm-specific covariates and macro factors. We follow the reduced-form forward intensity approach by Duan, Sun, and Wang (2010) and adopt the maximum likelihood estimation method. The analyses of out-of-sample default prediction are also performed. We incorporate the variables associated with credit contagion, firm specific variables and macro factors in the empirical tests. We find a significantly positive relationship between default event and default probabilities of survival companies. After controlling for macro factors and firm-specific covariates, the intra-industry contagion effect still persist, but no evidence for the inter-industry contagion effect.