주가정보를 이용한 신용위험 평가모형에 관한 연구

Abstract

Most of all, KMV model is practical and accurate method because it estimates a firm's default probability using traded stock price. This paper empirically tests KMV model using the financial data of Korean firms for the period from 1996 to 2000. First this paper estimates DD( Distance to Default ) and EDF( Expected Default Frequency ). Second this paper tests if the value of DDs are different between default firms and matched healthy firms. As the result, first EDF generally decreases as DD increases. Second Wilcoxon rank sum test shows that difference of DDs between default firms and healthy firms isn't valid for distributions at 95% of the confidence intervals. But sign test shows that medians of DDs of default firms differ from those of healthy firms at 95% of the confidence intervals. These results implies that KMV model could have been useful for predicting the defaults of Korean firms in advance.