재무비율을 이용한 부도예측에 대한 연구: 한국의 외부감사대상기업을 대상으로

After the 1997 Asian financial crisis South Korea experienced a serious financial hardship. Immediate economic effects of this crisis were felt in all areas of the Korean business world,and many companies were filed for bankruptcy. The main purpose of this study is to develop abankruptcy prediction model and to examine the power and accuracy ratio of the model usingthe data of Korean companies. We use all manufacturing companies audited by external auditorsas sample firms during the period from 2003 to 2007 and use the transaction information of thecurrent bank account from the KFTCI(Korea Financial Telecommunications &Clearings Institute)to identify the right time for bankruptcy. Multi-variate Logistic Regression was used to form aprediction model with 117 financial ratios. The major findings of this study are as follows. First, we obtained 17 financial ratios from the stepwise selection method of logistic model andused these ratios in our bankruptcy prediction model based on multi-variate analysis as keyinput variables. The estimation results show that the concordant ratio which represents thediscriminant power of the model is 82%. Second, The predictive power of the bankruptcy predictionmodel, AUROC is 85% and Kolmogorov-Smirnov statistics is 55.04%. These results mean thatthe model developed in this study can be used as the bankruptcy prediction model for externalaudit-companies in the Korean markets.