내부회계관리제도 취약기업 예측에 관한 연구- 코스닥상장기업을 중심으로 -

The purpose of this study is that the prediction model of firms with internal control deficiencies is formulated with various index of financial data and stock returns. We use KOSDAQ samples in 2008 because auditors disclosed internal control deficiencies more aggressively than ever. First of all, we select the variables included in the prediction model of firms with internal control deficiencies. Through logistic regression analysis, debt ratio, operating cash flow ratio, net income growth rate, total asset turnover, volatility, size, and opinion are included in the prediction model. And through discriminant analysis, debt ratio, ROA, net income growth rate, total asset turnover, volatility, size, and opinion are included in the prediction model. Based on predictability, hit rate of logistic regression model is higher than that of discriminant model, and based on verifiability, logistic regression model and discriminant model are similar in AUROC value and hit rate. We conclude that discriminant model is superior to logistic regression model because type Ⅰ error of discriminant model lower than that of logistic regression model. This study contributes that investors will be able to determine for theirselves which firm has internal control deficiencies.