Logit bankruptcy model of industrial product firms / Asmahani Nayan ... [et al.]

Logistic regression or logit model is one of the statistical methods that has been widely used in bankruptcy studies. Logistic regression is appropriate when the dependent variable is binary while the independent variables are either discrete or continuous. In bankruptcy studies, the dependent variable that is being used has only two categories which are failed firm and non-failed firm. Besides logistic regression there are other methods that can be used in bankruptcy studies such as Altman’s z-score model and multiple discriminant analysis. These methods act differently to different data sets which also give different accuracy rate. The purpose of this study is to compare the performance of logit model and Altman’s z-score model in predicting failed and non-failed firms. A total of 30 industrial product firms in Malaysia (15 failed firms and 15 non-failed firms) are used in this study. The firms were divided into training and validation sample then replicated into three groups and in each group will have 70 percent estimation sample and 30 percent validation sample. The performance of the two models were measured using accuracy rate, type I error and type II error. Results of the training and validation samples implied that logit model is slightly better than Altman’s z-score model with higher value of accuracy rate and lower value of type I and type II error.