A study of Models for Predicting Financial Distress in Chinese Listed Companies

For limitations in financial ratio selection,discriminant time point definition and exact model confirmation with studies of financial distress prediction in China,we make an in—depth analysis of financial ratios,then pick up 120 ST firms and their matched 240 non—distressed firms through 2005—2009 as the sample,use stepwise discriminate analysis to establish Z—score models based on information from 2002—2008.Conclusions are:(1)Compared with the model constructed by traditional financial ratios,the discriminate accuracy is higher with the model constructed by us.(2) With studying movements of the stock price between ST firms and non—distressed firms,we confirm the discriminant time point should be t—3,and the discriminant model should be constructed by data in t—3.(3) With another sample(96 firms) that is not for model establishment to test the model's predictive power,to the financial distressed firms, when we make a discrimination three years before being distressed,the accuracy is 81.25%,thus it's a well discriminator for distinguishing distressed firms three years ahead.