A Study Of Models For Predicting Failure In China's Listed Companies

With the rapid development of china stock market, the listed companies grow in quantity and scale, the industry and operation of them getting better. They have become the most dynamic and growth companies in the modern market economy. But some listed companies' performance is bad, even lead to failure. The listed companies' failures not only give investors and creditors loss, but also affect the allocation effectiveness of resources. Therefore the prediction of listed companies' failure is very important.The process of listed companies' failure is gradual, so we can use effective methods to predict the failure. This article analysis China listed companies' features and reasons, then redefinition the failure. Measure the failure from flowing and stocking, which is different from the general research definition ST as failure.Because manufacture industries are different from non-manufacture industries, we construct prediction model separately. Using fisher linear discriminant model and logistic regression analysis model, using financial variables and non-financial variables, we are looking for the most accurate variables and models to predict the failure.The results show that fisher linear discriminant model and logistic regression analysis model are effective when used to analysis manufacture industries, non-manufacture industries and all industries. However, manufacture industries model is better than all industries model.