Empirical Research On Predicting Financial Crisis Of Listed Companies

The research on predicting financial crisis is at early stage in China, and it has high practical value. So it is a field full of attraction, an extremely arduous and challenging exploration process. Based on the former research achievements, this paper makes an empirical study on prediction of financial crisis. It begins with the analysis of the basic conception and factors, followed by the reviews and comments on the literatures both domestically and overseas in this area. And then, Based on the theory analysis above, the empirical research is done. There are one hundred and sixteen listed companies in Shanghai and Shenzhen Stock Exchange. In this paper, fifty-eight ST (Special Treatment) listed companies and fifty-eight non-ST listed companies in 2003 are chosen as research objects, and Logistic Regression and Fisher Binary Discriminate Analysis are selected to be the methods of research. According to the principles and present condition of the listed companies of China, the paper sets up an index system made up of twenty financial indexes, such as Current Ratio and Return On Equity, which synthetically reflect the financial condition of listed companies at all directions. The paper computes the three years' value of the financial indexes before the crisis of the object companies, and through the section analysis, seven notable indexes for predicting financial crisis are found. So they are defined as the variables to establish the predicting models. Through Logistic Regression and Fisher Binary Discriminate Analysis, the predicting models ultimately are established. Furthermore, the paper compares the predicting effect of the different methods, and draws some conclusions from the different methods.