| With market economy system reform deepening and capital market increasingly developing in China, study on corporate finance early warning becomes an urgent issue. It is of great practical significance for all market attending parts to predict listed corporates'financial trend.Referring to Professor Altman's philosophy about financial early warning system, this article does an empirical study to design financial early warning models based on financial datum and ratios. It employs statistic method as multiple discriminant analysis and cross validation. 27 financial distress companies and the same number non financial distress companies of Huadong area are sampled to establish models according to the definition of financial distress companies who came to'ST'for the first time during 2004 since it is listed in Shenzhen or Shanghai. First, there are 61 financial ratios in the study, and then 10 ratios are left after the stepwise method. Furthermore, the correlation coefficients between the 10 ratios are not so high, which means that they stand for different aspects of the operating. As a result, all the 10 ratios can come into the model. Unstandardized canonical discriminant functions and the discriminant rules are established after the process of discriminant. At the same time, the ratios'proportion is also analyzed from financial point. At last the test result of the model is analyzed.In order to test the applicability of the model, predicted samples from both Huadong and non-Huadong are used to get a further test; the result shows the model plays an important role in the financial warning discriminant. In the end, the article gets four conclusions and the same number shortcomings of the research. The trend of financial early warning system is also referred. |
