| Empirical research on financial distress prediction is one of the most important research subjects. As an important part to the economic and financial prediction system, the research has not only high academic value but also enormous practical value/With the constant development of the capital market in our country, the number and scale of the listed companies are extending. But at the same time, the number of the listed companies, which are special treated, is increasing. Companies' get into financial distress is a developmental process; the characteristic of the stage in the financial is different. Discovering the financial distress in an earlier time is very important to the enterprise and its customers. Due to the great interests to the topic and special concern about the value of its application, in this Paper, we try to construct financial distress stage prediction models of Chinese listed firms based on the review of previous empirical results.From the definition of financial distress, the measure of the stage, the design of study examples to the choice of the variable and statistical methods on financial failure early-waning research, the author puts forward the thought way about it, choosing 44 serious financial crisis companies, 44 light financial crisis companies and 44 health financial companies as the sample of estimating, using Fisher Discrimination method, take the single year sample date and average sample date respective and build two financial distress stage forecasting model ,then testing the results of the model using both Self-validation and Cross-validation methods. The results are as the following: firstly, the accounting data are not imprecise, but they show some information. Secondly, it is possible that we divided the financial distress into different stages ; Thirdly, no matter use Self-validation method or Cross-validation method, the results of the average sample date is better than single year sample, it is shown that the... |
PDF Full Text Request