Model Construct And Demonstration Of Listed Corporations' Financial Crisis And Forewarning Based On Data Mining

In furious market competition's environment, many listed corporations get into finance plight and bankruptcy because of imperfect management. Therefore, stand or fall of listed corporations' quality is related to stabilization and development of securities business. It is important to build an effective forewarning system of financial crisis for every interest group, so that the system can help listed corporations forecast financial risk and reduce loss.Based on the former studies, the paper executes a forecast research for financial crisis of our country's listed corporations through analytical method.Using specially treated corporations in 2006 as sample and the net asset profit rate as partnership sample standard, I choose 32 samples of corporation with financial crisis and 32 samples of corporation without financial crisis between 2001 and 2005. We choose 21 financial indexes such as developing ability ratio, management ability ratio, profiting and repaying ability as well as circulating capacity of cash. Then we analyze the data of sample corporation through Fisher, Logistic and clustering analytical method. The forecast result is preferable.From the research result, we draw some conclusions. First, three methods have better forecast effect. However, the forecast result of data mining is better than these of Fisher and Logistic. Second, as far as financial crisis is concerned, the short-term debt rate and net asset profit rate have better forecast ability than circulating capacity of cash. Third, the method of data mining is suitable to long- time financial crisis forecast. Fourth, as a whole, financial data of listed corporation in our country is true.