Research on PCA-entropy model for financial crisis early warning of Chinese listed companies

Listed companies are open, complex nonlinear system. Based on the theory of entropy, the paper select a total of 126 companies that is the 2008-2010 ST companies and non-ST companies as the sample, propose the 30 financial and non-financial indicators with the actual situation of the listed companies financial crisis, use SPSS13.0 software for non parametric text and factor analysis to process a series of data and calculate entropy weights of the sample. The empirical results show that there are a significant difference between ST companies and non-ST companies. Early warning model of ST companies reach 80.95% of accuracy rate, which shows that the model is prediction accuracy and stability. The model has wide application space and can be applied to the prediction of financial crisis of listed companies.