Research On Financial Crisis Early-Warning Of Listed Manufacturing Companies In China

At present, there are increasingly concerned about the financial crisis forecast at home and abroad. Not only because it has high academic value, but also has extensive application value.With the rapid development of China's capital market and the increase of the market uncertainties, the company's financial risk becomes significant. Under this background, the financial crisis has become more important to the stakeholders. Therefore, enhancing the forecasting and management of the risks is becoming especially important.There are five parts in this thesis:The first chapter introduces the background and significance of the research .Then the paper puts forward the ideas and methods of this study.The second chapter mainly introduces the research presented both at home and abroad. Through literature study, we compared the results of the financial crisis prediction at home and abroad.The third chapter introduces the theoretical basis for the financial crisis forecasting.The fourth chapter describes the process of modeling and analysis. Firstly, we designed the sample and established the initial prediction index system.secondly, We conducted K-S test on the data distribution. Thirdly, we made the mean difference significance test on the indicators of financial forecasts. By Factor Analysis, we selected indicators of financial projections to eliminate the correlation between them. At last, we constructed the forecasting model of the enterprise's financial crisis by using the method of Logistic Regression Analysis, and got the correct rate of 95.9%. From the research, we find that the three financial indicators including Accounts Payable Turnover Ratio, Interest Coverage Ratio and the Type of Audit Opinion play important roles in predicting the enterprise's financial crisis.In the fifth chapter, we tested the prediction model and made analysis on the misclassification cost of it. Then we did quadratic discriminant, and selected 43 healthy listed manufacturing companies held by 180ETF to do the second test, to test the validity of the model. By misjudgment cost analysis, we can see the model established in this study can reduce the overall sub-error cost to some extent.The sixth chapter is a summary of this article and prospects.