| With the development of market economy, people have paid more and more attention to the financial distress early warning research. Owing to the imperfection of the securities market in China, the listed company’s financial position will have a direct relationship with the investors, the creditors and so on. Therefore, stakeholders need do more accurate financial distress early warning research before making investment or lending decisions. At the same time, in the daily operating and managing process, the managers need to timely warn the financial distress to take appropriate measures to prevent the company from going bankrupty.In this paper, the theoretical research and empirical studies are combined. In the theoretical research part, on the basis of previous research, the author defines financial crisis and analyzes the causes of financial crisis, from external factors to internal factors. Then, we present the concept and functions of the financial crisis early warning, and describe the theoretical basis of the financial crisis early warning theory, including the theory of cyclical fluctuations, corporate warning theory, economic early warning theory and agency theory.In empirical research part, based on the sample selection criteria,48listed manufacturing companies, which come from Shanghai and Shenzhen A securities market in2010or2011, have been picked out and divided into a group of financial crisis samples and the other group of financial normal samples. Secondly, in accordance with the principle of index selection, we choose the23financial indicators and9non-financial indicators. And then we ultimately determine the logistic regression model is used in this paper by comparing4kinds of financial crisis early-warning models.Next, remove indicators which the normality tests and non-parametric test results are not significant. At last, we apply factor analysis and logistic regression model into the study and set up mode1(which only includes financial indicators in T-3),mode2(which only includes financial indicators in T-2) and mode3(which includes both financial indicators and non-financial indicators in T-2).Next we compare the three models and draw the conclusion that non-financial indicators is useful to predict the probability of falling into financial crisis for listed companies.Finally,this paper proposes the research conclusions and points out the deficiencies and suggestions in the future. |