Study On Chinese Listed Companies' Financial Crisis Early-warning Models

In recent years, with the further deepening of China's market economy, intense competition bring companies not only opportunities but more challenges. In the market which is full of challenging competition, inevitably, some companies fall into financial difficulties, and even progress to show a new crisis. In 2008 the global financial crisis, China's economy has suffered a number of micro-entities, negative impact on enterprises are faced with the emerging financial risks. How to timely and accurately monitor and forecast financial risk has become a problem that modern enterprise financial management must be resolved, therefore, build an efficient, sensitive and practical financial risk early warning system is very necessary.This paper is based on the study of theories and methods of the financial crisis early warning at home and abroad. This paper use empirical research methods to construct suitable financial early warning models for listed companies in China according to their financial data.In this paper, Enterprise Financial Crisis warning is the research object, using quantitative analysis and qualitative analysis combined standardized research and empirical studies of combining research methods. First of all, this paper reviewed and summarized the financial early-warning model of classical literature at home and abroad. The second chapter defines a clear theoretical basis of this article and the concept of the financial crisis. Then, the third chapter builds a set of 20 indicators in five areas of the financial crisis early-warning indicator system. Select 51 crisis companies which were first special treat in the industry with 51 paired samples of healthy companies as modeling from the year 2003 to the year 2007 in China. The forth chapter use integrated statistical methods such as: single-sample K-S tests, T tests, Mann-Whitney U test, factor analysis, correlation analysis, and selected 12 indicators as a model for independent variables; in order to model a sample of financial crises based on data from the previous two years, using statistical software SPSS16.0 to establish an early-warning categoryⅡlogistic regression model of listed companies in China and a discriminated Fisher linear model. Empirical results of the study showed that the financial crisis two years ago, two types of models of modeling the overall return on behalf of the samples were 86.3% accuracy rate, but the distribution of the miscarriage of justice rate is different.Finally, the paper selected crisis of 2003-2007,21 companies and 21 samples tested in healthy companies, using logistic regression models and Fisher discriminated analysis of early warning models to test, test results showed that the overall logistic regression model accuracy rate was 92.9%, Fisher discriminated analysis model's overall prediction accuracy rate of 88.1%, while selected 5 first special treat listed companies and 5 matched companies in 2008 then 90 manufacturing industry listed companies in 2003-2007 into the test, the results prove that Logistic regression model forecasts more accurate than Fisher discriminated model. Study Conclusion:The established model in this paper are feasible and practical, applicable industry-wide in preventing and resolving crises, play a role in improving enterprise management level of crisis early warning.