The Study On Prediction Model Of Financial Distress In Listed Companies In China

The financial distress is not only a common phenomenon of the market economy, but also a global problem. Especially after China's accession to the WTO, China's enterprises are faced with greater risks and challenges, which has greatly increased the possibility of financial difficulties, so that the financial distress has become an issue of great concern. Financial difficulties of enterprises not only have impact on the enterprises themself, but also will give investors, banks, securities companies or even the economy in a region a great loss. Since the 60's of the last century, with the increase in the number of enterprises with financial difficulties and even bankruptcy, many scholars at home and abroad have carred out a large number of related researches, trying to predict the financial situation of enterprises in advance. Therefore, the research on the prediction of financial distress has an important practical significance.In this paper, the first chapter is about the background of subjects, the significance of research and the research framework of papers. In the fierce market competition, enterprises will have the possibility of financial distress at any time.With the improvement of China's stock market and the more and more detailed control on lossed listed companies, once listed companies have abnormalities in the financial situation, the listed companies will be introduced to the special treatment (ST) by Stock Exchange, so the listed companies are faced with greater the risk of the ST. Enterprises in financial difficulty have a step-by-step process, usually the normal state from the financial gradually developed into financial difficulties and even lead to bankruptcy. Therefore, the financial ditress of enterprises not only has portent , but also is predictable. Wheather the financial situation of companies is good or bad is often the focus of much attention. Making financial distress prediction in advance and understanding of the company's financial situation in time will help companies improve their business management strategies to protect the interests of investors and creditors, as well as the securities market regularity and healthy development.Chapter II of this paper describes the definition of financial distress, as well as its related research in detail. Companies in financial difficulty is a result of a continuous process, there is no clear delineation by which a company will be divided into two categories of financial distress and non-financial distress, so the scholars at home and abroad have no unified argument for the definition of financial distress. In this paper, results from previous studies and the current status of China's market, the listed companies of special treatment (ST) due to the the abnomal financial situation is regard as a symbol of falling into financial distress. This not only accords with the actual situation in China, but is as the same with most of the studies to facilitate comparison. And then the research results of domestic and foreign financial distress are introduced, since since the 60's of the last century, the financial distress had been develop widely in Europe and the United States, making a series of fruitful research results. Compared to the financial distress of the international research, scholars in China started rather late, from the last century until the mid-and late nineties, but has achieved some notable results in recent years.Chapter III of this paper is the design part of financial distress prediction model and is the preparatory work to establish a good prediction model. This chapter introduces sample selection, the choice of variables and the tests of variables. According to the data of listed companies is provided by CCER, Shenhu 6 years of financial data for the ST and non-ST listed companies bewteen 1998 and 2007 selected from the A-share market is as a research sample. Different financial indicators reflect the company's financial situation from a different perspective, so this article follows the results of previous studies and focuses on the model taking into account financial indicators of a significant contribution rate. The paper selects the four major categories of reflecting the profitability, solvency, asset management and business development capacity, including 23 financial indicators and then maks all the mean differences test and related test to indicative variables to select 21 indicators variables ultimately, so as to ensure that the models have the forecast ability with high accuracy.Chapter IV of this paper introduces the financial distress model and it is the core of this paper which uses selected financial data and indicators of the previous chapter to establish a dynamic variable prediction model.First of all, a "dynamic" multiple discriminant analysis models and "dynamic" Logistic regression analysis model is established. Different form Most of other studies here, firstly financial data is on the dynamic handling considering the time factor. The average of six years'data of each company's financial indicators is calculated and then the slope of the linear regression of each company's financial indicators to the time is calculated, in this way each indicator derives two indicators, so the 21 indicators of original variables becomes 42 indicators variables. We use 42 indicators to establish "dynamic" multiple discriminant analysis models and "dynamic" Logistic regression analysis model and makes forecasts by the group. Furthermore, in order to better test the merits and demerits of "dynamic" model, the data are divided into two groups, a group is used as model's predictions group, while the other group is used as a model's test group. Here we samples the company accorrding to the parity of company stock code. From the accuracy of discriminant results of the model, we can see that regardless the influence of whether the data is grouped, "dynamic" multiple discriminant analysis models and "dynamic" Logistic regression model has a higher accuracy rate of discrimination and discriminant accuracy rate of "dynamic" Logistic regression model is slightly higher than one of "dynamic" multiple discriminant analysis model. The results show that this "dynamic" model is used to assess the financial position of listed companies with high credibility, providing a certain reference for the forcast of listed companies's financial distress.And then the Logistic Regression Analysis model based on panel data is found. Because of all the data used by the paper are panel data. Panel data are the two-dimensional data of time and space at the same time, taking into account the continuous changes in the data, so have character of time-persistent features to overcome the deficiencies of the isolation of time-series information and cross-section information. Therefore, in Section III of this chapter attempt to use panel data analysis method to make the Logistic regression model based on the panel data. At the same time in order to better test the model's strengths and weaknesses, forecast data are divided into groups and test groups. From the results of discriminant accuracy of the model, the same thing can be seen, regardless of whether the data are grouped, the Logistic regression model based on panel data also has a higher accuracy rate of discrimination, which descries the validity of panel data analysis methods, as well as the practical value of the Logistic regression model based on panel data, in order to provided a new reference for forecast of listed companies'financial distress.Finally, the conclusion part of this paper makes the further summary about the findings and points out the shortcoming of this paper, hoping to build better and more effective financial distress prediction model in the near future.