The Prediction Research Of Corporate Financial Distress Based On The Multiple Discriminant Analysis

With the gradual development and perfection of the reform of the Economic Structure and capital market, there are increasing needs of the prediction and research on financial distress of the company, which requires complete system and method of economic forecast to perform effective synthesis early warning. The most major characteristic this article analyzes is to analyze the present situation of the companies in financial distress of our country. In advance, know about the overall situation of ST Company and complete the synthetical early-warning research on listed companies starting with financial targets link. There have been many demonstration articles dealing with early warning research in our country at present. Through certain processing, this article tries to apply various discriminant methods to analyze and compare in order to make the conclusion more comprehensive and reliable which can act on objective practice.This article is composed of four parts. The first part is the financial early warning research summary. This part analyzes relative theories about the domestic and foreign financial early warning management study, introduces and analyzes financial early -warning demonstration articles published in the authoritative publication at home and abroad.The second part analyzes the present situation of financial distress companies and the definition of financial distress study object.The third part will make use of statistical method to build the basic evaluation model of the early warning and carry on the analysis and the examination to the validity of the model.The fourth part compares the effects from several different forecast methods.It is noted that the 15 of 22 financial ratios are significantly related with financial distress, and the model by Principal Component is more precise in predicting financial distress with lowest error, but each target's information is different, the rate of accuracy is different. In the single variables analysis, the determination effect of the Gross Profit Margin on Sales is better. The multivariable predicting model surpasses the single variable one.