电子信息板块上市公司财务危机预警与综合评价

【Abstract】wWw.shuoshilunwen.com The paper researches and analyzes in depth the listed companies in the electronic information industry of our country from the standpoint of industry sector of securities business. We build the financial distress prewarning model and the distress level prediction model of the sector, give an analysis and comprehensive evaluation to the normal listed companies in the sector by cluster analysis and principal component analysis, and provide theoretic references which may help managers and stakeholders of the listed companies in the electronic information sector make correct decisions. Theses will promote healthy development of the companies in the electronic information industry of our country.The paper builds financial distress prewarning models of the electronic information sector in the first step. Considering the difference of financial targets in the diverse industries, we select indicators by hypothesis testing to distinguish effectively ST listed companies from normal ones in the electronic information sector. Considering the correlation of indicators, we try to introduce Fisher discriminance based on Mahalanobis distance to the financial distress prewarning, and build the financial distress discriminance models respectively based on threshold method and Mahalanobis distance method.We conclude through the related data checking that these two models are effective in the financial distress prewarning of the electronic information sector, furthermore he prewarning ability ahead of four years. The comparison shows that the financial distress prewarning model based on Mahalanobis distance is superior to the one based on threshold as a whole, which provides a new train of thought for the prewarning research.In order to predict listed companies’financial situation next year which he already involved in financial distress, we build distress level prediction model of the electronic information sector further. We still select indicators by hypothesis testing which can distinguish effectively ST and *ST companies, and build the distress level prediction models respectively based on threshold method and Mahalanobis distance method. We find out through data checking that two models he high back substitution accuracy rate, and the distress level prediction model based on Mahalanobis distance is superior to the one based on the threshold as a whole. The presentation of the distress level prediction model is a flash point in this paper. The paper provides a prediction flow chart by combining the financial distress prewarning model and the distress level prediction model which shows clearly the whole flow of forecasting the next year’s financial situation of one listed company in a particular year.After discussing the financial prewarning, we classify the normal listed companies in the electronic information sector by cluster analysis, and elaborate the company’s feature of each class. Then we abstract six principal components by principal component analysis, figure out each principal component’s score, and build a comprehensive evaluation model with these principal components’scores to give a overall ranking of the listed companies.