Financial Early Warning Of Manufacturing Listed Companies Using Logit Model

The U.S. financial crisis happened in2008, caused a wide range of economicqueasiness, affectted many countries and companies. The spread of the crisis made thecompany’s financial problems were no longer their own business, but also the country’seconomic problems and political troubles. In particular, the further deepening ofeconomic globalization makes the enterprises in a more uncertain external environment.While facing more challenges and opportunities, how to use available information to takea preparation is very important. With the development of information technology and theimprovement of accounting systems, more and more useful financial information issecurable. The research methods of data are updated, which makes the construction ofthe financial early warning model possible.In the process of eary warning model’s construction, this article first outlines therelevant research of financial early warning, analyzes the reason of financial crisis, andthen introduces the financial early warning model. Because the differences betweenindustries and the variance of equity structure may impact on the model, this paper takesChinese manufacturing listed companies’ A-shares as the research object, excluded Bshares, H shares and so on. According to the proportion of1:2, we selected65financialcrises companies and130normal companies as our sample. Considering thecompanies’ operating capacity, solvency, profitability, developing and corporategovernance ability, we use financial indicators and non-financial indicators to establishearly warning logit model. In order to eliminate the impact of industry differences, wefurther investigated the manufacturing differences of secondary category. This articleuse component analysis and factor analysis to eliminate multicollinearity amongindicators, and then analyze the differences between cross-section data logit model andpanel data logit model. We use the curves of the probability distribution of ST companiesand non-ST companies to find the critical value and warning area. The results show that:cross-sectional data logit model is better than panel data logit model. Based on the dataof t-1, the critical value of enterprise is0.6, warning region is [0.6,0.85].In the end of thisarticle, we present the corresponding policy recommendations.