The Empirical Study Of Financial Crisis Early-warning For Listed Companies

With the development of globalization of markets, enterprises are facing more andmore challenges and risks, and venturing into financial crisis is one of the mostimportant risks. Companies venturing into financial crisis is a gradual process, thefinancial position of the enterprise early warning assessment helps companies identifyproblems early, make measures to reduce the risk of falling into financial crisis.In the past, many studies tend to use the model in a simple financial indicatorsestablished to predict the financial crisis, but these financial indicators do not fullyreflect the real situation of enterprises, many financial indicators improved as well asnon-financial indicators can be used to predict the company’s financial crisis. Therefore,this paper based on the traditional financial indicators with prior modeling, then addpart of the financial indicators improved and non-financial indicators to build acomprehensive early-warning model, hoping to provide a reference value for the studyof enterprise financial crisis early warning.This paper reviews the literature on financial crises, discusses the tendency ofscholars of different definitions and indicators selected on the financial crisis earlywarning research and other aspects of the theoretical approach draws some scholars.First, this article set of samples, paired sample selection criteria and indicators,selected in2013by the50ST companies and50non-ST companies; Second, thesecompanies collect the first three years of crisis data, from a financial five aspects levelselected20financial indicators as a research base; Third, make distribution test andtest of significance for13indicators to remove its multicollinearity,9indicators areleft; Fourth, this paper uses software SPSS16.0to make logit regression in backwardstepwise method to establish a model of a financial crisis early warning, and todevelop their criterion. Fifth, then from the financial indicators of improvements andthe introduction of non-financial indicators aspects, introduce11new variables to make a new model on the base of model one. Sixth, compare the two models andvalidate the accuracy rate of these models.The results show that the model of early warning capable for financial indicatorsand non-financial indicators established has improved greatly and the predictionaccuracy rate has risen about7%, as well as3.34%about the accuracy test sample rate,indicating that the early-warning model two are more practical. Secondly, comparingwith ten studies from latest three years, model two’s accuracy rate91.67%are higherthan these studies’ accuracy rate85.12%. The accuracy rate has been improved by6.55%. At last, this paper presents the final policy recommendations and identifiesshortcomings and limitations that exist in the article.