The Study On Financial Crisis Prediction Model With The Introduction Of Non-Financial Indicators

In recent years, listed companies are set in danger in the market for financial crisis,and it is even the major cause for bankrupt. In general, the financial crisis of a company isin a process of developing accumulating, and qualitative change. In the process, somechanges of sensitive financial index can reflex the possibility of the crisis. So by study ofthe changes of financial index, the sign of financial crisis can be found in advance bybuilding an effective enterprise financial early warning model, and take measures to keepthe company going well.Firstly, the present study concluded the research about financial crisis early warninghome and abroad. Then，from the facts of a shares manufacturing listed company, and byanalysis to define financial crisis, and analyze the reason and feature of listed companyfinancial crisis. Secondly, in the recognition and choice of the sample and variable offinancial crisis, the present study divided the financial state of a listed company intofinancial crisis, health and sub health, and made up modeling sample and inspectionsample. Thirdly, selected samples of the use of financial and non-financial indicatorsvariables mean difference test, there are financial and non-financial primaries variableswere significant differences in the correlation between test, remove highly correlatedvariables primaries, and primaries after removing highly correlated variables variablemulticollinearity test; by mean difference test of the chosen sample, and with modelingsample of the year T-2and year T-3financial variables and non-variables, the SPSSsoftware was used to build Logistic regression early warning model of financial indexvariables. And the non-financial index variable Logistic regression early warning modelwas also introduced. Finally, the accuracy of Logistic regression early warning model offinancial index variables and non-variables were tested by samples. The accuracy andstability of2models were compared.Findings showed that as the approaching of financial crisis in a company, the number of financial index variables with significant difference among financial crisis company,healthy company and sub healthy company is rising, showing that the company financialcrisis is a process of accumulating. In different years, financial crisis company, healthycompany and sub healthy company showed significant difference in financial indexvariable, showing that the public information of listed company is reliable. Theintroduction of non-financial index variable into Logistic regression early warning modelcan improve the accuracy of prediction. After the introduction of Logistic regressionmodels in many types of non-financial indicators variables, regardless of financial status ofthe forecast period from near and far, the internal control system has a certain scoreprediction information content.As the barometer of enterprise operation states, financial crisis early warning systemhas raised highly attention abroad. The present study built an effective financial crisis earlywarning model, and helps the company to find the early signs of crisis to avoid it fromhappening.