The Study On Early-Warning Of Financial Crisis For Listed Companies Based On BP Neural Work

With the rapid development of our country,the extent of damage onlisted companies made by financial risk is deepen increasingly, it affectthe existence and development of company. It is necessary to build aneffective financial warning system in order to guard against financialrisks. Early warning of financial crisis, in the beginning of the financialcrisis on the timely early warning signals to promote corporate vigilanceand take appropriate steps to protect themselves.The paper first reviews and summarizes a large number of domesticand international research literature on the financial crisis and distressprediction, analysis of the advantages and disadvantages of researchmethods at home and abroad, and a clear definition of the financial crisis,warning financial theory. According to our company’s specificcircumstances, to determine the appropriate warning indicator system fordifferent industries. Several traditional financial early-warning modelproposed discrimination method based on BP neural network buildfinancial early warning system of a listed company.50listed companiesin Shanghai and Shenzhen Stock Exchange to be special treatmentbecause of abnormal financial status(ST), select the Shanghai andShenzhen ST50and its paired non-ST as the study sample, the use ofSPSS17.0software single-sample K-S test data, T-test and Mann-WhitneyU test, and filter out the representative indicators to establish financial early-warning model. The results show that the build early-warningmodel has high prediction accuracy. Discovery and delivery of a signal ofcrisis, to urge corporate managers to take effective measures to avoid orreduce the damaging effects of the crisis; To help operators in a timelymanner to identify the root causes that lead to financial crisis, to make upfor the company’s existing management defects, prevent and reduce risk,to avoid the financial crisis from happening again. Build early-warningmodel evaluation, and pointed out the strengths and weaknesses of theearly warning model, the prospect of early warning research.