Research On Financial Early Warning Model Combined EVA With Neural Network

Since the accession to the WTO a decade， facing the competitive pattern ofsurvival for the fittest， Chinese enterprises can be described to have been stormy， fromtrade barriers to the financial crisis， all of which are strong tests to Chinese enterprises.With the acceleration of the trend of economic globalization， domestic and foreignmarkets tied closely into an international market， not only the scope of the financialbenefits expand but also corporate financial risk extends to the domestic financial riskswith foreign financial risks. Relative to the domestic financial risks， the uncontrolleddegree of foreign financial risk are higher and the risk is greater. In this case， financialsecurity has become a prominent issue， but how to find the problems in corporatefinancial management timely， predict the existence of financial risk in business， all ofwhich requires companies to strengthen its own financial early warning construction，take the financial risk prevention measures timely， in order to cope with financial risksfrom home or abroad， avoid and reduce financial losses.In view of the scholars has made a lot of work on corporate financial early warningresearch， this article start from the concept of the financial crisis and early warning，drawing on the basis of existing research results， and construct a financial early warningindicator system from four areas of solvency， operating capacity， profitability and pershare indicators， and introduce a new indicators---EVA， building a financial earlywarning indicators based on EVA and a warning indicators based on the traditionalfinancial indicators.And then using empirical research on the listed companies’ financialdata to collect， analyze and model， and compare two different models and twodifferent indicators of system. First， we let Chinese40listed companies as the studysample， based on2007-2009actual financial data of listed companies， and then builda financial indicators system which have28financial indicators and can fully reflectlisted companies’ financial situation， while taking EVA to amend， then do correlationanalysis and factor analysis to the revised financial indicators， ultimately we select15financial indicators into the financial early-warning model such as the current ratio，accounts receivable turnover ratio，EVA and earnings per share cash flow，and then weget the result that the financial early warning model of the improved BP neural networkbased on EVA has a higher accuracy， which get by an empirical comparative analysisfrom the F-score model and the improved BP neural network model based on EVA; Atlast， we construct index system based on the traditional financial indicators，and usethe improved BP neural network to build model， compare the prediction accuracy oftwo financial indicators system， and obtain that the accuracy of the financial indexsystem based on EVA are slightly higher than the traditional financial indicators system.