The Research On Enterprise Credit Assessment Under Financial Early Warning

Today, the rapid development of global economy is formed by the integration of worldeconomy, however, the enterprise credit assessment is particularly important. In such a developedfinancial market and capital market, if the enterprise can give an objective and fair Standard， creditassessment can play an important role, these problems have aroused people’s attention. Therefore, areasonable selection of the index as the support vector machine for credit assessment, can be onguard effectively in the process of business operation risk, it also can reduce transaction costs. Thestudy is focused on the early warning model, find out the influence of credit assessment of keyfinancial indicators, and these indices are constantly tracking control, finally realizes theenterprise financial risk, financial risk is the enterprises operating in the process of facing one of themost important risks, only to control financial risk and achieve the enterprise steadily, eventuallywill not affect the evaluation mechanism on enterprise credit evaluation.In recent years, domestic and foreign scholars have been devoted to financial early warningmodel of research and development, they generally use statistical analysis method, enterprise ofsome key financial ratios, such as the liquidity ratio, quick ratio and assets turnover rate asexplanatory variables model calculation and detection. But because of accounting fraud of financialindex, calculated from the ratio does not truly reflect the management state of the enterprise,therefore, use statistical analysis to derive prediction effect will be reduced, so that the reliabilityis not strong, it influences the decision maker’s business strategy. In this case, use the application ofdata mining technique to build the financial early warning model of emerge come out as the timesrequire. Because the general statistical models and most of them have a lot of conditions, only tomeet the prerequisites to model calculations, but in market, enterprise data is difficult to meet therequirements of a variety of conditions, statistical models, which will affect the prediction accuracyand effectiveness. Data mining methods are not limited by conditions, at this point, it can make upthe defects existing in the statistical analysis method. In this case, some need to expand the scale ofinvestment managers need a kind of new financial crisis early warning model. This urgent makes usfrom another new point of view to study enterprise’s financial crisis early warning system. Thispaper introduct a new financial indexes, EVA, namely the company net operating profit less totalcost of capital after the net, this index is relatively accurately reflect in a certain period of enterprises to create value, but also can be reduced because the accounting fraud, the reportinformation and enterprise’s real contrary to fact. Therefore, introducing new variable EVA, coupledwith the data mining technology in the financial crisis early warning model, can improve theforecasting effect of the model, more accurately predicted the financial crisis, enterprise managerscan be adjusted in a timely manner cause financial crisis index, it can reduce the enterprises creditrisk, and then in the credit evaluation plays an important role.In this paper, the new financial index--EVA will be introduced, the decision tree in datamining technique will be used in financial early-warning method, the predicted results and thestatistical analysis method--Logistic regression prediction results are compared, finally, theseindexes also affect the enterprise credit. If these indicators can be controled for real-time, it canreduce the enterprise risk.