Financial Performance Prediction Of Chinese Manufacturing Listed Companies Based SVM

The fall of the company’s financial performance has considerable influence on thecorporation. It not only make the enterprise into financial distress,but also bring theinfluence in different level for various stakeholders of the company.Therefore, setting up ascientific financial performance prediction system and timely discover potential risks andtake appropriate measure to prevent or reduce unnecessary loss.The purpose of this paperis to put forward a method of scientific performance prediction which is suitable for listedcompanies in China and widely used, so that it can provide scientific information which isused for making decision for companies and their stakeholders when the company appearthe fall of financial performance.This paper is mainly divided into five parts: the first part mainly introduces thebackground and the method of the research,and summarizes the domestic and foreignscholars about the enterprise financial distress prediction literature;the second part mainlyintroduces the financial performance prediction theory;the third part selects the requiredfinancial indicators of this paper studies, and then pretreats the original index,so as to getthe ultimate indicator to get into the model;the fourth part is the main contribution of thispaper,which constructs for financial performance prediction of Logit regression model andsupport vector machine model;the fifth part summarizes this research work of this paper,and pointed out that the disadvantages of the research and the future research trend.Through the empirical study can get the following conclusions:first,the constructionof enterprise financial performance prediction model has a certain practical value.Although this is only on financial performance trend forecast,not on the financialperformance of specific changes in numerical prediction,but according to financialperformance trend,enterprise and its stakeholders can understand whether the company’sfuture will gain or loss.Second,financial performance prediction accuracy is not high.Either the return result or prediction outcome,the Logit regression model and supportvector machine model prediction accuracy is not high,the rate of miscarriage of justice is in40%or so.The accuracy of financial performance forecast is much lower than the financial distress prediction accuracy. In third,the predict effect of support vector machine model issuperior to the Logit regression model.Both the return result and prediction outcome,theprediction accuracy of support vector machine model is higher than that of the Logitregression model.The conclusion is consistent with the scholars’ study on the financialdistress. In fourth, support vector machine model is more suitable for financialperformance prediction.Empirical research shows that:whether the return result orprediction outcome,the accuracy to recognize the decline companies of support vectormachine model is obviously higher than that in the Logit regression model.