摘要
针对商业银行法人客户信用风险甄别问题,采用距离判别法,利用40家上市企业连续两年的财务数据建立了甄别模型,并用该模型对上市公司进行甄别,并将甄别结果与真实情况进行比较,模型对高风险企业的甄别正确率1999-2000年分别为82.5%和90%。与现有的基于神经网络甄别方法比较,此方法效果明显优于后者。
Analyzing the credit risk of enterprise from finance information is an old and widely used practice. Expert and enterpriser tried to choose some financial indicators to set up analysis models. It can analyze and evaluate the financial situation of enterprise and also can analysis the credit risk of enterprise. By means of this method we distinguish good enterprise from risky enterprise. This article uses distance diseriminant method to set up a model to analysis the credit risk of listed company based on the financial data of enterprise. The risky enterprise will be recognized easily, the ratio of success is 82.5 % and 90 %. According to the result, it is better to Neural Network method.
出处
《统计与信息论坛》
CSSCI
2009年第11期83-86,共4页
Journal of Statistics and Information
关键词
距离判别分析
甄别模型
信用风险
distance discriminant method
discrimiant modeling
credit risk
