Fisher判别分析模型在上市公司信用风险度量中的应用

以我国上市公司为研究对象,运用统计方法选取有效建模变量,建立了Fisher判别分析预测模型,对上市公司的信用状况进行了实证分析。研究结果显示该模型总的判别准确率为90.38%,与国内已有的研究成果相比有较高的判别准确性,对于证券投资者和分析人员在现有的条件下运用现代信用风险度量模型具有重要的应用价值。

Application of Fisher Discriminant Analysis Model into Credit Risk of Listed Companies

Based om China's listed companies, this paper analyzes empirically the credit status of them, and selects valid modeling variables to build a Fisher discriminant model by means of the statistic method. It indicates that this model's accurate rate is 90.38%, higher than other model's results in China. And it is a useful and important tool for security investors and analysts to apply modern credit risk model under the existing conditions.