基于主成分分析和Logistic回归方法的财务预警模型比较——来自我国制造业上市公司的经验证据

Comparison of Financial Early-Warning Model Constructed Based on Principal Component Analysis and Logistic Regression:——Evidence from China's Listed Companies in Manufacturing Industry

在对国内外财务预警研究进行系统研究的基础上,选取我国制造业上市公司作为研究样本,综合运用主成分分析法和Logistic回归法构建制造业上市公司财务预警模型,并对其判别效果进行比较分析。研究结果表明,通过建立预警模型帮助上市公司防范财务风险是一种行之有效的方法,主成分分析模型与Logistic回归模型的判别准确率均较高,但整体而言,主成分分析预警模型的判别效果更好。

In this paper,we first provide a literature review on research concerned with financial early-warning,then use Chinese listed companies in manufacturing industry as a sample to construct financial early-warning model based on principal component analysis and logistic regression,and finally make comparison between the predictive effects of the two models.Result shows that the forecast accuracy of either model is good,while on the whole,financial early-warning model based on principal component could provide a even better predictive effect.