企业信用风险判别的主成分Logistic回归分析  

Logistic Regression Based on Principal Component Analysis in Resolving Credit Risk Discrimination of Corporate

作　　者：李文清[1] 许晓娜[2] 

机构地区：[1]河南工程学院数理科学系,河南郑州451191 [2]湘潭大学商学院,湖南湘潭411105

出　　处：《河南工程学院学报（自然科学版）》2009年第2期28-30,共3页Journal of Henan University of Engineering：Natural Science Edition

摘　　要：采用2007年第2季度至2008年第1季度期间中国沪深两地证券交易所上市公司的财务数据作为样本,选取了50家ST公司和50家非ST公司.利用主成分分析法对衡量信用风险的财务指标进行选取和剔除,并得到主成分因子,然后进行logistic回归得到违约率方程.The sample is from China＇s two stock exchanges in Shanghai and Shenzhen from the second quarter of 2007 to the first quarter of 2008, including 50 ST and 50 non - ST enterprises. This paper uses Principal Component Analysis to extract and remove the financial indicators which measure the credit risk of enterprises, and gets the principal component factors, then makes use of Logistic regression, the default rate equation is obtained.

关 键 词：信用风险 主成分 LOGISTIC模型 违约率 

分 类 号：O212[理学—概率论与数理统计]