信用风险定量指标选取的实证研究  

Empirical Study on the Selection of Quantitative Indicators of Credit Risk

作　　者：岳爱东 Yue Aidong(College of Management and Economics, Tianjin University, Tianjin 300350, P.R.China)

机构地区：[1]天津大学管理与经济学部

出　　处：《山东大学学报（哲学社会科学版）》2019年第3期151-160,共10页Journal of Shandong University(Philosophy and Social Sciences)

摘　　要：通过运用信息熵理论和数据分析技术,计算企业定量财务数据的WOE及Ⅳ值,筛选出信用风险区分能力较强的财务指标,为金融机构量化企业信用风险提供了理论依据。结合企业微观数据并使用Logistic回归方法建立备选信用风险模型,运用ROC、CAP及KS曲线综合分析后,最终确定信用风险量化模型和相关定量分析财务指标。By using the information entropy theory and data analysis technology, the quantitative calculating enterprise financial data of WOE and Ⅳ value were calculated, it has strong ability to distinguish between credit risk of the financial indexes, which provides a theoretical basis for financial institutions to quantify corporate credit risk. Combining enterprise micro data and using logistic regression method to establish an alternative credit risk model, the credit risk quantitative model and related quantitative analysis financial indicators were finally determined after comprehensive analysis of ROC, CAP and KS curves.

关 键 词：信用风险 WOE Ⅳ值 LOGISTIC 

分 类 号：F275[经济管理—企业管理]