我国农村信用社客户信用评级系统研究  

Study on the Client Credit Grading System for Rural Credit Cooperatives in China

作　　者：蒋华[1] 卫功琦 

机构地区：[1]中国矿业大学管理学院,徐州221008 [2]中国银监会徐州银监分局,徐州221009

出　　处：《科技导报》2007年第8期54-60,共7页Science & Technology Review

摘　　要：客户信用评级系统是银行业为规范授信业务、降低贷款风险而采用的内部评级系统,用于对被评级对象履行还款责任的能力及其可信任程度进行客观公正的评价。建立信用评级体系需要考虑的主要因素是违约概率,用于测算违约概率的评价指标设置的科学性与合理性是系统有效的关键。农村信用社与商业银行虽然同样面临信用风险,但由于二者的服务对象不同,评级指标体系应有所差异,商业银行的信用评级方法不能直接适用于农村信用社。本文综合利用因子分析、聚类分析、多元排序选择模型及Logit模型对农村信用社内部评级系统进行了探讨,并对模型的有效性进行了检验,系统总体可行,但仍需进一步修订完善。Client credit grading system is an internal ratings-based system and reduce loan risks. It gives an objective and fair appreciation of a which banks use in order to regulate the credit business client＇s loan-repaying liability and credibility. Default probability is the main factors to be considered. The selected indicators should be scientific and reasonable, which will influence the usefulness of this system. Though rural credit cooperatives face the same credit risk as commercial banks, the current credit rating methods can not be adopted directly because their clients are different and the selected indicators may not be totally adequate. This paper uses factor analysis, cluster analysis, multi-variable discrete choice model, logistic model and practical data to test the client credit grading system proposed. The result shows that the system is applicable, but should be further improved.

关 键 词：信用评级 信用风险 内部评级系统 排序多元离散模型 

分 类 号：F832.2[经济管理—金融学]