信息不对称下银行对中小微企业的最优信贷策略研究--基于Logistic回归的违约率测算模型  

Research on Banks' Optimal Credit Strategy for MSMEs UnderInformation Asymmetry--Default Rate Measurement Model Based on Logistic Regression

作　　者：孙雨忱 Sun Yuchen(School of Statistics,Renmin University of China,Beijing 100872,China)

机构地区：[1]中国人民大学统计学院,北京100872

出　　处：《金融发展研究》2021年第6期78-84,共7页Journal Of Financial Development Research

摘　　要：通常认为,中小微企业融资难的主要原因在于抵押担保能力不足、银行贷前及贷后管理成本高。但究其深层次根源,则是在信息不对称情况下,银行难以对中小微企业的违约率进行可靠测算,从而难以估量这部分授信的资本占用,加大了银行自身资本充足率管理的难度。本文利用银行较易获取的企业交易信息(发票信息)、信用评级、信贷记录等有限信息,通过构建Logistic回归的违约率测算模型,测算出不同信用水平下中小微企业的预期违约率。在此基础上,以银行利润最大化为目标,构建银行最优信贷策略的非线性规划模型,并以此模型指导银行对潜在中小微企业客户开展授信活动。It is generally believed that the main reasons for the financing difficulties of micro,small and mediumsized enterprises are insufficient mortgage guarantee capabilities and high pre-and post-loan management costs of banks. However,the deep root cause is that under the situation of information asymmetry,it is difficult for the banks to reliably measure the default rate of small,medium and micro enterprises,which makes it hard to estimate the capital occupation of this part of the credit,increasing the difficulty of the bank’s own capital adequacy ratio management.This paper uses the limited information such as corporate transaction information(invoice information), credit ratings,and credit records that are easily available to banks,and builds a logistic regression model of default rate calculation to calculate the expected default rate of small,medium and micro enterprises at different credit levels. On this basis,with the goal of maximizing bank profits,a non-linear programming model of the bank’s optimal credit strategy is constructed,and this model is used to guide the bank to carry out credit activities for potential small,medium and micro enterprise customers.

关 键 词：中小微企业 违约率 信贷 LOGISTIC回归 

分 类 号：F830.5[经济管理—金融学]