부산대학교 도서관

In recent years, with the rapid development of the Internet era and the surging trend of emerging platforms such as e-commerce, the capital demand of medium, small and micro enterprises(MSMEs) has expanded year by year, and their credit decisions have gradually become a core issue for banks' management. In order to solve the problem in depth in the field, this paper adopts the statistical data related to 123 enterprises with credit records, 302 enterprises without credit records, and the relationship between loan interest rate and customer churn rate in 2019. Based on the strength assessment model and credit rating, the credit risk assessment model is used to determine the annual interest rate and the loan amount.For enterprises with no credit history, the Fisher discriminant method is used to discriminate enterprises with known indicators from those with unknown ratings, and the discriminant credit rating of enterprises with no credit history is obtained. Based on the strength assessment model, this study combines creditworthiness with risk assessment of MSMEs eligible for loans to obtain risk scores and ratings, and eventually determines the annual interest rate and loan amount for each enterprise according to the annual interest rate matrix model.This paper develops a credit decision model for MSMEs with the aim of providing banks with better credit strategies to reduce credit risk and maximize bank profits.