Credit Risk Evaluation of SME Loans based on Logistic Regression Model

Nan Li ； Xiaohao Zhao

《Frontiers in Economics and Management》 3卷10期 (2022/10) Pp. 216-222

https://doi.org/10.6981/FEM.202210_3(10).0026

摘要

Credit risk evaluation of corporate loans is important for preventing credit default risk. Based on a national village bank's SME loan data, this paper provides a comprehensive analysis of the influencing factors of loan default and builds a logistic regression model to predict the default rate of corporate loans. The credit rating master scale designed according to the borrower default rate classifies borrowers into 10 different credit classes, which effectively differentiate the risk categories of borrowers and thus provides theoretical support for bank loan approval decision. The empirical results show that years of operation, number of loans, annual sales revenue, execution interest rate, loan term, and lawyer-population ratio are significant in the model; both modeling sample and prediction sample have good default prediction ability.

關鍵字

SME Loans ； Default Rate ； Logistic Regression Model ； Credit Risk Evaluation

參考文獻

J. Carvalho, J. Orrillo, F. R. Gomes da Silva: Probability of default in collateralized credit operations for small business, North American Journal of Economics and Finance, Vol.52 (2020), 101129.

G.He, X.Yang: Empirical Test on Determinants of Credit Default for Small and Micro Enterprises: Evidence from a Regional Branch of One Chinese State-owned Commercial Bank, Journal of Shanghai University of Finance and Economics, Vol.17 (2015) No.6,p.67-79. (In Chinese).

Y. Zhang, S. Chen: Research on the Default Risk of Relationship Lending of Small and Medium-sized Enterprises, Journal of Financial and Economics Theory (2019), No. 2 p.102-112. (In Chinese).

L. Qian: symmetric Information and Risk Mitigation for SMEs, Journal of Financial Research (2015) No.10, p.115-132. (In Chinese).

Y. Zhou: Research on the SME loan risk assessment model based on logistic regression, (MS., Shanghai Jiao Tong university, China 2016), p.14-18. (In Chinese).