Research On Early-warning Model Of The Credit Risk Of Supply Chain Finance

Global supply chain is the inevitable result of economic globalization, with the continuous development of supply chain, the concept of supply chain finance is born. Commercial Banks as the important carrier of economic activities, developed the business of the supply chain finance the risk evaluation supply chain finance credit risk evaluation is very critical.Through the study of evaluating the credit risk of supply chain finance, it is a beneficial reference for commercial banks to making better credit decisions. It not only increase the bank's profit space but also lower the threshold of the small and medium-sized enterprise financing, a win-win situation for banks and enterprises.This article focus on evaluating the credit risk of supply chain finance of auto industry. Firstly, summarize the domestic and foreign scholars'research achievements of the credit risk of financial supply chain. According to the research results of scholars, define the concept of supply chain finance. Through the research that the characteristics of the credit risk of supply chain finance has strong sudden, travel faster, more and more destructive. And then introduce the evaluation methods of credit risk, introduce the Logistic regression model and deduced the Lasso-logistic model. Then identify credit risk against the storehouse financing model, financing storehouse financing model, and accounts receivable financing model. Setting up evaluation system of the credit risk of supply chain finance according to the established principles of index. Use the data of listed 44 small and medium-sized enterprises of our country's automobile industry as sample, make regression analysis using the Logistic model and Lasso-logistic model. Then test two kinds of regression models, and then compare the results of the two models.Through regression analysis, the overall accuracy of Logistic regression model based on principal component is 81.8%, which has the much higher predicting accuracy of observed risk-free enterprises than observed risk enterprises. The model predicting accuracy of the Lasso-logistic regression is slightly lower than the logistic regression model based on principal components, the predicting accuracy is 79.5%. Through the empirical analysis, enterprise's solvency has negative correlation with the credit risk of supply chain finance of auto industry, obtained that the operating margin, total asset turnover, liquidity ratio, industry status, and partnership of supply chain finance have great influence on credit risk of supply chain finance of auto industry and the influence coefficient respectively -0.0944,-0.0968, -0.0699,-0.0646 and -0.1551.Finally, according to the results, combine with the practical situation of the supply chain's development in China, put forward the corresponding policy recommendations on the government and bank aspects.