The Study Of SMEs Credit Scoring Based On Logistic Regression Model

As financial market of China becoming more open, the lack of an effective risk management mechanism has become the greatest risk of commercial banks of China. The commercial banks of China have serious challenges how to expand the scale of credit loan while guaranteeing the quality of credit assets. SMEs play an increasingly important role in the national economy development. But the present situation of SMEs are small scale, less standardized financial management, frequent needs of funds and the less loan amount, so Banks are unwilling to lend to SMEs. It is one of important topics of Commercial banks solving SMEs credit risk that developing a rational and effective method of SMEs credit scoring in Commercial banks.SMEs credit scoring which have tremendous potential for development is a financial innovation, But SMEs credit scoring method in our country has never been great concern. This paper summed up the main methods of establish credit scoring, basic principle and basic supposition of Logistic regressing, the multistage classifications of Logistic regressing in foundation of the related domestic literature and foreign literature have already had. Explained Logistic regressing is suitable for SMEs credit scoring because Logistic regressing has the superiority that guaranteeing scoring result objectivity and processing qualitative data. This paper analyzed SMEs scoring indicator system, established specific indicator system for SMEs credit scoring in our country, and established credit scoring model based on Logistic regressing, analyzed cases for SMEs credit data of several commercial banks in different regions. The results show that, in 14 variables of this paper, "Assets-liabilities ratio" and "Accounts receivable turnover ratio" of financial indicators; "Whether is the investor" of the entrepreneur characteristics; enterprise characteristic variables and regional variables affected the model notably. On the model itself, Logistic regressing for the classification accuracy rate of the quality of customer could reach 82.4%. At the same time, analysis obstacles of credit scoring model in China SMEs Loan Application, and proposed the corresponding countermeasures.