Credit Assessment Of Small And Medium-sized Enterprises Based On Logistic Regression

From the reform and opening up to today,China's economy has undergone a great transformation after more than 40 years of reform and development.But China's smes(small and medium-sized enterprises)are growing,reflects that China's smes can well according to the market changes,make adjustments to adapt.At present,smes are a greatly part of the national economy,and their positive role is becoming more and more obvious,more and more people take seriously.But,China's smes are generally faced with such challenges as high cost,increasing downward pressure on market demand,and insufficient orders caused by the lack of competitiveness of products Vice minister of the ministry of industry and information technology Wang Jiangping analysis,on the one hand,smes independent innovation is not enough,innovation cost is high,long cycle of risk,facing a lack of capital,talent,information and other difficulties.On the other hand,the professional ability and level of smes are not high.In China,there are still a large number of smes concentrated in traditional industries and at the middle and low end of the value chain.There are problems such as lack of core technologies,homogeneous development and insufficient focus on main industries.Data show that in the process of smes' transformation and upgrading,capital is the main problem,and credit risk is still the main risk in the loan process.Therefore,the establishment of an objective,fair,reasonable and effective risk assessment system is essential for the development of the sme market,which is an important way to improve the information asymmetry and reduce costs between smes,Banks and other lending institutions.It is of great significance and value to individual investors,smes themselves,commercial Banks and other lending institutions.In terms of credit evaluation system,credit evaluation system in our country started late compared with foreign countries,but the development is also relatively mature,many scholars of our country use the international mainstream credit assessment model for reference,take the essence of the actual situation of our country to optimize.A sound credit evaluation system can enable lenders to extend credit to smes more often,and ultimately achieve a "win-win" situationThis paper first explains the background and significance of the article,and explains that the research of this paper is of practical significance.It introduces the relevant literature on credit risk assessment at home and abroad,and explains the key points and difficulties in the research process,as well as the main problems and ideas of the research.Then it briefly summarizes the current situation of smes and credit risk in China,describes the definition,characteristics,role and development trend of smes,the definition,causes and characteristics of enterprise credit risk.Then,three representative credit evaluation theories are described and their advantages and disadvantages are analyzed,which are expert evaluation method,credit score method and modern credit risk measurement model.After comparing some models,this article finally selects the logistic regression model,the empirical mainly uses the financial index data and non-financial index data of the sme board in the whole year of 2019,whether the smes is selected as ST company is the criterion to determine whether the smes is in breach of contract.In terms of evaluation indicators,this paper choose eight indicators to screening,including:solvency indicators,operating ability index,profitability index,development capacity index,level of risk index,social responsibility index,ownership concentration index,executive incentive index.Finally,the tested factors were selected to extract the principal components for modeling.The final results show that the accuracy of the regression model with non-financial indicators is better than that with pure financial indicators.Besides,the logistic regression model established in this paper is based on non-industry division,and joint consideration of financial indicators and non-financial indicators,with a high degree of fitting.