Research On Credit Risk Evaluation Model Of Technology-based SMEs

With the development of the times,the innovation of science and technology play an increasingly important role.Technological SMEs transform scientific and technological achievements into productivity,and at the same time can promote the growth of the national economy and absorb a large amount of labor.Therefore,technology-based SMEs play an important role in the stability of the national economy,and China also pays more and more attention and attention to the development of technology-based SMEs.However,at present,technology-based SMEs have not established a complete risk control evaluation system,resulting in their credit risk cannot be accurately measured.Under such circumstances,technology-based SMEs are trapped in financing difficulties.The credit risk assessment method,which is easy to be applied in practice,can provide a theoretical reference for the banking industry when making loan decisions for technological SMEs.This article first briefly analyzes the status quo and characteristics of China’s technology-based SMEs,and introduces the concept of credit risk and classic credit evaluation methods.Then combined with the actual situation in China and the comparison of several credit evaluation methods,the Logistic regression model is selected to measure the credit risk of technological SMEs.In terms of indicators,21 indicators were selected as the independent variables of the model from the six aspects of corporate profitability,solvency,operating capacity,development capacity,per share indicators,and research and development capabilities.This article defines ST-listed companies as financially distressed companies and non-ST-listed companies as financially normal companies,and selects 21 ST-listed companies and 63 non-ST-listed SME board-listed companies as samples for 2017-2019 analysis.In order to avoid the influence of index multicollinearity on the model,before establishing the Logistic model,first perform factor analysis on the data to extract the common factor,use the extracted common factor as the independent variable of the model,and then perform binomial logistic regression.Empirical results show that the overall prediction is good.