소상공인 신용평가모형 구축에 관한 연구 - 설문조사 자료를 이용하여 -

In general, it is very difficult for small business owners to obtain the minimum amount of data necessary to secure the reliability of the credit evaluation model because financial transaction records are insufficient. The reliability problem of the small business credit evaluation model can be partially overcome by using relational financing technique which utilizes qualitative information. But it is not easy to achieve. Therefore, the purpose of this paper is to verify the feasibility of constructing a credit evaluation model using self-filling questionnaire data without objective informations. Analysis data are 2,884 respondents to the "2017 Small Business and Small Business Owners Financial Status and Credit Guarantee Support Effectiveness Survey" of the KOREG and the algorithm used is a logistic regression model. As a result of model construction, 10 independent variables are finally selected. The selected variables are monthly rent, growth stage, motivation for start-up, occupation before start-up, difficulties in management, current fund operation situation, main borrowing institution, reasons for loan rejection, average turnover, age of business. The response rate gradually increased from low to high posterior probability of bad, and the sequencing is progressing well. Therefore, it can be concluded that it is possible to use survey data to build a credit rating model for small businesses. In order to establish a credit evaluation model with high reliability and accuracy, the answer of the borrower's questionnaire should be faithful.