Credit Decision of Small and Medium Sized Enterprises
Download as PDF
DOI: 10.23977/EMCG2020.006
Author(s)
Haisen Wang, Yixuan Ma, Tiewen Chen
Corresponding Author
Haisen Wang
ABSTRACT
This paper studies the credit strategy of banks to different enterprises, establishes a scoring system to quantitatively analyze the credit risk of enterprises, and establishes a credit strategy decision-making model for enterprises with different credit grades. Then, taking the data as training set, the credit rating and default are classified by random forest algorithm, and the planning model is established to calculate the specific loans. Finally, considering the impact of unexpected factors on the credit strategy in the credit decision-making model, the credit strategy is adjusted by industry. Firstly, we select a series of indicators that affect the credit risk of enterprises, and select the indicators that are completely multicollinearity to establish the disordered multivariate logistic regression model. Considering the probability that an enterprise belongs to different levels, the credit risk of the enterprise is quantified. The descriptive analysis of the purchase and sales invoice is convenient for the subsequent selection of indicators and the formulation of credit strategy. Taking Xinguan epidemic as a typical emergency factor, this paper analyzes its economic impact on different industries and industries.
KEYWORDS
Logistic regression model, multivariate programming model, random forest algorithm