Research on Computer Aided Risk Evaluation Model through Fuzzy Hierarchical Analysis

Abstract

China does not yet have a set of standardized and easy-to-operate decision-making system for banks to refer to and apply to MSMEs, so it is crucial to establish a set of credit risk assessment strategies that meet the characteristics of MSMEs. The financial risk of MSMEs can be assessed by the combination of both financial and non-financial indicators. The financial indicators can be refined into three indicators: solvency, profitability and growth ability, among which, solvency can be reflected by credit rating and default or not, profitability can be reflected by operating profit margin and return on assets, and growth ability can be reflected by the growth rate of total assets; non-financial indicators can be refined into industry prospect risk, enterprise operation risk, management quality and enterprise's influence on upstream and downstream. influence. Finally, in order to quantitatively analyze the credit risk of the enterprise, the weights of the indicators of the credit risk assessment system should be determined. In this paper, we use fuzzy hierarchical analysis to quantitatively describe the problem. Thus, the comprehensive credit risk rating score of each enterprise is calculated, and it is graded to find the loan amount and operating profit margin under the corresponding level, and finally the bank credit strategy is given. Then, the total sales data and credit rating of several enterprises are used as the input and output layers of the 3-layer neural network for training, respectively. Based on this, the relationship between annual loan rate and customer churn rate is analyzed, and a multi-objective planning model is built with the objective of minimizing customer churn rate and maximizing bank profitability to determine the bank lending strategy through MATLAB.