Study On Credit Risk Assessment Of Commercial Banks Based On BP Neural Network

Since 2014, China has seen a terrible trend of its economy, with a concentrated outbreak of the credit risk of commercial banks, which means the current management of credit risk can’t work well, especially for the evaluation, because of its complexity and the lack of scientific method.The paper constructs the evaluating model of commercial banks’ credit risk, from a perspective of neural network, which can help to enrich the theory of credit risk assessment and expand the scope of the study in this field. Making full use of corporate disclosure information, the establishment of credit risk assessment model for commercial banks to raise profitability and reduce non-performing loan ratio of great practical significance.Based on the reality, the paper first arranges the domestic and foreign research literature. Then it explains the purpose of the establishment of commercial bank credit risk assessment model based on neural networks, and also analyzes the significance, research framework, contents and methods. Secondly, the paper makes a theoretical analysis of the credit risk of commercial banks and also makes theoretical explanation of its meaning, causes, management processes and assessment methods. After that, the paper builds evaluation index system of commercial banks’ credit risk from the practice view, on the basis of which a BP neural network is finally based. To test the model, the paper selects the data of 267 listed companies, and then improves the algorithm by PSO. The results show that the improved model has a high accuracy and robustness, which have high reference value for the credit risk management practice of commercial banks. Finally, the paper makes several suggestions from the algorithm and data of the model.