摘要

China's commercial banks are prevalent in the asymmetric information of "principal-agent" relationship in financial transactions, because of the asymmetric between banks and borrowers, the banks have to confront a problem which is moral hazards. In this paper we proposed a credit rank appraisal index system in view of China's commercial bank's actual operational mechanism, and has constructed a model being on guard against moral hazard based on the neural network technology, this model will help the banks to learn about the future credit of the borrowers before loan providing. Finally, the guard model gave a detailed argument on the feasibility analysis through commercial bank recent years' financial data, which shows the feasibility and validity.