Comprehensive Evaluation Of The Efficiency Of Commercial Banks Based On Neural Networks

As a core of the financial system, the efficiency of banking will directly affects the development of national economy. Thus, hotspot of theoretical research, abstractly and realistically, essential analysis based on efficiency appraisal of commercial banks has become the important aspect of policy formulation to ensure effective development of the financial market.This article is made up of five parts. First chapter: foreword. The second chapter is the construction of the efficiency index system. The third chapter is the appraisal model based on BP neutral network and its practical calculation. The fourth chapter is the analysis of out comings with corresponding policy advises and the last one, conclusion.On the ground of the widely chosen efficiency indexes of commercial banks, we set up a new index system after principal component analysis on original data, and, we also set up an efficiency appraisal model based on BP neutral network. There are also pertinent advices based on practical calculation and analysis.Appraisal result shows that the efficiency of the joint-stock commercial banks is better than that of the non-joint-stock ones, and four state-owned commercial banks did badly in their efficiency appraisal. Among the joint-stock commercial banks, Shanghai Pudong Development Bank has the highest efficiency and China Communication bank finishes last. Of the state-owned banks, China Bank finishes the first and China Agriculture Bank gets last. Analysis on appraisal results indicates a remarkable diseconomy on scale in china commercial banks, that is to say the increase of assets will not enhance the efficiency of commercial banks. This article suggests a prudential increase of assets and to develop an efficient management system. Also, commercial banks should definite the property right to enhance their efficiency by stock reformation.The first innovation of this paper lies in the new index system which keeps 92.876% of the original information and avoids the influence of the coefficient between different indexes on later appraise. Secondly, new index system gets even more information from the widely chosen original indexes. Thirdly, we avoid the influence comes from the lack of loan quality through taking into account of those indexes of loan quality. Fourthly, we successfully keep the information of diversity and inner influence between different indexes through promoted standardizingmethod of formal data-average method. Fifthly, we find the mainly influencingfactors through direct oblimin method and, based on which, we put forward correlative measures and advises to improve the efficiency of China's Commercial Banks. Sixthly, we set up a appraisal model benefits from the self-studying and self-organizing BP neutral network which has get rid of the influence of the factitious subjective factors and fuzzy stochastic ones. Lastly, we proved diseconomy on scale through co efficiency analysis and significance analysis.