The Warning Study Of Commercial Banks' Liquidity Risk Based On BP Neural Network

Liquidity is the lifeline of commercial Banks.management of Liquidity is very important to commercial Banks' stable operation.Conducting liquidity management, commercial banks will inevitably regard liquidity risk as a key study object. Only if controling the liquidity risk within a certain ra nge and ensuring high profitability of commercial banks, commercial banks can operate safely, stability and lasting.Since the risk of commercial banks' liquidity can be quantified as a number of specific indicators,In the event of a liquidity crisis these corresponding indicators often reflect something, therefore we can monitor these indicators before the liquidity crisis happened in order to achieve early warning of liquidity crisis.The first chapter of the paper describes the background and significan ce of research in warning of China's commercial banks' liquidity risk and sort out the existing literature on early warning of commercial banks' liquidity risk in the domestic and abroad.The second chapter of the paper describes the related theories about early warning of commercial banks' liquidity risk.Third chapter of the paper introduce the needs and feasibility of early warning system for commercial banks ' liquidity risk.The fourth chapter of the paper designs the early warning system for commercial banks' liquidity risk in the basis of the previous theoretical analysis. At first,the paper completes the overall design the motivation mechanism of the early warning system.Then, on the basis of the overall mechanism the paper conduct the details to specific modules,including constructing the evaluation system of commercial banks' liquidity risk for deciding the input and output modules, building the early warning system on the basis of BP neural network model and deciding the structure of network and the excitating function.The fifth chapter of the paper train and test the BP neural network by using the data of sample banks.The results show that the effect of the training simulate is very good, and the accuracy is 100% which indicating that the early warning system can be used for warning of commercial banks' liquidity risk.The sixed chapter summarizes the conclusions, elaborate the main issues addressed in this study,points out the deficiency of this paper and clears the directions of future research.The warning system of BP neural network established in this paper is complete and have high accuracy of prediction.this system has important practical significance to the management of commercial banks' liquidity risk.