Optimized Design Of Bank Liquidity Risk Early Warning Program From The Perspective Of New Regulatory Regulations

The liquidity of commercial Banks has always been highly concerned by regulators,bank managers and financial service demanders,because the liquidity of Banks to some extent determines their future lending capacity and ability to meet the demand for cash,thus affecting their long-term development.Only by controlling the bank's liquidity within a certain range and striving to improve its own profitability can the bank maintain stable and efficient operations.The monitoring of bank liquidity level has always been the focus of research,and the establishment of a sound bank liquidity risk early warning program is an important part of timely detection and effective management of liquidity risk.With the implementation of the new regulation on liquidity risk management,the regulation on liquidity risk of commercial Banks has been improved.As commercial Banks,they should also enhance their ability of liquidity risk identification and early warning,and timely find the root causes of liquidity problems.This paper studies the current liquidity status of Chinese Banks and the advantages and disadvantages of their monitoring and identification methods,and finds that it is necessary to construct and optimize the liquidity risk early warning scheme of commercial Banks.From the perspective of new rules on liquidity risk management,the liquidity risk index system is constructed for listed commercial Banks with assets of over 200 billion yuan.Then,factor molecules are carried out to reduce the dimension,and on this basis,factor comprehensive score is calculated to determine the level of liquidity risk in each period.Finally,based on RBF neural network and BP neural network,an early warning model is constructed,and the prediction accuracy and prediction effect of the two are compared,so as to choose a better early warning model of liquidity risk of listed commercial Banks,and on this basis,genetic algorithm is used for optimization,in order to improve the prediction accuracy and prediction effect.By using sample bank data to train and simulate two early-warning models,the prediction accuracy and prediction effect of the models are obtained.The results show that the BP neural network is better than the RBF neural network.After GA is solved and the optimal weights and thresholds are given,the prediction accuracy and effect of the BP neural network are improved.Therefore,for listed banks with assets exceeding 200 billion yuan,GA_BP neural networks can be selected to optimize their liquidity risk early warning programs.