The application of improved Markov Chain Monte Carlo method in liquidity management of commercial banks

Abstract

This paper proposes an improved Markov Chain Monte Carlo (MCMC) estimation method. First, it introduces improved Markov Chain Monte Carlo estimation method, then uses extended Kalman filter to approximate the nonlinear transfer equation, and ultimately completes the estimation of the redefined asymmetric SV model combining with the forward filtering and backward sampling algorithm. The experimental results indicate that the new algorithm has a faster convergence rate. Finally, this improved MCMC method of asymmetric SV model is applied to the liquidity management of commercial banks. The fitting results not only illustrate its efficiency and accuracy, but also show the empirical evidence of leverage effect.