Were There Regime Switches in China's Monetary Policy?

Abstract

This paper develops and estimates the quantity based Markov-switching(MS) monetary policy rule for China. The MS regression results show that China's monetary policy rule does exist regime changes, and the best fit is with a version that allows time variation both in disturbance variance and intercept. The volatility and steady state of M2 growth rate under the volatile regime are much larger than that in the moderate regime. I then build an MS-DSGE model allowing for regime switching in monetary shock's variance and steady-state of M2 growth rate. Based on the Bayesian estimation of MS-DSGE model, I find that monetary shock becomes the main driver of the variations in GDP growth rate and inflation rate under the volatile regime,and monetary policy regime shift between moderate regime and volatile regime can lead to the time-varying volatilities of inflation and output in China's bussiness cycle.