Abstract: The authors of this paper first use the convergence cross mapping causality test(CCM) model to judge the causal link between nominal interest rate and different output gaps, and then use the time-varying parametric regression model to estimate the parameters of Taylor rule constructed by different output gaps in the framework of forward-looking and backward Taylor rules. The results show that there is a significant causal relationship between the nominal interest rate and the pseudo output gap constructed based on the real output growth rate and the target growth rate, while not between the nominal interest rate and the output gap calculated based on the unobservable component model and the production function method. The further research shows that the central bank has a policy preference of adjusting nominal interest rate for inflation gap and pseudo output gap whether based on rational expectations or adaptive expectations. At the same time, the central bank's control mode shows significant time-varying characteristics.

Keywords: monetary policy; Taylor rule; output gap; CCM model; time-varying