An Asymmetrical Analysis of Inflation,Inflation Expectations and Monetary Policy in China

In this paper,the authors iteratively use a VAR-based methodology to construct out-of-sample forecasts of inflation and then estimate the past expectations of current inflation(■) based on the sticky-information hypothesis. While the authors construct nonlinearities using logistic smooth transition regression(LSTR) models to describe inflation asymmetry.In the models,lagged ■ is used as a transition variable and it suggests that if inflation expectations exceeds particular threshold values,its effect on the inflation will be reduced.Furthermore,by comparing the effects of interest rate and M_1 on inflation,it is found that if the inflation expectations is less than 2.8%,M_1 can curb inflation, and if the inflation expectations is between 2.8%and 3.9%,both interest rate and M_1 are disinflationary,and if inflation expectations is larger than 3.9%,thus interest rate has significant influence on inflation.