Inflation, Output, Employment And Monetary Policy Shock

This paper investigates the impulse responses of inflation, output and employment to monetary policy shock using New--Keynsian Dynamic Stochastic General Equilibrium (DSGE) model, Vector Auto Regression (VAR) model and Factor--Augmented Vector Auto Regression (FAVAR) model from the theoretical and empirical perspective Separately, and make comparative analysis on the similarities and differences between the results of two models. Also explore the reason between differences. Using these two models to do analysis mainly bcause they represent two different modeling idea. The DSGE model on the one hand has a strong theoretical microeconomic background; on the other hand the FAVAR model is factor-augmented VAR model to overcome the defects that VAR model can not contain more variables, therefore the FAVAR model depends on data to a large extent, by this way these two models can complement each other.Firstly we develop a DSGE model with prices sticky and wages sticky, and stimulate the impulse responses of inflation, output and employment to monetary policy shock after parameters calibration and model solution; then we provide empirical evidences based on macroeconomic seasonal data of China using VAR model and FAVAR model with different factors respectively, also get the impulse response of these three variables to monetary policy shock, meanwhile to study the role of number of different factors; we observe the results of impulse response use one factor and three factor respectively in FAVAR model, lastly make comparison between empirical impulse obtained from VAR model and FAVAR model with number of different factors and the simulated theoretical impulse. We find that the impulse from FAVAR model with three factors is the most approximative to the theoretical impulse,and can better fit the characteristics of the actual operation of China's economy, which shows that FAVAR model indeed overcome the defects in VAR model, is relatively successful in extracting relevant information from a numbr of macroeconomic variables, thence FAVAR model is more suitable for empirical analysis of DSGE model than VAR model in our country.