Does the DSGE Model Fit the Chinese Economy? A Bayesian and Indirect Inference Approach by

This thesis makes three main contributions to the literature on Dynamic Stochastic General Equilibrium (DSGE) models in Macroeconomics. As no previous studies have studied the Chinese economy from the perspective of DSGE, the first contribution of this thesis is estimating a DSGE model for China through a Bayesian approach using the Chinese quarterly post-economic reform data representing the main macro-economic time series 1978.Q1-2007.Q4. Second, this thesis adopts a new method of evaluating macro-economic models in its evaluation of the estimated DSGE model for China. Rather than the classical methods used to evaluate a macro-economic model such as the Maximum Likelihood method, the method of Indirect Inference is used to test the DSGE model. This method differs from other methods in its adoption of a VAR as the auxiliary model that mimics reality. A hybrid model is adopted to improve the ability of the DSGE model to replicate real world results and compared to the original New Keynesian version of the DSGE model developed by Smets and Wouters. Third, considering the restrictions that the prior distribution imposed on the estimated parameters of the model in the Bayesian estimation, the estimation method of Indirect Inference is used in the last chapter of this thesis and compared with the Bayesian estimation. The results of the Bayesian estimation are in agreement with most of the existing literature on DSGE models. However, the results of Indirect Inference testing suggest that the adopted DSGE model does not closely resemble the real data, with a Hybrid model with 50% weight on the NK part performing significantly better. Indirect Inference estimation produces the same results and provides a better estimation of the model.