Our discussion is structured by three concerns - model design, matching the data and operational requirements. The paper begins with a general discussion of the structure of dynamic stochastic general equilibrium (DSGE) models where we investigate issues like (i) the type of restrictions being imposed by DSGE models upon system dynamics, (ii) the implication these models would have for "location parameters", viz. growth rates, and (iii) whether these models can track the long-run movements in variables as well as matching dynamic adjustment. The paper further looks at the types of models that have been constructed in central banks for macro policy analysis. We distinguish four generations of these and detail how the emerging current generation, which are often referred to as DSGE models, differs from the previous generations. The last part of the paper is devoted to a variety of topics involving estimation and evaluation of DSGE models.
Keywords: DSGE model, Bayesian estimation, model evaluation.
Issued: November 2006
Download CNB WP No. 6/2006 (pdf, 352 kB)