Dynamic stochastic general equilibrium (DSGE) models represent the state of the art in macroeconomic modelling by providing a coherent framework for policy discussion and analysis. DSGE models are basically complex, non-linear systems of equations and are characterized by the following theoretical aspects (Woodford, 2003). Firstly, the models build on explicit micro-foundations involving rational and forward-looking optimising behaviour of individual economic agents. Secondly, they provide a coherent representation of the interactions between households, firms and policy makers in a dynamic micro-founded context. Moreover, the DSGE models have been enriched as to also account for nominal rigidity or price stickiness by the introduction of a new class of models typically referred to as New Keynesian models. They are useful in the analysis of structural changes and of the sources of fluctuations in the economy, or they can be applied to forecast or to predict the effects of policy changes. For these reasons, DSGE has attracted the attention of central banks across the world, some of which have already developed their own model and employ them for monetary policy analysis and forecasting; in particular, the Bank of Canada (ToTEM), the Bank of England (BEQM), the Central Bank of Chile (MAS), the Central Reserve Bank of Peru (MEGA-D), the European Central Bank (NAWM), the Norges Bank (NEMO), Sveriges Riksbank (RAMSES), and the US Federal Reserve (SIGMA). Also, multilateral institutions like the IMF (i.e. GEM, GFM, or GIMF) and the European Commission (QUEST III) have developed their own DSGE models for policy evaluation.  Basic DSGE models reconcile the instances of the New Keynesian paradigm, of the New Classical School and of the real business cycle approach (RBC) to give birth to a new theoretical approach commonly known as the “New Neoclassical Synthesis”. Basic DSGE adopt the real business cycle approach introduced by the seminal paper of Kydland and Prescott (1982), which is based on an impulse-response structure built around optimising agents in a general equilibrium setting. However, DSGE models differ from the RBC benchmark in the ways they explain business cycles. Some realistic assumptions are introduced by a class of DSGE models, as the Keynesian versions of DSGE models and VAR models (see, for example, Gray and Malone (2008) on the source of rigidities and imperfections in the reference markets). However, in order to take advantage from a high degree of flexibility, this class of models often involves strong simplifying assumptions (i.e. homogeneity across households and firms). At present, the modelling of the financial sector is one important challenge of the general DSGE. Finally, the development of estimation procedures of DSGE models (i.e. Bayesian techniques such as Markov Chain Monte Carlo (MCMC) methods and the expectations-maximisation (EM) algorithm increased the usefulness of DSGE models in the context of policy analysis.


Camilo E. Tovar (2008), DSGE models and central Banks, BIS Working Papers, No 258, Monetary and Economic Department, September 2008.
Kydland F. and Prescott E. (1982), Time to Build and Aggregate Fluctuations, Econometrica, 50(6) 1982, pp. 1345-70.
Woodford M. (2003), Interest and Prices: Foundations of a Theory of Monetary Policy, Princeton University Press.

© 2011 ASSONEB