Evaluating and estimating a DSGE model for the United Kingdom
Published in Bank of England Working Papers, 2010
(with Ozlem Öomen)
We build a small open economy dynamic stochastic general equilibrium model, featuring many types of nominal and real frictions that have become standard in the literature. In recent years it has become possible to estimate such models using Bayesian methods. These exercises typically involve augmenting a stochastically singular model with a number of shocks to structural equations to make estimation feasible, even though the motivation for the choice of these shocks is often unspecified. In an attempt to put this approach on a more formal basis, we estimate the model in two stages. First, we evaluate a calibrated version of the stochastically singular model. Then, we augment the model with structural shocks motivated by the results of the evaluation stage and estimate the resulting model using UK data using a Bayesian approach. Finally, we reassess the adequacy of this augmented and estimated model in reconciling the dynamics of the model with the data. Our findings suggest that the shock processes play a crucial role in helping to match the data.