|
Activity Number:
|
497
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Thursday, August 10, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #306017 |
|
Title:
|
A Full Information Bayesian Approach to the Evaluation and Estimation of DSGE Models
|
|
Author(s):
|
John Landon-Lane*+
|
|
Companies:
|
Rutgers University
|
|
Address:
|
75 Hamilton Street, New Brunswick, NJ, 08901,
|
|
Keywords:
|
DSGE models ; MCMC ; full information ; segmented markets
|
|
Abstract:
|
In recent years, the estimation of DSGE models using Bayesian methods has become commonplace. One important issue in this literature is that DSGE models are typically degenerate in that there are fewer stochastic terms than the number of endogenous variables in the model. In this paper, I show how all endogenous variables can be used in the calculation of the likelihood function without the need to add shocks to the model. I use a dynamic factor model to approximate a DSGE model and use this to estimate the structural parameters of a segmented markets cash-in-advance monetary DSGE model using MCMC techniques. I compare these estimates to the case of using only a subset of the endogenous variables and to the case of adding nonstructural shocks to the model and find that the impact on the structural estimates is considerable in both a statistical and economic sense.
|
