By Martin Kliem and Harald Uhlig
http://d.repec.org/n?u=RePEc:zbw:bubdps:372013&r=dge
This paper presents a novel Bayesian method for estimating dynamic stochastic general equilibrium (DSGE) models subject to a constrained posterior distribution of the implied Sharpe ratio. We apply our methodology to a DSGE model with habit formation in consumption and leisure, using an estimate of the Sharpe ratio to construct the constraint. We show that the constrained estimation produces a quantitative model with both reasonable asset-pricing as well as business-cycle implications.
To continue with the theme about models addressing both business cycles and asset pricing: The point of this paper is quite simple. If an estimated model cannot satisfy both business cycle and asset pricing aspects, one can try to force it that way with Bayesian estimation. And as long as the model can work in theory, it should have a good shot at working with estimated parameters. This also means that we do not yet have a model that, at least in this respect, fits the data naturally enough to not need being guided by tight priors.