Data Cloning Estimation and Identification of a Medium-Scale DSGE Model

Abstract

We apply the data cloning method to estimate a medium-scale dynamic stochastic general equilibrium model. The data cloning algorithm is a numerical method that employs replicas of the original sample to approximate the maximum likelihood estimator as the limit of Bayesian simulation-based estimators. We also analyze the identification properties of the model. We measure the individual identification strength of each parameter by observing the posterior volatility of data cloning estimates and access the identification problem globally through the maximum eigenvalue of the posterior data cloning covariance matrix. Our results corroborate existing evidence suggesting that the DSGE model of Smeets and Wouters is only poorly identified. The model displays weak global identification properties, and many of its parameters seem locally ill-identified.

Testing Data Cloning as the Basis of an Estimator for the Stochastic Volatility in Mean Model

“…Tis paper proposes using a diferent approach called "data cloning" [34] to estimate the parameters, utilizing the computational simplicity of MCMC algorithms while also enabling frequentist inferences, such as maximum likelihood estimates and standard errors, to be made. Te method involves applying a Bayesian methodology to a dataset constructed by cloning the original dataset as many times as necessary for the solution to approximate the maximum likelihood estimate [35,36]. Te main advantage of using data cloning over other Bayesian methods is that the inferences are invariant to the choice of the prior distributions, and no likelihood estimation is required.…”

“…We propose using this methodology to estimate the parameters of SV and SVM models as it has been shown to be particularly useful for complex models, as discussed in studies by authors in [34][35][36][37]. Recently, this method has been successfully used to estimate the parameters of other complex fnancial models in [38,39].…”