Methods for Computing Marginal Data Densities from the Gibbs Output

Cristina Fuentes-Albero and Leonardo Melosi

Departmental Working Papers from Rutgers University, Department of Economics

Abstract: We introduce two new methods for estimating the Marginal Data Density (MDD) from the Gibbs output, which are based on exploiting the analytical tractability condition. Such a condition requires that some parameter blocks can be analytically integrated out from the conditional posterior densities. Our estimators are applicable to densely parameterized time series models such as VARs or DFMs. An empirical application to six-variate VAR models shows that the bias of a fully computational estimator is sufficiently large to distort the implied model rankings. One estimator is fast enough to make multiple computations of MDDs in densely parameterized models feasible.

Keywords: Marginal likelihood; Gibbs sampler; time series econometrics; Bayesian econometrics; reciprocal importance sampling (search for similar items in EconPapers)
JEL-codes: C11 C15 C16 (search for similar items in EconPapers)
Pages: 20 pages
Date: 2011-10-17
New Economics Papers: this item is included in nep-ecm and nep-ets
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://www.sas.rutgers.edu/virtual/snde/wp/2011-31.pdf (application/pdf)

Related works:
Journal Article: Methods for computing marginal data densities from the Gibbs output (2013)
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:rut:rutres:201131

Access Statistics for this paper

More papers in Departmental Working Papers from Rutgers University, Department of Economics Contact information at EDIRC.
Bibliographic data for series maintained by ().