Efficient posterior simulation for cointegrated models with priors on the cointegration space

G.M. Koop, R. Leon-Gonzalez, R. Strachan

Research output: Contribution to journalArticle

21 Citations (Scopus)

Abstract

A message coming out of the recent Bayesian literature on cointegration is that it is important to elicit a prior on the space spanned by the cointegrating vectors (as opposed to a particular identified choice for these vectors). In previous work, such priors have been found to greatly complicate computation. In this paper, we develop algorithms to carry out efficient posterior simulation in cointegration models. In particular, we develop a collapsed Gibbs sampling algorithm which can be used with just-identifed models and demonstrate that it has very large computational advantages relative to existing approaches. For over-identifed models, we develop a parameter-augmented Gibbs sampling algorithm and demonstrate that it also has attractive computational properties.
Original languageEnglish
Pages (from-to)224-242
Number of pages18
JournalEconometric Reviews
Volume29
Issue number2
DOIs
Publication statusPublished - 2 Mar 2010

Fingerprint

Cointegration
Simulation
Gibbs sampling
Relative advantage

Keywords

  • Bayesian
  • collapsed Gibbs sampler
  • error correction model
  • Markov Chain Monte Carlo
  • parameter-augmentation
  • reduced rank regression

Cite this

@article{fbc337823ec44ce9b2e85c461fe544c8,
title = "Efficient posterior simulation for cointegrated models with priors on the cointegration space",
abstract = "A message coming out of the recent Bayesian literature on cointegration is that it is important to elicit a prior on the space spanned by the cointegrating vectors (as opposed to a particular identified choice for these vectors). In previous work, such priors have been found to greatly complicate computation. In this paper, we develop algorithms to carry out efficient posterior simulation in cointegration models. In particular, we develop a collapsed Gibbs sampling algorithm which can be used with just-identifed models and demonstrate that it has very large computational advantages relative to existing approaches. For over-identifed models, we develop a parameter-augmented Gibbs sampling algorithm and demonstrate that it also has attractive computational properties.",
keywords = "Bayesian, collapsed Gibbs sampler, error correction model, Markov Chain Monte Carlo, parameter-augmentation, reduced rank regression",
author = "G.M. Koop and R. Leon-Gonzalez and R. Strachan",
year = "2010",
month = "3",
day = "2",
doi = "10.1080/07474930903382208",
language = "English",
volume = "29",
pages = "224--242",
journal = "Econometric Reviews",
issn = "0747-4938",
number = "2",

}

Efficient posterior simulation for cointegrated models with priors on the cointegration space. / Koop, G.M.; Leon-Gonzalez, R.; Strachan, R.

In: Econometric Reviews, Vol. 29, No. 2, 02.03.2010, p. 224-242.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Efficient posterior simulation for cointegrated models with priors on the cointegration space

AU - Koop, G.M.

AU - Leon-Gonzalez, R.

AU - Strachan, R.

PY - 2010/3/2

Y1 - 2010/3/2

N2 - A message coming out of the recent Bayesian literature on cointegration is that it is important to elicit a prior on the space spanned by the cointegrating vectors (as opposed to a particular identified choice for these vectors). In previous work, such priors have been found to greatly complicate computation. In this paper, we develop algorithms to carry out efficient posterior simulation in cointegration models. In particular, we develop a collapsed Gibbs sampling algorithm which can be used with just-identifed models and demonstrate that it has very large computational advantages relative to existing approaches. For over-identifed models, we develop a parameter-augmented Gibbs sampling algorithm and demonstrate that it also has attractive computational properties.

AB - A message coming out of the recent Bayesian literature on cointegration is that it is important to elicit a prior on the space spanned by the cointegrating vectors (as opposed to a particular identified choice for these vectors). In previous work, such priors have been found to greatly complicate computation. In this paper, we develop algorithms to carry out efficient posterior simulation in cointegration models. In particular, we develop a collapsed Gibbs sampling algorithm which can be used with just-identifed models and demonstrate that it has very large computational advantages relative to existing approaches. For over-identifed models, we develop a parameter-augmented Gibbs sampling algorithm and demonstrate that it also has attractive computational properties.

KW - Bayesian

KW - collapsed Gibbs sampler

KW - error correction model

KW - Markov Chain Monte Carlo

KW - parameter-augmentation

KW - reduced rank regression

UR - http://www.scopus.com/inward/record.url?scp=77950978163&partnerID=8YFLogxK

UR - http://dx.doi.org/10.1080/07474930903382208

U2 - 10.1080/07474930903382208

DO - 10.1080/07474930903382208

M3 - Article

VL - 29

SP - 224

EP - 242

JO - Econometric Reviews

JF - Econometric Reviews

SN - 0747-4938

IS - 2

ER -