Bayesian inference in a time varying cointegration model

Gary Koop, Roberto Leon-Gonzalez, Rodney Strachan

Research output: Contribution to journalArticle

20 Citations (Scopus)

Abstract

There are both theoretical and empirical reasons for believing that the parameters of macroeconomic models may vary over time. However, work with time-varying parameter models has largely involved vector autoregressions (VARs), ignoring cointegration. This is despite the fact that cointegration plays an important role in informing macroeconomists on a range of issues. In this paper, we develop a new time varying parameter model which permits cointegration. We use a specification which allows for the cointegrating space to evolve over time in a manner comparable to the random walk variation used with TVP–VARs. The properties of our approach are investigated before developing a method of posterior simulation. We use our methods in an empirical investigation involving the Fisher effect.
LanguageEnglish
Pages210-220
Number of pages11
JournalJournal of Econometrics
Volume165
Issue number2
DOIs
Publication statusPublished - Dec 2011

Fingerprint

Cointegration
Bayesian inference
Vector Autoregression
Time-varying
Time-varying Parameters
Macroeconomics
Random walk
Vary
Model
Specification
Specifications
Range of data
Bayesian Inference
Simulation
Vector autoregression
Time-varying parameter model

Keywords

  • Bayesian
  • time varying cointegration
  • error correction model
  • reduced rank regression
  • Markov chain Monte Carlo

Cite this

Koop, Gary ; Leon-Gonzalez, Roberto ; Strachan, Rodney. / Bayesian inference in a time varying cointegration model. In: Journal of Econometrics. 2011 ; Vol. 165, No. 2. pp. 210-220.
@article{ca329de8d8f14c998a5b4097e4704486,
title = "Bayesian inference in a time varying cointegration model",
abstract = "There are both theoretical and empirical reasons for believing that the parameters of macroeconomic models may vary over time. However, work with time-varying parameter models has largely involved vector autoregressions (VARs), ignoring cointegration. This is despite the fact that cointegration plays an important role in informing macroeconomists on a range of issues. In this paper, we develop a new time varying parameter model which permits cointegration. We use a specification which allows for the cointegrating space to evolve over time in a manner comparable to the random walk variation used with TVP–VARs. The properties of our approach are investigated before developing a method of posterior simulation. We use our methods in an empirical investigation involving the Fisher effect.",
keywords = "Bayesian, time varying cointegration, error correction model, reduced rank regression, Markov chain Monte Carlo",
author = "Gary Koop and Roberto Leon-Gonzalez and Rodney Strachan",
year = "2011",
month = "12",
doi = "10.1016/j.jeconom.2011.07.007",
language = "English",
volume = "165",
pages = "210--220",
journal = "Journal of Econometrics",
issn = "0304-4076",
number = "2",

}

Bayesian inference in a time varying cointegration model. / Koop, Gary; Leon-Gonzalez, Roberto; Strachan, Rodney.

In: Journal of Econometrics, Vol. 165, No. 2, 12.2011, p. 210-220.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Bayesian inference in a time varying cointegration model

AU - Koop, Gary

AU - Leon-Gonzalez, Roberto

AU - Strachan, Rodney

PY - 2011/12

Y1 - 2011/12

N2 - There are both theoretical and empirical reasons for believing that the parameters of macroeconomic models may vary over time. However, work with time-varying parameter models has largely involved vector autoregressions (VARs), ignoring cointegration. This is despite the fact that cointegration plays an important role in informing macroeconomists on a range of issues. In this paper, we develop a new time varying parameter model which permits cointegration. We use a specification which allows for the cointegrating space to evolve over time in a manner comparable to the random walk variation used with TVP–VARs. The properties of our approach are investigated before developing a method of posterior simulation. We use our methods in an empirical investigation involving the Fisher effect.

AB - There are both theoretical and empirical reasons for believing that the parameters of macroeconomic models may vary over time. However, work with time-varying parameter models has largely involved vector autoregressions (VARs), ignoring cointegration. This is despite the fact that cointegration plays an important role in informing macroeconomists on a range of issues. In this paper, we develop a new time varying parameter model which permits cointegration. We use a specification which allows for the cointegrating space to evolve over time in a manner comparable to the random walk variation used with TVP–VARs. The properties of our approach are investigated before developing a method of posterior simulation. We use our methods in an empirical investigation involving the Fisher effect.

KW - Bayesian

KW - time varying cointegration

KW - error correction model

KW - reduced rank regression

KW - Markov chain Monte Carlo

U2 - 10.1016/j.jeconom.2011.07.007

DO - 10.1016/j.jeconom.2011.07.007

M3 - Article

VL - 165

SP - 210

EP - 220

JO - Journal of Econometrics

T2 - Journal of Econometrics

JF - Journal of Econometrics

SN - 0304-4076

IS - 2

ER -