Performance of model selection criteria in Bayesian Threshold VAR (TVAR) models

Yongjae Kwon, Hamparsum Bozdogan, Halima Bensmail

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

This article presents a new Bayesian modeling and information-theoretic model selection criteria for threshold vector autoregressive (TVAR) models. The analytical framework of Bayesian modeling for threshold VAR models are developed. Markov Chain Monte Carlo (MCMC) simulation and importance/rejection sampling methods are used to estimate the parameters of the model and to obtain posterior samples. We propose reliable modeling procedures using Bayes factor, and the information-theoretic model selection criteria such as, Akaike's (1973) Information Criterion (AIC), Schwarz (1978) Bayesian Criterion (SBC), Information Complexity (ICOMP) Criterion of Bozdogan (1990, 1994, 2000), Extended Consistent (AIC) with Fisher Information (CAICFE), and the new Bayesian Model Selection (BMS) Criterion of Bozdogan and Ueno (2000). We study the performance of these criteria under different design of the simulation protocol with varying sample sizes in TVAR models. Our results show that these criteria perform well in small sample as well as large samples to avoid heavy computational burden in conventional procedures.

Original languageEnglish
Pages (from-to)83-101
Number of pages19
JournalEconometric Reviews
Volume28
Issue number1-3
DOIs
Publication statusPublished - 1 Jan 2009
Externally publishedYes

Fingerprint

Model selection criteria
VAR model
Bayesian modeling
Burden
Bayes factor
Markov chain Monte Carlo
Sample size
Importance sampling
Monte Carlo simulation
Vector autoregressive model
Simulation
Bayesian model
Small sample
Sampling methods
Fisher information
Bayesian information criterion
Akaike information criterion
Modeling

Keywords

  • Bayesian modeling
  • Information-theoretic model selection criteria and model selection
  • Threshold autoregressive models

ASJC Scopus subject areas

  • Economics and Econometrics

Cite this

Performance of model selection criteria in Bayesian Threshold VAR (TVAR) models. / Kwon, Yongjae; Bozdogan, Hamparsum; Bensmail, Halima.

In: Econometric Reviews, Vol. 28, No. 1-3, 01.01.2009, p. 83-101.

Research output: Contribution to journalArticle

Kwon, Yongjae ; Bozdogan, Hamparsum ; Bensmail, Halima. / Performance of model selection criteria in Bayesian Threshold VAR (TVAR) models. In: Econometric Reviews. 2009 ; Vol. 28, No. 1-3. pp. 83-101.
@article{d3e42cf103f8408492fd83773db8d429,
title = "Performance of model selection criteria in Bayesian Threshold VAR (TVAR) models",
abstract = "This article presents a new Bayesian modeling and information-theoretic model selection criteria for threshold vector autoregressive (TVAR) models. The analytical framework of Bayesian modeling for threshold VAR models are developed. Markov Chain Monte Carlo (MCMC) simulation and importance/rejection sampling methods are used to estimate the parameters of the model and to obtain posterior samples. We propose reliable modeling procedures using Bayes factor, and the information-theoretic model selection criteria such as, Akaike's (1973) Information Criterion (AIC), Schwarz (1978) Bayesian Criterion (SBC), Information Complexity (ICOMP) Criterion of Bozdogan (1990, 1994, 2000), Extended Consistent (AIC) with Fisher Information (CAICFE), and the new Bayesian Model Selection (BMS) Criterion of Bozdogan and Ueno (2000). We study the performance of these criteria under different design of the simulation protocol with varying sample sizes in TVAR models. Our results show that these criteria perform well in small sample as well as large samples to avoid heavy computational burden in conventional procedures.",
keywords = "Bayesian modeling, Information-theoretic model selection criteria and model selection, Threshold autoregressive models",
author = "Yongjae Kwon and Hamparsum Bozdogan and Halima Bensmail",
year = "2009",
month = "1",
day = "1",
doi = "10.1080/07474930802387894",
language = "English",
volume = "28",
pages = "83--101",
journal = "Econometric Reviews",
issn = "0747-4938",
publisher = "Taylor and Francis Ltd.",
number = "1-3",

}

TY - JOUR

T1 - Performance of model selection criteria in Bayesian Threshold VAR (TVAR) models

AU - Kwon, Yongjae

AU - Bozdogan, Hamparsum

AU - Bensmail, Halima

PY - 2009/1/1

Y1 - 2009/1/1

N2 - This article presents a new Bayesian modeling and information-theoretic model selection criteria for threshold vector autoregressive (TVAR) models. The analytical framework of Bayesian modeling for threshold VAR models are developed. Markov Chain Monte Carlo (MCMC) simulation and importance/rejection sampling methods are used to estimate the parameters of the model and to obtain posterior samples. We propose reliable modeling procedures using Bayes factor, and the information-theoretic model selection criteria such as, Akaike's (1973) Information Criterion (AIC), Schwarz (1978) Bayesian Criterion (SBC), Information Complexity (ICOMP) Criterion of Bozdogan (1990, 1994, 2000), Extended Consistent (AIC) with Fisher Information (CAICFE), and the new Bayesian Model Selection (BMS) Criterion of Bozdogan and Ueno (2000). We study the performance of these criteria under different design of the simulation protocol with varying sample sizes in TVAR models. Our results show that these criteria perform well in small sample as well as large samples to avoid heavy computational burden in conventional procedures.

AB - This article presents a new Bayesian modeling and information-theoretic model selection criteria for threshold vector autoregressive (TVAR) models. The analytical framework of Bayesian modeling for threshold VAR models are developed. Markov Chain Monte Carlo (MCMC) simulation and importance/rejection sampling methods are used to estimate the parameters of the model and to obtain posterior samples. We propose reliable modeling procedures using Bayes factor, and the information-theoretic model selection criteria such as, Akaike's (1973) Information Criterion (AIC), Schwarz (1978) Bayesian Criterion (SBC), Information Complexity (ICOMP) Criterion of Bozdogan (1990, 1994, 2000), Extended Consistent (AIC) with Fisher Information (CAICFE), and the new Bayesian Model Selection (BMS) Criterion of Bozdogan and Ueno (2000). We study the performance of these criteria under different design of the simulation protocol with varying sample sizes in TVAR models. Our results show that these criteria perform well in small sample as well as large samples to avoid heavy computational burden in conventional procedures.

KW - Bayesian modeling

KW - Information-theoretic model selection criteria and model selection

KW - Threshold autoregressive models

UR - http://www.scopus.com/inward/record.url?scp=58149115213&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=58149115213&partnerID=8YFLogxK

U2 - 10.1080/07474930802387894

DO - 10.1080/07474930802387894

M3 - Article

VL - 28

SP - 83

EP - 101

JO - Econometric Reviews

JF - Econometric Reviews

SN - 0747-4938

IS - 1-3

ER -