Lag length selection in panel autoregression

Chirok Han, Peter C B Phillips, Donggyu Sul

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Model selection by BIC is well known to be inconsistent in the presence of incidental parameters. This article shows that, somewhat surprisingly, even without fixed effects in dynamic panels BIC is inconsistent and overestimates the true lag length with considerable probability. The reason for the inconsistency is explained, and the probability of overestimation is found to be 50% asymptotically. Three alternative consistent lag selection methods are considered. Two of these modify BIC, and the third involves sequential testing. Simulations evaluate the performance of these alternative lag selection methods in finite samples.

Original languageEnglish
Pages (from-to)1-16
Number of pages16
JournalEconometric Reviews
DOIs
Publication statusAccepted/In press - 2016 Mar 1

Fingerprint

Lag
Autoregression
Simulation
Fixed effects
Dynamic panel
Inconsistency
Model selection
Finite sample
Testing

Keywords

  • BIC
  • dynamic panel
  • lag selection
  • sequential testing
  • X-differencing

ASJC Scopus subject areas

  • Economics and Econometrics

Cite this

Lag length selection in panel autoregression. / Han, Chirok; Phillips, Peter C B; Sul, Donggyu.

In: Econometric Reviews, 01.03.2016, p. 1-16.

Research output: Contribution to journalArticle

Han, Chirok ; Phillips, Peter C B ; Sul, Donggyu. / Lag length selection in panel autoregression. In: Econometric Reviews. 2016 ; pp. 1-16.
@article{439d01b03da84fe8a01cb9ccaaac8fa3,
title = "Lag length selection in panel autoregression",
abstract = "Model selection by BIC is well known to be inconsistent in the presence of incidental parameters. This article shows that, somewhat surprisingly, even without fixed effects in dynamic panels BIC is inconsistent and overestimates the true lag length with considerable probability. The reason for the inconsistency is explained, and the probability of overestimation is found to be 50{\%} asymptotically. Three alternative consistent lag selection methods are considered. Two of these modify BIC, and the third involves sequential testing. Simulations evaluate the performance of these alternative lag selection methods in finite samples.",
keywords = "BIC, dynamic panel, lag selection, sequential testing, X-differencing",
author = "Chirok Han and Phillips, {Peter C B} and Donggyu Sul",
year = "2016",
month = "3",
day = "1",
doi = "10.1080/07474938.2015.1114313",
language = "English",
pages = "1--16",
journal = "Econometric Reviews",
issn = "0747-4938",
publisher = "Taylor and Francis Ltd.",

}

TY - JOUR

T1 - Lag length selection in panel autoregression

AU - Han, Chirok

AU - Phillips, Peter C B

AU - Sul, Donggyu

PY - 2016/3/1

Y1 - 2016/3/1

N2 - Model selection by BIC is well known to be inconsistent in the presence of incidental parameters. This article shows that, somewhat surprisingly, even without fixed effects in dynamic panels BIC is inconsistent and overestimates the true lag length with considerable probability. The reason for the inconsistency is explained, and the probability of overestimation is found to be 50% asymptotically. Three alternative consistent lag selection methods are considered. Two of these modify BIC, and the third involves sequential testing. Simulations evaluate the performance of these alternative lag selection methods in finite samples.

AB - Model selection by BIC is well known to be inconsistent in the presence of incidental parameters. This article shows that, somewhat surprisingly, even without fixed effects in dynamic panels BIC is inconsistent and overestimates the true lag length with considerable probability. The reason for the inconsistency is explained, and the probability of overestimation is found to be 50% asymptotically. Three alternative consistent lag selection methods are considered. Two of these modify BIC, and the third involves sequential testing. Simulations evaluate the performance of these alternative lag selection methods in finite samples.

KW - BIC

KW - dynamic panel

KW - lag selection

KW - sequential testing

KW - X-differencing

UR - http://www.scopus.com/inward/record.url?scp=84961217875&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84961217875&partnerID=8YFLogxK

U2 - 10.1080/07474938.2015.1114313

DO - 10.1080/07474938.2015.1114313

M3 - Article

SP - 1

EP - 16

JO - Econometric Reviews

JF - Econometric Reviews

SN - 0747-4938

ER -