Lag length selection in panel autoregression
Lag length selection in panel autoregression
Model selection by BIC is well known to be inconsistent in the presence of incidental parameters. This paper 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.
BIC, dynamic panel, lag selection, x-differencing, sequential testing, C33
1-35
Han, Chirok
acedb9ff-f19b-4313-aa72-a69cf89b4fc6
Phillips, Peter C.B.
f67573a4-fc30-484c-ad74-4bbc797d7243
Sul, Donggyu
7b86aaed-8c4a-4372-adef-a199c3c3f82b
Han, Chirok
acedb9ff-f19b-4313-aa72-a69cf89b4fc6
Phillips, Peter C.B.
f67573a4-fc30-484c-ad74-4bbc797d7243
Sul, Donggyu
7b86aaed-8c4a-4372-adef-a199c3c3f82b
Abstract
Model selection by BIC is well known to be inconsistent in the presence of incidental parameters. This paper 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.
Text
__userfiles.soton.ac.uk_Library_SLAs_Work_for_ALL's_Work_for_ePrints_Accepted Manuscripts_Han_Lag.pdf
- Accepted Manuscript
More information
e-pub ahead of print date: 5 November 2015
Keywords:
BIC, dynamic panel, lag selection, x-differencing, sequential testing, C33
Organisations:
Social Sciences
Identifiers
Local EPrints ID: 384957
URI: http://eprints.soton.ac.uk/id/eprint/384957
ISSN: 0747-4938
PURE UUID: fd760eab-5c52-4230-a6d7-2854729214ac
Catalogue record
Date deposited: 10 Dec 2015 16:37
Last modified: 14 Mar 2024 22:07
Export record
Altmetrics
Contributors
Author:
Chirok Han
Author:
Donggyu Sul
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
