F versus t tests for unit roots: a comment

In this note we provide justification for some Monte Carlo results presented by Elder and Kennedy (2001). In particular we show that the severe size distortions observed by Elder and Kennedy are due to the presence of nuisance parameters in the data generation process, but ignored in the test regression. As is shown in a small Monte Carlo exercise, correct size for the statistics is obtained when an adequate test regression is considered.

MoreLess

Year of publication:	2004-04-09
Authors:	Rodrigues, Paulo M. M. ; Tremayne, Andrew
Published in:	Economics Bulletin. - AccessEcon, ISSN 1545-2921. - Vol. 3.2004, 12, p. 1-7
Publisher:	AccessEcon
Subject:	asymptotically invariant tests

More details

Extent:	application/pdf
Type of publication:	Article
Classification:	C4 - Econometric and Statistical Methods: Special Topics ; C2 - Econometric Methods: Single Equation Models
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10011208200