The asymptotic global power comparisons of the GMM overidentifying restrictions tests

In this paper, the asymptotic power comparisons of two versions of GMM overidentifying restrictions tests are conducted globally through the concept of approximate slopes. It is found that the GMM overidentifying restrictions test with the consistent mean deviation variance-covariance matrix estimator is more powerful than the test with the conventional non-mean deviation one. The results shed new light on the findings of Chang (2005) and Hall (2000).

MoreLess

Year of publication:	2007
Authors:	Chang, Sheng-Kai
Published in:	Economics Bulletin. - AccessEcon, ISSN 1545-2921. - Vol. 3.2007, 44, p. 1-6
Publisher:	AccessEcon
Subject:	Approximate Slopes Overidentifying Restrictions Test

More details

Extent:	application/pdf
Type of publication:	Article
Language:	English
Classification:	C1 - Econometric and Statistical Methods: General ; C5 - Econometric Modeling
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10005110755