Size matters: covariance matrix estimation under the alternative

The purpose of this paper is to investigate, using Monte Carlo methods, whether Hall's (2000) centred test of overidentifying restrictions for parameters estimated by generalized method of moments (GMM) is more powerful, once the test is size-adjusted, than the standard test introduced by Hansen (1982). The Monte Carlo evidence shows that very little size-adjusted power is gained over the standard uncentred calculation. Copyright Royal Economic Society 2007