A NOTE ON BOOTSTRAPPING GENERALIZED-METHOD OF MOMENTS ESTIMATORS

Recently, Arcones and Gine (1992, pp. 13-47, in R. LePage & L. Billard [eds.], Exploring the Limits of Bootstrap, New York: Wiley) establish ed that the bootstrap distribution of the M-estimator converges weakly to the limit distribution of the estimator in probability. In contras t, Brown and Newey (1992, Bootstrapping for GMM, Seminar note) discove red that the bootstrap distribution of the GMM overidentification test statistic does not converge weakly to the chi(2) distribution. In thi s paper, it is shown that the bootstrap distribution of the GMM estima tor converges weakly to the limit distribution of the estimator in pro bability. Asymptotic coverage probabilities of the confidence interval s based on the bootstrap percentile method are thus equal to their nom inal coverage probability.