Optimal Inference in Regression Models with Nearly Integrated Regressors

   (426 K)

This paper considers the problem of conducting inference on the regression coefficient in a bivariate regression model with a highly persistent regressor. Gaussian power envelopes are obtained for a class of testing procedures satisfying a conditionality restriction. In addition, the paper proposes feasible testing procedures that attain these Gaussian power envelopes whether or not the innovations of the regression model are normally distributed.

This paper is available as PDF (426 K) or via email.

Machine-readable bibliographic record - MARC, RIS, BibTeX