This paper studies testing for a unit root for large n and T panels in which the cross-sectional units are correlated. To model this cross-sectional correlation, we assume that the data are generated by an unknown number of unobservable common factors. We propose unit root tests in this environment and derive their (Gaussian) asymptotic distribution under the null hypothesis of a unit root and local alternatives. We show that these tests have significant asymptotic power when the model has no incidental trends. However, when there are incidental trends in the model and it is necessary to remove heterogeneous deterministic components, we show that these tests have no power against the same local alternatives. Through Monte Carlo simulations, we provide evidence on the finite sample properties of these new tests.
Bibliographical noteFunding Information:
We would like to thank Cheng Hsiao, an associate editor, and a referee, Francis Diebold, Gloria Gonzalez-Rivera, Soyoung Kirn, Peter Phillips, Frank Schorfheide, and Mark Watson for their comments and discussions. We also appreciate the comments and discussions of seminar participants at UC Riverside, University of Pennsylvania, the Université de Montréal, and the 2002 North American Summer Meeting of the Econometrics Society. Moon gratefully acknowledges financial support from the USC Faculty Development Fund. Perron gratefully acknowledges financial support from FCAR (Fonds pour la Formation de Chercheurs et l'Aide à la recherche) and MITACS (Mathematics of Information Technology and Complex Systems).
All Science Journal Classification (ASJC) codes
- Economics and Econometrics