Asymptotic distribution theory for econometric estimation with integrated processes : a guide

The purpose of this paper is, in the absence of a textbook which incorporates in a comprehensive form the increased diversity of applications and methodology of integrated processes, take stock of the most important results in this field, interpreting such results and, also, comparing them to conventional central limit theory for stationary processes. Section 2 develops some preliminary notation and introduces the basic concepts of the appropriate limiting distribution for integrated processes of order one ; in section 3 is applied the previous theory to derive the distributions of several tests for the existence of unit roots ; section 4 examines results in multivariate regression models, including spurious regressions, detrending and cointegrating regressions, as well as issues related to causality tests in a framework of integrated variables. Finally, section 5 extends some of the previous results to higher order integrated and near integrated time series. References