Paper number: 95/07

Year: 1995

Paper Category: Discussion Paper


K.M. Abadir*

Department of Economics, Amory Building, Rennes Drive
University of Exeter, Exeter EX4 4RJ, UK.


Testing for cointegration is now widespread in economics. Although the principle is sound, the practice has not always been so. In this note, an attempt is made to reveal flaws in some applied testing procedures. Incomplete nonstationary-null procedures make cointegration seem more likely than it actually is, while incomplete stationary-null procedures reject it too often. Complete definitions of integration and cointegration are given, some of them extending earlier definitions of these concepts.

JEL Classification: C32.

Keywords: (Fractional) Integration, Cointegration, Time Series.
* This note is based on a 1990 mimeographed paper by the author at the American university in Cairo, which was entitled "A warning on cointegration tests". I wish to thank David Hendry and Kaddour Hadri for helpful comments on that paper. I lay sole claim to all remaining errors.