COINTEGRATION TESTS IN THE PRESENCE OF STRUCTURAL BREAKS

Structural breaks in stationary time series can induce apparent unit r oots in those series. Thus, using recently developed recursive Monte C arlo techniques, this paper investigates the properties of several coi ntegration tests when the marginal process of one of the variables in the cointegrating relationship is stationary with a structural break. The break has little effect on the tests' size. However, tests based o n estimated error correction models generally are more powerful than E ngle and Granger's two-step procedure employing the Dickey-Fuller unit root test. Discrepancies in power arise when the data generation proc ess does not have a common factor.