The main goal of this paper is to analyze the behavior of the ECM non-cointegration test when there are additive outliers in the time series under different co-breaking situations. We show that the critical values of the usual ECM test are not robust to the presence of transitory shocks and we...

We show that OLS and GLS are asymptotically equivalent in the linear regression model with AR (p) disturbances and a wide range of trending regressors_ and that OLS based statistical inference is still meaningful after proper adjustment of the test statistics.

This paper is concerned with testing the null hypothesis of no cointegration among I(1) variables when the cointegration residuals are I(d) with 0 < d <1. This possibility is entertained with increasing frequency in many applications, (see e.g. Cheung and Lai 1993 Baillie and Bollerslev 1994 Booth and Tse 1995 or Baillie 1996 for examples. We consider the power of various cointegration tests both for the stationary case d < 5 and for the nonstationary case d > 5.

We show that OLS and GLS are asymptotically equivalent in the linear regression model with AR (p) disturbances and a wide range of trending regressors_ and that OLS based statistical inference is still meaningful after proper adjustment of the test statistics.

This paper is concerned with testing the null hypothesis of no cointegration among I(1) variables when the cointegration residuals are I(d) with 0 < d <1. This possibility is entertained with increasing frequency in many applications, (see e.g. Cheung and Lai 1993 Baillie and Bollerslev 1994 Booth and Tse 1995 or Baillie 1996 for examples. We consider the power of various cointegration tests both for the stationary case d < 5 and for the nonstationary case d > 5.