Cointegration testing under structural breaks: A robust extended error correction model

It is well known that unit root tests and non-cointegration tests depend on the deterministic elements like: constants, trends, breaks, outliers, segmented trends, etc., that are present under the null hypothesis and maybe also under the alternative hypothesis. This is a serious inconvenience for empirical work since one could arbitrarily influence the result of the unit root test by a convenient selection of the deterministic elements that are included in the regression test. In this paper eve analyze if those problems could be reduced by forming the cointegration tests on extended error correction models. The analysis is done based on Monte Carlo simulation experiments allowing for several structural breaks in the data generating process.