Outliers - robust ECM cointegration tests based on the trend components

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 suggest a procedure based on signal extraction to bypass this problem. These procedure renders ECM tests with a left tail of distribution under the null that is robust to the presence of additive outliers in the series. The small sample critical values and the empirical power of the test are analyzed by Monte Carlo simulations for several low frequency filters. The proposed empirical methodology is applied to the CPI-based US/Finland real exchange rate. Copyright Springer-Verlag Berlin/Heidelberg 2004