I'm dealing with the Johansen cointegration test before running an ECM model.

I cannot figure out which type of test execute: with/without constant, with/without trend, ...

How can I choose the correct version? Do I have to do it a priori?

And if the test tells variables are cointegrated with trend, what does it mean? Do I have to estimate an ECM with trend?

I've read some post on this topic, but no one answers my quesions. Any help would be much appreciated!

  • $\begingroup$ does this answer your question? stats.stackexchange.com/questions/180854/… $\endgroup$
    – utobi
    Commented Oct 11, 2022 at 12:33
  • $\begingroup$ @utobi not really... I need to understand what "trend cointegration" means and if it implies the necessity of a trend in my ECM model. And is there a method to decide a priori if I have to run the cointegration test with or without trend? $\endgroup$ Commented Oct 11, 2022 at 12:52


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.