To my understanding the Johansen cointegration test, will test first if the time series have zero cointegrations, if rejected, the Johansen test will test for one cointegration, if rejected, it will test for two cointegrations of more.. If you cannot reject the null hypthesis for the second time in the row it means the times series has one cointegration.. I use matlab and using the Johansen test I get the following result: see pic.

enter image description here

I think the results are weird, because you cannot reject the null hypothesis of zero cointegration, but you can reject the null hypothesis of one cointegrations.. Has someone an explanation?


Such a result could mean that your two time series are already independently stationary. In such a case, you would not need to look for a cointegration relationship.

Recall that Johansen's method supposes that you test a group of time series that are integrated of order 1. You might want to check the stationarity of your 2 time series first.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.