Co-integration Test

[alim]

Hi everyone
I would be grateful if someone helps me to overcome the problem because sometimes simple answer gives more pain. I have 5 time series data. One is at level stationary, three is 1st difference stationary and one is 2nd difference stationary. My question is that how do I do Johansen Co-integration Test. Will I take 2nd difference the data of all variables then will do co-integration test or something also. If anyone know this please inform. I will really appreciate you.

Thanking
Abual, Australia

[buratino]

As far as I know all the variables should be integrated of the same order for you to be able to do Johansen...otherwise you simply cannot.

I read somewhere, there is a way to play with your data to make them somehow stationary and then run the test...but it said...it is quite complicated and tricky...

Best wishes,
Buratino.

[chrishayes]

I have the same issue. One of my variables when I tested it using DF and phillips perron tests, it gave me the variable as being I(1) when including a constant but I(2) when including constant and trend. Not sure how to proceed I used the Johansen test, the result of which was 1 cointegrating equation. Is it possible to get that result with one of the variables being I(2) and the other variables I(1)???