Johansen cointegration test set up in Eviews 9

[eviews1]

Hello,

I have a question regarding a set up of the Johansen cointegration test in Eviews. The process I am planning to follow is below. Could you kindly correct me if anything is wrong.

1) I have three REAL price series that I would like to test for cointegration in EViews 9. Please see the file attached JohansenCoint.docx (initially 54 obs.) The ADF test shows that all three series are I(1)

Visually the series do not exhibit any distinctive trend so it is likely that their non-stationarity comes from being Random Walk ones, i.e. they are neither RW with drift nor series with deterministic trend.

2) In the group dialog window of the Johansen cointegration test I made the following settings (here is my main concern) Please see the dialog box snapshot in the file attached. So I chose 1) No intercept or trend in CE or test VAR and lag intervals 0 1.

Based on the Eviews User my choice means that there is no intercept or detemenistic trend in cointegration VAR (my data support that, it means that there is no B*x(t) term in VAR eq. 47.1 in the User guide http://www.eviews.com/help/helpintro.ht ... _Test.html).

The lag choice 0 1 means that in the VECM equation (user guide eq. 47.2) on the RHS of the equation there is PI*y(t-1) (it corresponds to 0 in the lag selection) and one lag of the 1st difference delta_y(t-1).

What does choice 1 1 mean?

Thank you.

eviews1