According to Lütkepohl and Krätzig (2004, Applied Time Series Econometrics, p. 89) it is viable to relax the assumption that all variables should be I(1) in levels so that the concept of cointegration can is extended by including any stationary linear combination (providing variables are I(0) in first differences).

Lars
Från: Cottrell, Allin <cottrell@wfu.edu>
Skickat: den 30 oktober 2022 01:04
Till: Gretl list <gretl-users@gretlml.univpm.it>
Ämne: [Gretl-users] Re: small Johansen test - minimum sample size
 
On Sat, Oct 29, 2022 at 3:27 PM Lars Ahnland <lars.ahnland@outlook.com> wrote:
>
> Also, what do you mean by only "one component is stationary". The point is that residuals are stationary with cointegration?

No, what Sven is pointing out is a common "gotcha" with the Johansen
test: one of the series in the llst proposed for testing is I(0). This
in itself does _not_imply that there's any cointegration going on.

Allin Cottrell
_______________________________________________
Gretl-users mailing list -- gretl-users@gretlml.univpm.it
To unsubscribe send an email to gretl-users-leave@gretlml.univpm.it
Website: https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/