[R-SIG-Finance] Co-integration question

[Megh Dal]

Hi dears, let say I have few cointegrated variables and those are placed in 2 groups with length n1 & n2. Cointegration may present within and between the groups. As usual information on the cointegration relationship is in beta matrix and the corresponding weights are in alpha matrix.

Now suppose I partition the alpha matrix in 2 part viz alpha1 & alpha2 according to the lengths of the 1st group and 2nd group. Now if I say "alpha2 = 0" then does it signify that: there are no cointgration within the variables of the 2nd group? (however cointegraion may exist within 1st group as well 1st & 2nd group combined).

Thanks for your views.