Page 1 of 1

Johansen Cointegration Test

Posted: Sun Jun 27, 2010 1:11 pm
by youtu
Dear all,

I am doing a Johansen Cointegration Test.

In the output, both "At most 1" and "None" have P-Values bigger than 0.10. The P-Value of "At most 1" is a bit higher than the P-Value of "None". Which hypothesis should I accept ?

Thank you very much

Re: Johansen Cointegration Test

Posted: Sun Jun 27, 2010 1:49 pm
by youtu
I forgot to say I was dealing with two variables.

I learned that in a bivariate VAR(p) model, the Johansen test testes the values of the eigenvalues of a 2*2 matrix. It is in two parts :
- first it testes if the two eigenvalues are zero.
- second it testes if at least one eigenvalue is zero.

I have a problem with E-Views because it is not presented in the same way with "None" and "At most 1". Can somebody explain it to me ?

Thank you very much