Hi! In a VEC system with four equations (and two lags) and rank three, I get three variables that are significant (according to a Chi2-test of the parameters), and one that isn't. How do I interpret that? Is it an indication that the non-significant variable is long-run exogenous to the others, and that it is correctly set up as the one Granger-causing the others? Or am I totally wrong here? See the attached pic.
Are you interpreting VEC granger causality through chi square? Cannot see the attachment. The t statistics of the VECM table also reveals the causal relationship if it is significant (short term relationship) and the coefficient of the ECT reveals long term relationship
Yes, isn't Chi-squared interesting then? The pic seems to be too small...
Dear Lars,
It is so easy. You have two different parts in your output. Short-run relationship and long-run nexus (that is your ECT vector). Unfortunately I coukd not see your attached file in my phone. But reading a related paper can be useful to you.
Yours,
Ehsan
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis