Page 1 of 1

panel cointegration and GMM

Posted: Wed Jan 14, 2009 4:27 pm
by kelaynak
I am using Eviews 6 in panel data.

I used panel cointegration tests (Fisher and Pedroni). I found cointegration in my data in "individual intercpet" and no "intercept case". There is no cointegration when deterministic trend specifcation was "individual intercept and individual trend". I dont know what this option refers in model equation estimation (GMM panel).

Individual intercept in cointegration means fixed effect for cross section in the panel equation (GMM /Panel option)?
Individual trend in cointegration means fixed effect for period option in panel equation (GMM /panel option)?

Should I select both the "fixed effect" for cross section and "fixed effect" for period option to simulate" individual effect" and "individual trend" mentioned in cointegration test? Or "individual intercept and individual trend" case in cointegration refer to "random effect" in cross section in Panel option of GMM Method?

If not, what are the options I should choose for them in model equations estimation (GMM panel) according to above cointegration result in my data?

Thanks in advance