Unit Root Panel tests
Posted: Wed Apr 02, 2014 5:27 am
I'm estimating a 5-equation time-series system and thus retrieving 5 time-series residuals for each equation. I'm interesting in testing for the stationarity of these residuals. I see e-views offers panel unit root test (e.g., IPS, ADF, PP).
Are these tests somehow reserved for an original panel estimation method, or are they just (as I think) a pooled way of testing the joint stationarity of those residuals, irrespective of what econometric tool generated those residuals.
So a p-value of 0.000 implies a unit root is strongly rejected in all residuals, ?
Thanks.
Are these tests somehow reserved for an original panel estimation method, or are they just (as I think) a pooled way of testing the joint stationarity of those residuals, irrespective of what econometric tool generated those residuals.
So a p-value of 0.000 implies a unit root is strongly rejected in all residuals, ?
Thanks.