Page 1 of 1

VAR for Panel Data

Posted: Sat Feb 13, 2010 5:08 pm
by weber
My dataset has a panel structure (multiple years for multiple firms), and I'd like to ask if there are any problems of using "var var1.ls var1 var2 ……."
Because this EVIEWS VAR statement uses an OLS approach, any assumptions are violated if we use it to analyze panel data? Is the estimation still efficient?


Thanks

Weber

Re: VAR for Panel Data

Posted: Tue Feb 16, 2010 10:20 am
by EViews Glenn
While there is no definitive answer to that question, using the built-in var estimator in a panel data set quite likely will violate one or more assumptions.