Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | "Sabina Kummer" <sabina.noo@postmail.ch> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | st: VAR model and significance of coefficients |
| Date | Wed, 13 Mar 2013 18:10:59 +0100 (CET) |
Dear all, I have a question regarding the interpretation of the coefficients of a VAR/VEC specification. In a VAR/VEC model, the coefficients are all endogenous and dependant from each other (multicollinear). This implies that the t-statistics (Student statistics) of the various coefficients are biased. Thus, I would like to know how to interpret the significance of the coefficients and how to be sure that the choice of the explanatory variables is adequate. Should I consider the F-test for the “global significance” of the model (i.e. the significance of all the coefficients altogether)? Or should I consider the t-stat but only as a "soft indication" about the significance of the coefficients? Thank you very much for your help and your consideration. Best regards, Sabina * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/