Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Richard Herron <richard.c.herron@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: How to use -suest- with -xtfmb- (test Fama-MacBeth regression coefficients) |
Date | Tue, 8 Nov 2011 09:56:58 -0500 |
I wasn't able to figure out a neat Stata solution, but a dummy variable interaction and adjusting the table output does the trick. Regress y on 1, X, D, and D*X, where D is an indicator variable and X doesn't contain a vector of 1s. The coefficients on D and D*X are the differences between the two models. With the -keep()- option in -esttab- you can focus on the differences by showing only the variables you choose (-esttab- is from the -estout- package on SSC). On Sun, Nov 6, 2011 at 23:23, Richard Herron <richard.c.herron@gmail.com> wrote: > > I am using -xtfmb- from SSC to do Fama-MacBeth regressions (JPE 1973). > I estimate the model in two subsets and would like to use -suest- to > test if the models' coefficients are significantly different. But > -suest- gives the error that the model "was estimated with a > nonstandard vce (Fama-MacBeth)" > > * begin code: > webuse grunfeld, clear > xtfmb invest mvalue kstock if time < 10 > estimates store low > xtfmb invest mvalue kstock if time >= 10 > estimates store high > suest low high > * end code > > The help file for -xtfmb- says that coefficients can be tested if the > -lags()- option isn't used, but here -suest- fails. Is there a way > that I can use -suest- or an equivalent command? Thanks! > * > * For searches and help try: > * http://www.stata.com/help.cgi?search > * http://www.stata.com/support/statalist/faq > * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/