I am trying to calculate the long run effects following the estimation
of a dynamic model (dependent variable Yit), where I have around 20
independent variables (including dummies). To find out the long run
effects, I am dividing all RHS variables by the speed of adjustment
(1-Yit-1). I realised I could calculate these non-linear combinations
using 'nlcom' command in Stata and obtain the long run values of all
variables as well as their p-values, standard errors, etc.
Nevertheless, it would be a great headache for me to make a table out of
these long run effects, given so many regressors (whereas you can
easily format the short run results post estimation using 'outreg' or
'estout' and so forth).
Is anyone aware of any command in Stata to calculate the long run
effects after regression and allow you to make nice tables based upon
them? Or if I have to move from short run to long run manually (by using
'nlcom'), how can I make an output table out of the calculation of these
20 variables?
When you loop over the nlcom results, put them in a Stata matrix, and
use one of the user-written commands to produce an output table of your
liking from the matrix. outtable will do the job if you want LaTeX
tables (and why would one want anything lesser?!?)
