Dear Experts on Panel Data Econometric and Stata Program,
I will use two way fixed effect panel regression on my article. Because there are id(firm) and time(t) effect on my working.
But on stata instead of using two way fixed effect (xtreg y x i.t, fe),
by eliminating the time effect, i use one way(id/firm) fixed effects model.
For eliminating time effects, i make this Yit-Y.t (for dep. var.) and Xit-X.t (for indep. var.) transformation. And on new model there are not time effects.
And both result is look like each other much.
I have two aim here, First aim is not decreasing degree of freedom of model with dummy for time var. Second aim is that it is easy to work on stata for one way model, and it is easy to calculate that heterescedasdticity and autocorrelation and to product robust estimates and confidence interval.
Please, about this method you can give me your feedback. This is true or not true?

Thank you lots of.
Regards..