[Clyde Schechter]

In your data set, the firm indicator is NPC_FIC, You did enter it in the -reghdfe- command as it appears in -absorb()-. But it does not appear at all in your -reg- command.

[Paris Rira]

You are right.
What about adding id in -reghdfe- as in #14 ? because I assumed that I can substitute id by NPC_FIC. Thats why I entered NPC_FIC in -absorb()-

[Clyde Schechter]

In #14, there is no mention of region or sector. In fact, all of the variables are rather different in that post, and I don't know what any of them actually mean. The general principle applies: if you have panel data with id as the panel identifier, and you have any variable that does not vary within id, that variable cannot be included in the same regression as id. One or the other of them must be removed.

[Paris Rira]

Still, I am not sure whether the model is developed correctly or not.
In the model, there is a firm fixed effect.

[Clyde Schechter]

In the model, there is a firm fixed effect.

Yes, in the equation you show in #34 there is a firm fixed effect. But in the -reg- command there isn't, so the -reg- command does not estimate that model. You would need to add i.id or i.NPC_FIC (whichever is the firm identifying variable in your data set -- you have referred to both in this thread and I am not sure why) to that command's list of regressor variables.

[Paris Rira]

According to #29, I am not eligible to include firm fixed effect, sector and region simultaneously. Am I right?

[Clyde Schechter]

That is correct. However, you can include region#year and sector#year which are the second and third fixed effects in the equation in #34.