Including "Year" and Macro Variables in SYS-GMM regression

[Fabio biker]

Dear all,
I'm estimating a microeconometric model with bank-level data. A referee flagged an issue claiming: "You cannot use both the "year" variable and macro variables in one regression. This is a serious mistake!!!. Seeking guidance on how to handle this, as I do not understand iwhy. Any tips or resources would be greatly appreciated!
Thanks

[Fabio biker]

The study regards multicountry data

[Sebastian Kripfganz]

This subforum is for experimenting with the forum features. You should post your question in the General forum.

In any case, the referee is generally right. "Macro variables" - i.e., variables that do not vary by bank or country, such as global oil prices - are perfectly collinear with the time dummies. Normally, Stata would automatically omit either those macro variables or some of the time dummies. Whenever time dummies are dropped, you need to carefully assess why!

If those "macro variables" do not vary by bank but vary by country, then there should be no problem.