Dear all,

I am applying a panel Vector Autoregression model using the command pvar from Abrigo and Love (2016) in Stata.

There is an option to use "GMM-style" instruments and I have seen a couple of papers where people are doing this.
However, I haven’t seen an explanation why and when I should use the GMM-style instruments and when shouldn’t I use them.

Does anyone have an explanation or can point me to some literature that explains when to use gmm-style instruments?

Kind regards,
Michael

Abrigo, Love (2016), Estimation of panel vector autregression in Stata, Stata Journal, 2016, vol. 16, issue 3, 778-804