Currently I'm working on the replication (variation in time) of Charles W. L. Hill, & Phan, P. (1991) for an assignment.
In their analytical strategy they use a generalized-least-squares cross-sectional "heteroskedastic" and "time-wise" autoregressive model to correct for serial-correlation.
My question is, how to perform this test in STATA?

These are the options I have found, but I'm uncertain which one to use.
//set panel data
xtset gvkey fyear, yearly

xtgls dep.v ind.v, panels(correlated)
which specifies a heteroskedastic error structure with cross-sectional correlation.
However this one doesn't work as my panel data isn't balanced due to the fact that not all firms are present in the S&P1500 for consecutive years.

or:
xtreg dep.v ind.v, re


I'm new to STATA so any good advice would be helpful.
Thank you!


v. Stiphout


reference:
Charles W. L. Hill, & Phan, P. (1991). CEO Tenure as a Determinant of CEO Pay. The Academy of Management Journal, 34(3), 707-717. Retrieved January 15, 2021, from http://www.jstor.org/stable/256413