Newey-West HAC

noetsi

Fortran must die
#1
"The Newey-West HAC-robust standard errors for the OLS estimators are consistent when the error term is heteroskedastic, autocorrelated, or both, as long as the regressors are stationary and ergodic."

This confuses me since they are used with time series primarily and its unlikely that all the regressors will be stationary in time series. If they are all stationary then you don't really need these to correct the problem in the first place.