Hi Sven,

I've just had a look at the Stata manual and it looks like Newey West is only available for old panel regression. They use cluster approach for fixed and random effects.

I think I have enough to go back with a convincing argument. Thank you

P.S.: Maybe it's a bit too early to mention this, but the next gretl version will have more clustering options for the standard errors in the panel case. (It already had them for non-panel stuff.) Actually, it seems to me that then gretl will offer more flexibility than Eviews currently does, being closer to Stata in that respect, if I understand correctly.

There's no release date for the next gretl version, but I would expect that it's 2023c, not 2024a.

cheers

sven