Hello there,

I am running a linear regression with panel data and fixed effects using the xtreg, fe command.

Running the Breusch Pagan test and the xtserial command I have problems of heteroskedasticity and autocorrelation. I know that I can manage heteroskedasticity with the vce(robust) command but how can I manage the assumption of autocorrelation in my regression model?

Dou have any advice?

Thanks,
Meike