Hi all,
I need your help in choosing the appropriate technique in panel data analysis with large N (607 companies) and small T (5 years). The cross-sectional dependence is identified using xtcsd command, while group-wise heteroskedasticity and serial correlation are also detected after running fixed effect regression model.
What technique is appropriate in such a case please?
Hi,
I think you can choose Robust standard error estimates for linear panel model. For more details please read page 38 of below attachment. https://www.princeton.edu/~otorres/Panel101.pdf
Hi,
I think you can choose Robust standard error estimates for linear panel model. For more details please read page 38 of below attachment. https://www.princeton.edu/~otorres/Panel101.pdf
rolling window regression with robust standard errors can be used. for the details following link may be useful
https://www.statalist.org/forums/forum/general-stata-discussion/general/1445420-asreg-update-rolling-window-regression-with-robust-standard-errors-newey-west-standard-errors-and-noconstant-options
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running