Presence of Heteroscedasticity but no serial Correlation in Panel, how do I Proceed

[Akinyemi Afolabi]

I am working on the impact of Trade integration on poverty reduction in ECOWAS
Number of Countries: 15
Time: 10 years.

Below is my code.
After analysing, I can see that there is heteroscedasticity but no serial correlation. Can I proceed with GLS (xtgls)?
please help on what i should do next?

My codes and what I have done are in the picture below

I want to know what to do next and also the test to use in analysing the panel data after the presence of heteroscedasticity with no serial correlation.

[Andrew Musau]

Since you have only 15 countries, you may want to look at wild bootstrap standard errors.

ssc install boottest, replace
help boottest

Compare the wild-bootstrap standard errors with your non-clustered standard errors from -xtreg, fe-. If there is no big difference, then the heteroskedasticity does not affect your results.

[Rich Goldstein]

also see

help wildbootstrap

for StataCorp's version of the wild bootstrap