Hi,

I am trying to anlysis an (unbalanced) panel data which covers around 6000 companies (79000 frim-year observations) in around 99 industry subsection across 36 countries from 1990 to 2010.
I am planning to use the random fixed-effect model. And according to advice that 'Run a model with firm, country and time fixed effects. Then remove the country fixed effects and add the country variables. Then remove the firm fixed effects and add the firm variables.'

I am trying to use the code
HTML Code:
xtset company_ID year,yearly
xtreg Y X i.company_ID i.Country i.Industry i.year, re
esttab using test_pat1.xls, ar2(%8.4f) se(%8.4f) brackets nogap replace
but the result shows
HTML Code:
r(103);
(i.e., too many variables specified)

Assume I have a list of firm control variable: F1, F2, F3 and a list of country variables: C1, C2, C3. Could you please give me some suggestions about this?
1.which code should I use to run this random fixed-effect model and
2.especially how to get the firm fixed-effect variable?
3.how to set industry fixed effect variable when I have the firm fixed-effect model?

Thanks in advance.