Hello,

I am doing a panelregression with industry and year FE
Is it possible to create a new panelid

Code:
egen panelid = group(industry_key year)
and set this as the identifier

Code:
xtset panelid
and then run the regression like this

Code:
xtreg firm_beta_w esg_single_lag1 operating_leverage_lag1 diversification_lag1 RnD_lag1 advertising_lag1 leverage_lag1 capex_lag1 cash_lag1 size_lag1 earnings_variability_lag1 state_tax_lag1, fe

Fixed-effects (within) regression               Number of obs     =      5,011
Group variable: panelid                         Number of groups  =        413

R-sq:                                           Obs per group:
     within  = 0.0118                                         min =          1
     between = 0.0212                                         avg =       12.1
     overall = 0.0213                                         max =         60

                                                F(11,4587)        =       5.00
corr(u_i, Xb)  = 0.0711                         Prob > F          =     0.0000

-------------------------------------------------------------------------------------------
              firm_beta_w |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
--------------------------+----------------------------------------------------------------
          esg_single_lag1 |  -.0915714   .0362866    -2.52   0.012    -.1627106   -.0204322
  operating_leverage_lag1 |  -.0000463   .0010595    -0.04   0.965    -.0021234    .0020307
     diversification_lag1 |    .002271   .0033245     0.68   0.495    -.0042466    .0087887
                 RnD_lag1 |   .9691038   .1826924     5.30   0.000     .6109388    1.327269
         advertising_lag1 |  -.0395175   .0520266    -0.76   0.448    -.1415147    .0624798
            leverage_lag1 |   .0378296   .0380948     0.99   0.321    -.0368546    .1125138
               capex_lag1 |   .0393854   .1811958     0.22   0.828    -.3158456    .3946164
                cash_lag1 |  -.0029276   .0048622    -0.60   0.547    -.0124599    .0066047
                size_lag1 |   .0416424   .0143322     2.91   0.004     .0135443    .0697404
earnings_variability_lag1 |   .0049405   .0023344     2.12   0.034      .000364    .0095171
           state_tax_lag1 |  -.6195115   .1996369    -3.10   0.002    -1.010896    -.228127
                    _cons |    .715787   .1415617     5.06   0.000     .4382579     .993316
--------------------------+----------------------------------------------------------------
                  sigma_u |  .28925279
                  sigma_e |  .38517848
                      rho |  .36058832   (fraction of variance due to u_i)
-------------------------------------------------------------------------------------------
F test that all u_i=0: F(412, 4587) = 3.46                   Prob > F = 0.0000
The results look good to me but is this the rigth way to do it and to include industry and year FE?