Loop regression: store constants to calculate median

[Daniela Bonilla]

Hi,

I have the following code and would like to store the constants being produced from the loop regression so that I can calculate the median. Any help?

encode GVKEY, gen(GVKEY_num)
estimates clear
levelsof GVKEY_num, local(groups)

local start_group1 1
local end_group1 88

scalar sum_alpha1 = 0  // Initialize a sum variable for alpha1
scalar count_alpha1 = 0  // Initialize a count variable for alpha1

foreach group of local groups {
    if `group' >= `start_group1' & `group' <= `end_group1' {
        regress dealexcessreturns marketexcessreturns if GVKEY_num == `group', robust
        estimates store capm_`group'
        scalar alpha1 = _b[_cons] - (rf + benchmarkyrret * marketexcessreturns)
        display "`group' alpha1: " alpha1
        
        // Update the sum and count for alpha1
        scalar sum_alpha1 = sum_alpha1 + alpha1
        scalar count_alpha1 = count_alpha1 + 1
    }
}

[Maarten Buis]

. // reset the frames
. frames reset

.
. // open example data
. sysuse nlsw88, clear
(NLSW, 1988 extract)

.
. // We are going to overwrite the data with the coefficients
. // So doing it in a separate frame allows us to keep the original data
. frame copy default result

. frame change result

.
. // some categories are too small
. numlabel occlbl, add

. tab occupation

               Occupation |      Freq.     Percent        Cum.
--------------------------+-----------------------------------
1. Professional/Technical |        317       14.17       14.17
        2. Managers/Admin |        264       11.80       25.97
                 3. Sales |        726       32.45       58.43
    4. Clerical/Unskilled |        102        4.56       62.99
             5. Craftsmen |         53        2.37       65.36
            6. Operatives |        246       11.00       76.35
             7. Transport |         28        1.25       77.60
              8. Laborers |        286       12.78       90.39
               9. Farmers |          1        0.04       90.43
        10. Farm laborers |          9        0.40       90.84
              11. Service |         16        0.72       91.55
    12. Household workers |          2        0.09       91.64
                13. Other |        187        8.36      100.00
--------------------------+-----------------------------------
                    Total |      2,237      100.00

. recode occupation (9=10) (12=13)
(3 changes made to occupation)

.
. // estimate the regression for each occupation and store the coefficients and
. // standard errors
. statsby _b _se , by(occupation) clear : regress wage i.south, vce(robust)
(running regress on estimation sample)

      Command: regress wage i.south, vce(robust)
           By: occupation

Statsby groups
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
...........

.
. // find the median
. sum _b_cons, detail

                          _b[_cons]
-------------------------------------------------------------
      Percentiles      Smallest
 1%     3.519447       3.519447
 5%     3.519447       3.744918
10%     3.744918       5.535785       Obs                  11
25%     5.535785       6.570047       Sum of wgt.          11

50%     6.942338                      Mean           7.323996
                        Largest       Std. dev.      2.667199
75%     9.384748       7.760388
90%     11.01399       9.384748       Variance        7.11395
95%     11.93449       11.01399       Skewness       .2739177
99%     11.93449       11.93449       Kurtosis       2.260087