Event study question; how to create loops to measure the effect of earnings announcements of peer firms by industry and subindustry?

[Julien Maas]

Hello!

My name is Julien and I am a Finance master student. I have been trying for a while now to construct some loops yet I keep running into some issues in the preparation.

My main question is;
how can I create loops to measure the effect of earnings announcements of peer firms by industry and subindustry?

My code for this currently looks as following;

expand `Distance'

bys firmid eventdate: gen int tau=_n + `Estimationwindowstart' -1
assert(tau<=`Eventwindowend')

* creating eventid *
egen eventid = group(industrycode eventdate)
egen eventid2 = group(industrycode subindustrycode eventdate)

*generate real dates*
gen date = eventdate+tau
format date %tbtrdates

* loop 1 using eventid 1 *
rangestat (reg) ret mktrt if tau<=`Estimationwindowend', by(eventid) interval(tau . .)

bys eventid: egen cons_coeff=mean(b_cons)
bys eventid: egen mktrt_coeff=mean(b_mktrt)
gen NR10 = cons_coeff + mktrt_coeff*mktrt
drop if NR10==.
drop cons_coeff
drop mktrt_coeff
drop reg_nobs reg_r2 reg_adj_r2 b_mktrt b_cons se_mktrt se_cons


gen AR10 = ret - NR10


However I think I have made a mistake with the eventid(s) and that I need to do something else. Moreover I am currently looking at the effect of announcements of the firm itself, while I want to look at the effect of the announcements of peer firms. I am thus thinking of maybe something like including factor dummy variables into the loop, yet I haven't been able to do that as Stata gave me the error; factor-variable operators not allowed.

I'll gladly hear from you what you think is best and what tips and advice you may have,

Thank you in advance;

Kind regards, Julien Maas.

[Julien Maas]

Note. I think I solved this by creating seperate industry and subindustry variables connected to the event and by creating a new firmid.

[Julien Maas]

Note. This issue is solved, if anyone needs the code just let me know.