Producing table with aggregated results for different event windows

[Natalia Tsorapi]

Hello everyone!

I am currently conducting an event study where I calculate cumulative abnormal returns over different event windows for a specific sample of firms and a sample period. These event windows for example are (-1,+1), (-5,+5), (-10,+10) with t=0 being my event date. I want to test the significance of them for each event window and produce a table including the results for all the event windows aggregated. I use the following command for example for testing the significance of cumulative abnormal returns over one of my event windows which produces a table only for that event window.

reg CAR if dif==0, robust

How is it possible to produce a table including the results for all cumulative abnormal returns that correspond to all my event windows?

Thank you in advance,
Natalia

[Attaullah Shah]

You can try the following if you have numbered your event windows.

forval i=1/100{
reg CAR if dif==0 & window==`i', robust
outreg2 using CARs.doc, append ctitle(window`i')
}

[Natalia Tsorapi]

Dear Attaullah Shah,

Thank you very much for the response.

I have basically created different columns for the event windows so eg I have event_window1 in one column, event_window2 in another one and so on where I have given the value 1 when a day is included in my event window each time. So based on these different windows I have also created different columns for cumulative abnormal returns named CAR1, CAR2, CAR3 and so on. Consequently I think I cannot implement the above-mentioned code.

Practically I have somehow to test the significance choosing multiple variables (CAR1, CAR2, etc) and have the results simultaneously instead of taking individual results for each variable that means different tables.

Thank you in advance!

[Natalia Tsorapi]

Maybe there is some kind of option through the stata menu instead of a command? I tried to search but did not find anything so far.

[Attaullah Shah]

If you have different CARs in different columns, then name your CARs sequentially i.e. CAR1, CAR2, etc. and apply the following code which assume maximum of 100 CARs, which you can modify as per your requirement

  
 forval i=1/100{
reg CAR`i' if dif==0 , robust
outreg2 using CARs.doc, append ctitle(CAR`i')
}