Re: st: How to generate a dummy variable based on groups of variables in t-1: Simple solution

["Jean-Marie Meier" <jean-marie.meier@web.de>]

Hi,

first, thanks to all that helped me.

When I was testing the proposals I got for my problem, I suddenly had an idea how to do it in a quick and nice way:

The problem was: How do I generate a dummy variable which is 1 for a given firm x and year t if there was a takeover in t-1 in the same industry (same industry code) and which is 0 otherwise ?

First, for clarity, drop all variables except

industry code

dummy_whether_firm_was_taken_over

firm identifier

year


The 3 line solution then is:

keep if dummy_whether_firm_was_taken_over == 1

(only observations remaining firms and the year in they which were taken over (firms that are taken over leave the dataset anyhow))

duplicates drop industry_code year, force

(as I want a dummy variable which is 1 if there was a takeover in an industry in t-1, I use the duplicate command for industry code and year, so that for every industry code and year combination I have one observation if there was takeover in the industry and none if there wasn't

replace year = year+1

(as I wanted to know, whether there was a takeover in the industry the year before, I simply add 1 to the year variable, then I save the dataset and perform a m:1 merge with the original dataset as master file)

Kind regards,

Jean-Marie

On Apr 30, 2011, at 2:50 AM, "Jean-Marie Meier" <jean-marie.meier@web.de> wrote:

> Dear Stata list,
>
> It would be great if anyone could help me with this issue as I spent hours on it without solving it:
>
> I have a dataset with all companys listed on US stock exchanges from 1979 up to now.
> Every observation in my dataset are balance sheet and profit & loss account data for a given firm in a given year.
>
> I have a dummy variable, which is 1 for an observation if a firm was taken over and left the dataset - e.g. firm ABC was taken over in 1985, then for all observations up to 1984 for this firm the dummy variable is 0 and for the observation in 1985 the dummy variable is 1. If a firm was not taken over, than the dummy variable is obviously always 0.
>
> All firms/observations in my dataset are assigned an industry code between 1 and 48 (Fama French industry classification).
>
> My problem now is:
>
> How do I generate a dummy variable which is 1 for a given firm x and year t if there was a takeover in t-1 in the same industry (same industry code) and which is 0 otherwise ?
>
> Looking forward to your answers
>
> Kind regards,
>
> Jean-Marie Meier
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/