Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Event-study
From
Nick Cox <[email protected]>
To
[email protected]
Subject
Re: st: Event-study
Date
Wed, 27 Feb 2013 10:06:58 +0000
You need first to spread the benchmark value to other observations.
One general way to do that is
egen benchmark = total(sum_for_each_day / (dif == -1)), by(<whatever>)
Then divide by -benchmark- and multiply by 100.
See a more detailed and systematic review of similar problems at
http://www.stata-journal.com/article.html?article=dm0055
Nick
On Wed, Feb 27, 2013 at 9:10 AM, Sven Künzel <[email protected]> wrote:
> I have a question concerning the use of formula respectively scaling a
> value. I am doing an Event-Study. I have calculated the sum_for_each_day.
> Now I want to scale this sum_for_each_day in percent. The sum prior to the
> event (dif=-1) has to take the value 100% and has to represent the benchmark
> for the others sums in the event_window. My datasample with the new variable
> needed looks like this. Is there a simple way of fitting in a formula
> respectively rescaling a value?
>
> countrynam spreads group_id datum event_date dif event_window
> abnormal_return sum_for_each_day sum_for_each_day_Percent
>
>
> Portugal 0.28 1 20dec2010 23dec2010 -3 0
> 0.6
> Portugal 0.30 1 21dec2010 23dec2010 -2 1
> 0.5 1.80 102
> Portugal 0.33 1 22dec2010 23dec2010 -1 1
> 0.6 1.75 100
> Portugal 0.32 1 23dec2010 23dec2010 0 1
> 0.5 1.80 102=(((1.8-1.75)/1.75)+1)*100
> Portugal 0.33 1 27dec2010 23dec2010 1 1
> 0.6 2.05 117
> Portugal 0.35 1 28dec2010 23dec2010 2 1
> 0.4 2.00 114
> Portugal 0.34 1 20dec2010 23dec2010 3 0
> 0.5
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/