Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: [Stata 8] Merge question


From   "Scott Merryman" <smerryman@kc.rr.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: [Stata 8] Merge question
Date   Fri, 6 May 2005 19:39:22 -0500

Cameron,

Try this:

clear
use http://www-personal.umich.edu/~chooper/stata/returns

sort perm date

save "C:\Documents and Settings\Scott_2\Desktop\returns.dta", replace

use http://www-personal.umich.edu/~chooper/stata/event_dates

sort perm date

save "C:\Documents and Settings\Scott_2\Desktop\dates.dta", replace

use "C:\Documents and Settings\Scott_2\Desktop\returns.dta", clear

merge permno date using "C:\Documents and
Settings\Scott_2\Desktop\dates.dta"

gen type = 1 if _m == 3

replace type =1 if _m[_n + 1] == 3 | _m[_n + 2] == 3 |_m[_n - 1] == 3 |
_m[_n - 2] == 3 

replace type =3 if _m[_n - 3] == 3 | _m[_n - 4] == 3 | _m[_n - 5] == 3 |
_m[_n -6] == 3 | _m[_n -7] == 3 

replace type =2 if _m[_n + 3] == 3 | _m[_n + 4] == 3 | _m[_n + 5] == 3 |
_m[_n +6] == 3 | _m[_n +7] == 3 

l in 39/57


Hope this helps,
Scott

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-
> statalist@hsphsun2.harvard.edu] On Behalf Of Cameron Hooper
> Sent: Friday, May 06, 2005 1:56 PM
> To: statalist@hsphsun2.harvard.edu
> Subject: st: [Stata 8] Merge question
> 
> Dear list,
> 
> I want to merge event dates with a series of daily stock returns and
> subsequently define various event windows. What seemed like a straight
> forward problem is actually proving surprisingly difficult (for me
> anyway).
> 
> I have made up some illustrative data. Suppose I have a file containing
> daily returns for a couple of firms like this:
> 
> 
> . version
> version 8.2
> . use http://www-personal.umich.edu/~chooper/stata/returns
> . list in 1/10
> 
>       +-----------------------------------------+
>       | permno   ticker        date         ret |
>       |-----------------------------------------|
>    1. |  10107     MSFT   03 Jan 00    -.001606 |
>    2. |  10107     MSFT   04 Jan 00   -.0337802 |
>    3. |  10107     MSFT   05 Jan 00    .0105438 |
>    4. |  10107     MSFT   06 Jan 00   -.0334981 |
>    5. |  10107     MSFT   07 Jan 00    .0130682 |
>       |-----------------------------------------|
>    6. |  10107     MSFT   10 Jan 00    .0072911 |
>    7. |  10107     MSFT   11 Jan 00   -.0256125 |
>    8. |  10107     MSFT   12 Jan 00   -.0325714 |
>    9. |  10107     MSFT   13 Jan 00    .0189014 |
>   10. |  10107     MSFT   14 Jan 00    .0411594 |
>       +-----------------------------------------+
> 
> I also have a file of event dates:
> 
> . use http://www-personal.umich.edu/~chooper/stata/event_dates
> . list
> 
>       +-----------------------------+
>       | permno   ticker        date |
>       |-----------------------------|
>    1. |  10107     MSFT   04 Jan 00 |
>    2. |  10107     MSFT   10 Mar 00 |
>    3. |  10107     MSFT   22 Jun 00 |
>    4. |  10107     MSFT   18 Sep 00 |
>    5. |  10107     MSFT   28 Dec 00 |
>       |-----------------------------|
>    6. |  12490      IBM   12 Jan 00 |
>    7. |  12490      IBM   07 Apr 00 |
>    8. |  12490      IBM   25 Jul 00 |
>       +-----------------------------+
> 
> Merging is straightforward. However I also want to create a set of 3
> event windows for each event date: pre-event, event, and post-event.
> Define the event date as 0 then
> 
> pre-event = days -7 to -3
> event = days -2 to +2
> post-event = days +3 to + 7
> 
> ** These are trading dates not calendar dates (each observation is a
> trading date). **
> 
> I thought this would be easy, but I just can't seem to make it happen.
> Please note that in practice I have a sample containing daily data for
> thousands of firms, many of which have over 10 years of data. I also
> have 4 event dates per firm-year. This means that the method shown at
> 
> http://dss.princeton.edu/online_help/analysis/multiple_event_dates.htm
> 
> which creates a duplicate set of observations for each event date /
> company combination is impractical. The resulting file is simply too
> large.
> 
> Here is a mocked up file showing what I am aiming for. I have shown
> data around one event_date. The variable TYPE is defined as (0 =
> non-event, 1 = event, 2 = pre and 3 = post.
> 
> . use http://www-personal.umich.edu/~chooper/stata/desired_result
> . list in 39/57
> 
> 
> +-------------------------------------------------------------------+
>       | permno   ticker        date   event_d~e         ret   type
> indx |
> 
> |-------------------------------------------------------------------|
>   39. |  10107     MSFT   28 Feb 00           .    .0027379      0
> . |
>   40. |  10107     MSFT   29 Feb 00           .   -.0238908      0
> . |
>   41. |  10107     MSFT   01 Mar 00           .    .0160839      2
> -7 |
>   42. |  10107     MSFT   02 Mar 00           .    .0282175      2
> -6 |
>   43. |  10107     MSFT   03 Mar 00           .    .0294511      2
> -5 |
> 
> |-------------------------------------------------------------------|
>   44. |  10107     MSFT   06 Mar 00           .   -.0572172      2
> -4 |
>   45. |  10107     MSFT   07 Mar 00           .    .0248276      2
> -3 |
>   46. |  10107     MSFT   08 Mar 00           .    .0289367      1
> -2 |
>   47. |  10107     MSFT   09 Mar 00           .    .0464356      1
> -1 |
>   48. |  10107     MSFT   10 Mar 00   10 Mar 00         .01      1
> 0 |
> 
> |-------------------------------------------------------------------|
>   49. |  10107     MSFT   13 Mar 00           .    -.029703      1
> 1 |
>   50. |  10107     MSFT   14 Mar 00           .   -.0293367      1
> 2 |
>   51. |  10107     MSFT   15 Mar 00           .    .0026281      3
> 3 |
>   52. |  10107     MSFT   16 Mar 00           .           0      3
> 4 |
>   53. |  10107     MSFT   17 Mar 00           .    .0419397      3
> 5 |
> 
> |-------------------------------------------------------------------|
>   54. |  10107     MSFT   20 Mar 00           .   -.0201258      3
> 6 |
>   55. |  10107     MSFT   21 Mar 00           .     .055199      3
> 7 |
>   56. |  10107     MSFT   22 Mar 00           .    .0048662      0
> . |
>   57. |  10107     MSFT   23 Mar 00           .    .0835351      0
> . |
> 
> +-------------------------------------------------------------------+
> 
> Any suggestions greatly appreciated.
> 
> Cameron
> 
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index