Error r(459) while applying eventstudy2

[Alexander Kolb]

Hi Guys,

for my master thesis I need to conduct an Event Study. I investigate if there is a connection between announcements related to movie adaptions of novels and the return of the movie studio which is doing the adaption. As I am new to Stata, I was really thankful for the eventstudy2 Programm by Thomas Kaspereit.

I am trying to apply it with the following formula:

eventstudy2 security_id date using securityreturnsfile,returns(security_returns) model(MA)marketfile(marketfile) marketreturns(market_returns) idmarket(market_id) evwlb(-1) evwub(1) eswlb(-120) eswub(-2) aarfile(Averageabnormalreturns)

But when I do so, I get the following error message:

variables date market_id do not uniquely identify observations in the using data r(459);

I have three files:

-eventlist.dta which is stored in memory (Contains: event_id;security_id;date)
-marketfile.dta (Contains:market_id;date;market_returns)
-securityreturnsfile.dta (Contains:security_id;market_id;date;price;securit y_returns)

The files are not equally sized. Securityreturnsfile is the biggest file, because it includes the return data of 7 Companies for the time from 2007-2016, while the marketfile also contains return data from 2007-2016, but only for the SP500. The eventlist contains 437 Columns with events. I am not sure if that plays a role, since I know, that this kind of error typically occurs while merging two files.

I really researched a lot, but I could not find any solution for my problem.

See attached screenshots of these files.

I hope I did not forget any information which is needed to find a solution. I also hope I am not stealing your time with a really simple problem.

Thanks in advance,

Alexander Kolb
securityreturnsfile-screenshot.tiffmarketfile screenshot.tiffeventlist screenshot.tiff

[Nick Cox]

Please do read and act on

http://www.statalist.org/forums/help#stata 12,1 12.4 12.5

[Alexander Kolb]

Thanks for the advice Nick. I now tried to follow this rules and additionally I exctracted some example data with dataex (I am confused why date is not displayed in Format DMY when using dataex:

Securityreturnfiles

input byte security_id int(market_id date) str17 Price double security_returns
15 500 17076 "20.279994"                     .
15 500 17077 "20.260003" -.0009857497985451785
15 500 17078 "20.310004"   .002467966070883513
15 500 17079 "20.399998"   .004431018329686238
15 500 17080 "20.449999"  .0024510296520616416
15 500 17083 "20.700001"   .012225037272618061
15 500 17084 "20.950003"   .012077390720898926
15 500 17085 "20.989997"  .0019090212063454161
15 500 17086 "21.150005"   .007623059688860426
15 500 17087 "21.389999"   .011347231359992556
15 500 17090 "21.710004"   .014960496258087805
15 500 17091 "21.640001"  -.003224458180661773
15 500 17092 "21.729996"    .00415873363406953
15 500 17093 "21.77"      .0018409575409033534
15 500 17094 "21.720001" -.0022966926963711386
15 500 17097 "21.619999"  -.004604143434431699
15 500 17098 "21.529994"  -.004163043670816148
15 500 17099 "21.610005"  .0037162574220876467
15 500 17100 "21.640001"  .0013880607616703732
15 500 17101 "21.660004"  .0009243530071925183
end
format %tdnn/dd/CCYY date

Marketfile:

input int(market_id date) double market_returns
500 17076                     .
500 17077   .002095693832562754
500 17078   .012060449423890642
500 17079  .0022367205670265435
500 17080 -.0026824944043040456
500 17083  .0007928837846738941
500 17084  .0020434527790284536
500 17085  -.002563943851270394
500 17086    .00954109816475706
500 17087   .002047239266877385
500 17090  .0025190492322865357
500 17091   -.00365945231333344
500 17092   .001282944127514195
500 17093  .0008492546188216433
500 17094  .0011997535017779124
500 17097   .006152304652678119
500 17098  .0002614232144570066
500 17099   .003513893030558407
500 17100   .004963019739207669
500 17101  -.008451630123442681
end
format %tdnn/dd/CCYY date

Eventlist:

input byte(event_id security_id) long date
1 12 17443
1 12 17518
1 12 17518
1 13 17800
1 14 17821
1 15 17870
1 12 17927
1 16 17927
1 16 18007
1 18 18092
1 15 18211
1 16 18246
1 14 18238
1 14 18304
1 13 18346
1 12 18431
1 17 18659
1 15 18693
1 17 18704
1 18 18709
end
format %tdnn/dd/CCYY date

See attached screenshots of these files.

Eventlist:

securityreturnsfile:

marketfile:

I hope I did not forget any information which is needed to find a solution. I also hope I am not stealing your time with a really simple problem.

Thanks in advance,

Alexander Kolb

[Nick Cox]

Thanks for this. I don't see any source given for the user-written program eventstudy2

I note that price is a string variable.. You don't seem to use it, but that looks wrong.

See also http://www.stata.com/support/faqs/da...d-time-values/ and

help duplicates 
duplicates list  date market_id  
* etc.

[Alexander Kolb]

Thanks Nick.

The Source for eventstudy2 would be: https://ideas.repec.org/c/boc/bocode/s458086.html .

Thank you very much for the suggestion with the duplicates code. That was the error. In my marketfile, after the last date there were about 100 rows filled with ".", that caused the error and after deleting this, it finally worked.

Thanks for your fast help!