Adjusted Standardized Cross-Sectional or BMP Test

[Frank Schmidt]

Adjusted Standardized Cross-Sectional or BMP Test

Dear all,

I am currently working on an event study following the Princeton guidelines (http://dss.princeton.edu/online_help...ventstudy.html). To make my results (especially my AR/CARs) more robust, I am planning to conduct the following tests:

a.) Standardized Cross-Sectional or BMP Test (Abbr.: StdCSect Z), as proposed by Boehmer, Musumeci and Poulsen (1991) --> A standardized cross-sectional method which is robust to the variance induced by the event

Link to paper: http://www.sciencedirect.com/science...04405X9190032F

b.) Adjusted Standardized Cross-Sectional or BMP Test (Abbr.: Adjusted StdCSect Z), as proposed by Kolari and Pynnönen (2010)
--> Modification to the BMP-test to account for cross-correlation of the abnormal returns

Link to paper: http://rfs.oxfordjournals.org/conten.../3996.abstract

Therefore, I was wondering if anyone has heard of/seen commands or codes in Stata to perform either one of those tests. I would be more than happy about any hints or tips.


I highly apologize for my inability of not already providing a code here. I first wanted to ask the community if there is already some solution out there, before starting the task.

Any help is very much appreciated. Thanks in advance!

Frank

[Nina Scherl]

Hi Frank,

I am facing exactly the same problem right now.
My plan is to do an event study in STATA which accounts for event-induced variance increases and event-date clustering.
So far I have also followed the Princeton guidelines and I am searching for a way to do the Adjusted Standardized Cross-Sectional test (ADJ-BMP) as suggested by Kolari & Pynnönen (2010).

Did you find any solution which might also help me?

Thanking you in anticipation.
Best regards,
Nina

[Claudia Kuehn]

Hi Nina,

I found your post and am currently working on the same methodology. Have you found any user written code for this test?

Many thanks

Claudia

[Johnson Ementraubt]

Hi Claudia,

I'm also conducting an event study with standardized cumulative abnormal returns. Have you already found a way to code this in STATA?

Thanks in advance,

Johnson

[Amira El-Feiaz]

GUYS! Use the estudy command, it has all of these tests!

[Tina Dao]

Dear all,

I'm trying to use estudy command, but I meet a couple of problems.
I'm researching about the cartel in EU. I have 3 events (Investigation, Fine, Appeal) for total 441 firms, every firm has a different date due to the decision of EU committee.
So if I use the estudy command, do you know how to specify the evedate in the command?
How can I list my varlist? Can I use foreach here?
Should I separate my data to 3 by the events?

Thanks in advance,
Tina

[Rech Lee]

Hi Guys,

Why bother with all these yet you can prepare your data into three files: request file, firm file, and market file and use
event study tools. I am sure this is faster and accurate.
Unless you have reservations on this. Let me know if this assists!

RL

[Andrea Stringhetti]

Hi everyone, can you please explain better how you solved the issue? I have to implement a BMP test but I'm facing different issues. The first one is that the variable date is not a unique value.
Can you give me any hint?
Rech Lee do you have any idea?

Thank you very much,

AS

[Thomas Kaspereit]

Dear all,

I disagree with Rech Lee's comment. There are (and at the time of his comment already were) good user-written commands for event studies available. Please see my recent article on this topic: "Event studies with daily stock returns in Stata: Which command to use?"

https://www.researchgate.net/publica...command_to_use

Best
Thomas

[Willemijn Olree]

Hi everyone,

I am facing the same problem as above. Unfortunately, I cannot use the estudy or eventstudy2 command for my whole research. I am conducting an event study whereby the events consist of extreme changes in the daily price of oil, gas and coal. This result in 47 events. I examined the impact of each event on a sample of 1326 firms. So, every event occurs to every firm. First, I create the CAR per firm per event and thereafter the CAAR per event. However, I also need to aggregate this over the events itself. So, positive and negative events for oil, gas and coal. I want to use the adjusted standardized cross-sectional test, however, I really don't know how to do this in Stata. Has anyone some ideas? Thanks a lot!