Event Study with events on non-trading days.

[Michela Saba]

Hi there,
I am currently conducting an event study on Stata aiming to investigate the effect of Elon Musk's tweets on the stock market. However, some tweets seem to have been posted on non-trading days. Therefore, when I compute the following commands, the tweet dates are not matched with the date of the stock returns, making my analysis invalid.

sort group_id date
by group_id: gen datenum=_n
by group_id: gen target=datenum if date==event_date
egen td=min(target), by(group_id)
drop target
gen dif=datenum-td

I am attaching a screenshot of my data:
1. Company (or sub-group) whose tweet was posted on a trading day
2. Company (or sub-group) whose tweet was not posted on a trading day

How can I overcome this issue?
Many thanks!

[Devra Golbe]

Hi Michela,

I'm having trouble understanding what you want your code to do. I suspect you are trying to define the event and estimation windows. Here are two references that may help:

Event Study with Stata: a Step-by-Step Guide, from the Princeton U Library

Subhan Ullah, Ghasem Zaefarian, Rizwan Ahmed, Danson Kimani, How to apply the event study methodology in STATA: An overview and a step-by-step guide for authors,Industrial Marketing Management,Volume 99,2021

It will also be helpful for you to consult the Forum FAQs to learn how to formulate a question so as to maximize the likelihood of getting an answer. In particular, you will want to use -dataex- to produce a data example to illustrate your question rather than posting screenshots (or attachments many folks will not open.)

Best,
Devra