estudy2 & eventstudy cant offer me what I want

#1
Hi!

For our master thesis, me and my partner want to construct an event study, but the packages "estudy2" & "eventstudy" does not seem to get around with some aspects. I wonder if there are someone out here on this forum that could help get around those issues. We are struggeling quite as lot so we would be very thankful for your help!

We have all data on stock returns and market indexes. Variable name for country and the market index is the same.

Specifically the aspects I cant get around are:
1. We want to regress returns on different market indexes dependent on which country the firm is situated, not just one market index.
2. We want to segregate CARs with respect to dummy variables. That is: Calculate average CARs, standard errors, an t-value/p-value for all observations when dummy is 1 and when dummy is 0.
3. We want to construct three different event windows, not just one like the packages seems to offer: [-21,-11], [-10,5], [6,20]

We greatly appreciate any effort on trying to help us with these problems!

Best,

Kristoffer & Jonas
 
Last edited:
#3
What form of regression are you using?
To estimate the returns we are going to conduct a simple market model regression:

estimated Rit = ai + bi*MARKET_RETURNit

Where MARKET_RETURN is the return in the market of the company is mainly operating, EG for Apple we use US index (SP500), for Volkswagen we us German index (DAX) etc.

AR= Rit - Estimated Rit

CAR = All AR during the event windows etc..