Hello. I am a 3rd year university student, working on the quantitative part of my dissertation which is due in a week and am in panic as I only started using Stata for the first time last week.
Although I have read the instructions, I am still not unable to conduct the Event study that I need for my dissertation, so I apologise if this would be classed as 'elementary' but I am desperate.
A summary of my investigation and data:
1) I have a sample of 70 announcements made by the company I am investigating (It is a large company on the Ftse100).
I am trying to understand whether there is any impact from the announcements of interest on the company's share price.
There are in fact two types of announcements that are captured within these 70; 50 of which are one type (I have called this Type 1) and the other 20 type (Type 2).
I am interested in testing the effect of all of them together (as one test) and then the two subsamples separately )as a second test (hence I have created a dummy variable, Type1=1 and Type2=0).
2) I am using 3 estimation windows as this is testing the announcements of 1 company. Where day 0 = announcement day ( announcements made early on in the day)
Estimation window 1 = 6 day window (3,-2,-1 -> 0,1,2)
Estimation window 2 = 4 day window (-2,-1 -> 0,1)
Estimation window 3 = 2 day window (1 -> 0)
3) I have created a dummy variable for days before the announcement date (-3,-2,-1 = 0) and after the announcement date (0,1,2 = 1) (including the announcement date itself 0 as the announcement is made during the market day, so under perfect market efficiency assumption I expect immediate reaction).
4) I am using the Ftse100 as an independent variable to understand how much the general market movement is affiliated with observed share price movement for the particular company I am investigating.
The dependent variable (Y) will therefore be the share price (of the company). I have taken the logs for these.
Below is an example of the data that I have sorted.
Share price Day After Type Ftse100 logprice logftse100
£x -3 0 1 £ft log£x log£ft1
£x -2 0 1 £ft log£x log£ft1
£x -1 0 1 £ft log£x log£ft1
£x 0 1 1 £ft log£x log£ft1
£x 1 1 1 £ft log£x log£ft1
£x 2 1 1 £ft log£x log£ft1
£y -3 0 0 £ft log£y log£ft1
£y -2 0 0 £ft log£y log£ft1
£y -1 0 0 £ft log£y log£ft1
£y 0 1 0 £ft log£y log£ft1
£y 1 1 0 £ft log£y log£ft1
£y 2 1 0 £ft log£y log£ft1
etc.
I do hope the above is clear, I really would appreciate it if anyone knows the commands that I should use in order to conduct the regression for the event study.
Any support would be extremely appreciated!
Although I have read the instructions, I am still not unable to conduct the Event study that I need for my dissertation, so I apologise if this would be classed as 'elementary' but I am desperate.
A summary of my investigation and data:
1) I have a sample of 70 announcements made by the company I am investigating (It is a large company on the Ftse100).
I am trying to understand whether there is any impact from the announcements of interest on the company's share price.
There are in fact two types of announcements that are captured within these 70; 50 of which are one type (I have called this Type 1) and the other 20 type (Type 2).
I am interested in testing the effect of all of them together (as one test) and then the two subsamples separately )as a second test (hence I have created a dummy variable, Type1=1 and Type2=0).
2) I am using 3 estimation windows as this is testing the announcements of 1 company. Where day 0 = announcement day ( announcements made early on in the day)
Estimation window 1 = 6 day window (3,-2,-1 -> 0,1,2)
Estimation window 2 = 4 day window (-2,-1 -> 0,1)
Estimation window 3 = 2 day window (1 -> 0)
3) I have created a dummy variable for days before the announcement date (-3,-2,-1 = 0) and after the announcement date (0,1,2 = 1) (including the announcement date itself 0 as the announcement is made during the market day, so under perfect market efficiency assumption I expect immediate reaction).
4) I am using the Ftse100 as an independent variable to understand how much the general market movement is affiliated with observed share price movement for the particular company I am investigating.
The dependent variable (Y) will therefore be the share price (of the company). I have taken the logs for these.
Below is an example of the data that I have sorted.
Share price Day After Type Ftse100 logprice logftse100
£x -3 0 1 £ft log£x log£ft1
£x -2 0 1 £ft log£x log£ft1
£x -1 0 1 £ft log£x log£ft1
£x 0 1 1 £ft log£x log£ft1
£x 1 1 1 £ft log£x log£ft1
£x 2 1 1 £ft log£x log£ft1
£y -3 0 0 £ft log£y log£ft1
£y -2 0 0 £ft log£y log£ft1
£y -1 0 0 £ft log£y log£ft1
£y 0 1 0 £ft log£y log£ft1
£y 1 1 0 £ft log£y log£ft1
£y 2 1 0 £ft log£y log£ft1
etc.
I do hope the above is clear, I really would appreciate it if anyone knows the commands that I should use in order to conduct the regression for the event study.
Any support would be extremely appreciated!
Comment