Abnormal Returns using multi-factor rolling regression (even study) - Using SAS (or Stata)
Budget: $50 – $300 CAD
Looking for help running a multi-factor regression analysis.
The analysis will consist of an event study, to observe the price movements of the stocks around the time of their deletion from the S&P 500 Index. However, rather than using raw returns for abnormal returns, I will calculate the abnormal return for each security in the final deletions sample using a modified market model (to be provided)
I will run a regression for each of the deleted securities using data for the observed dependent variables starting 253 days before the AD+1 through 253 days following that date. I will use AD+1 as the “event” date since deletions are usually occurring after market close making AD+1 the actual date where markets can react to the news. That is fine but it is more typical to just use AD as the event day. I decided to use 253 days since it is the average number of trading days in a year for the NYSE. I will also be observing the security’s behavior in the few days following the AD as well as the abnormal return during the month following AD (AD-30: AD+30).
AD = Announcement day, as in the day where the deletion is announced.
I have all the data but simply do not know how to program it considering that each stocks in the sample have different AD
Ultimate goal is to obtain abnormal returns
The analysis will consist of an event study, to observe the price movements of the stocks around the time of their deletion from the S&P 500 Index. However, rather than using raw returns for abnormal returns, I will calculate the abnormal return for each security in the final deletions sample using a modified market model (to be provided)
I will run a regression for each of the deleted securities using data for the observed dependent variables starting 253 days before the AD+1 through 253 days following that date. I will use AD+1 as the “event” date since deletions are usually occurring after market close making AD+1 the actual date where markets can react to the news. That is fine but it is more typical to just use AD as the event day. I decided to use 253 days since it is the average number of trading days in a year for the NYSE. I will also be observing the security’s behavior in the few days following the AD as well as the abnormal return during the month following AD (AD-30: AD+30).
AD = Announcement day, as in the day where the deletion is announced.
I have all the data but simply do not know how to program it considering that each stocks in the sample have different AD
Ultimate goal is to obtain abnormal returns