Hi All,
New to stata (and statalist)!
I have been following princeton's guide on event study analysis. My stock variable is the aggregated average return for 15 nordic stocks, i.e., Stata interprets this as only one company with company_id "nordic" and different prices for different dates "price". The other variable is price_omxnordic, which is the nordic index that I am comparing to. I have the % returns as ret for "price" and market_return for price_omxnordic.
My goal is to do an event analysis for the date 2019-12-18, where I want to see if my 15 stocks react differently to the market index. As you can see, my event window is 17th-19th of december. My estimation window is from -119 days until -5 days prior to the 18th of December.
Does this look correct to you? I found that we have some abnormal returns for the whole event window, but I think that my ar_sd seems very low at 0.0016998 (according to the article https://dss.princeton.edu/online_hel...tudy.html#test, it should be >1.96 to have significance).
Furthermore, now that I have these results, I have no idea how to output them in a graph. Preferably, I would want like an ANNOVA table or similar (showing significance levels, p-values, etc). Or what is the standard output format for an event analysis and does anyone have a code for this?
One of my concerns is that my dataset is too narrow with the average return of the nordic companies already aggregated in "nordic". Perhaps I should include them separately in the analysis?
Hope my questions make sense.
Many thanks.
Best,
Sofia
'
New to stata (and statalist)!
I have been following princeton's guide on event study analysis. My stock variable is the aggregated average return for 15 nordic stocks, i.e., Stata interprets this as only one company with company_id "nordic" and different prices for different dates "price". The other variable is price_omxnordic, which is the nordic index that I am comparing to. I have the % returns as ret for "price" and market_return for price_omxnordic.
My goal is to do an event analysis for the date 2019-12-18, where I want to see if my 15 stocks react differently to the market index. As you can see, my event window is 17th-19th of december. My estimation window is from -119 days until -5 days prior to the 18th of December.
Does this look correct to you? I found that we have some abnormal returns for the whole event window, but I think that my ar_sd seems very low at 0.0016998 (according to the article https://dss.princeton.edu/online_hel...tudy.html#test, it should be >1.96 to have significance).
Furthermore, now that I have these results, I have no idea how to output them in a graph. Preferably, I would want like an ANNOVA table or similar (showing significance levels, p-values, etc). Or what is the standard output format for an event analysis and does anyone have a code for this?
One of my concerns is that my dataset is too narrow with the average return of the nordic companies already aggregated in "nordic". Perhaps I should include them separately in the analysis?
Hope my questions make sense.
Many thanks.
Best,
Sofia
'
Attached Files
