I would like to use gretl in my thesis for calculating short term abnormal return after a occuring a stock split and check its significance.
I use a model, where Abnormal return(ARi)=Ri (actual return)- alpha+(beta*Rm).
In the function alpha and beta is caculated from the data before the split (from the calculate window) and Rm is the actual market return (from the event window). I already have run an OLS on the data before the split, so I get the alpha and the beta of the stocks before the event. I have some problem how to run an other regression with the given alpha and beta (calculated from the past) and the market return, but now using the actual market return and not that of from the past? Could you please help me how to do it? And how to check the significance of the abnormal return?