is stock fixed effects necessary for a panel data where 80% of the firms are small firms and only 20% firms are medium to large? I suspect there would be little to no variation in majority of the sample in such case and adding a stock fixed effects would just kill the results. This is what I appear to understand from the paper (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3699777 . In my scenario, I would probably go with Industry fixed effects and a time (year fixed effects) then. A bit confused. Any advice is appreciated. Thanks. Note: I have panel data for 15 years.
Asked
Active
Viewed 27 times
