In my paper I want to know if there is correlation between sustainability and and the performance of a company. I want to explain the StockPrice with a lot of other variables and for sure the ESG-Score.
I created panel data and now I need to know wether I use the time FE, individual FE or twoways FE.
When I create the FE models the individual FE model is the only one which has a significant ESG-Score. The ESG-Score is very important to me so I thought let's use the individual FE-model, but doing the Hausmanntest it is loosing to the random effects model.
Later on I found out, that, surprise surprise you can't choose a model based on wether an variable that is important to you is significant or not.
So how do I decide which FE model I run against the random model?
Is there a test or something else that is answering that?
I also found the question (Effects in panel models "individual", "time" or "twoways"), but it's not really giving a final answer.
Fore sure I've been searching the internet twice, but I didn't found an answer specific for finance panel data, where you don't have things like gender that usually don' change over time.
I would be very greatful, if you could please help me. :)
