Best method to find out the effect ESG ratings have on Stock Returns

[Dave Guillermo]

I've tried using fixed and random effects but stata shows that my control variables (ROA, ROE, log of total assets, net profit and eps) were either insignificant or is negatively related to stock returns and ESG scores are both insignificant and negatively related. Below are the results of the fixed effects analysis. Any help to solve this problem would be greatly appreciated!

[Maxence Morlet]

Well first of all you haven't included time FE, but I don't really understand your problem; you want significant results but don't have them, right?

Unfortunately there's not much to do about it, you may want to try some form of potential outcome model, or diff in diff if you can find an exogenous treatment. Otherwise that's just what the data have to say, so to speak.

Also all your regressors are jointly insignificant. You should perhaps completely modify your model, incorporate more flexibility, respecify variables, include other regressors...

[Joro Kolev]

You can try including time fixed effects as Maxence suggested, and excluding the ROA and ROE, because the latter two are endogenous and just alternative (accounting based) measures of returns.

And it is totally consistent with previous literature that ESG can have positive/negative/insignificant effect on returns.