Regarding panel data random effects and fixed effects

[shubham padte]

Hello, I am studying finance and I am using stata for my dissertation which is on the topic the effect of digital payments on commercial bank performance, my dependent variables are ROA, ROE, NIM and my main independent variables are number of digital payment users and transaction value, I tried to run panel data fixed effects regression but my R squared is coming around 0.02 and for random effects it is 0.32. I am not able to understand why there is such a big difference and i tried transforming the data for multicollinearity as well but it isnt helping at all . If anybody could please help me out would be really grateful

[Carlo Lazzaro]

Shubham:
welcome to this forum.
1) the outcome table on the left shows no panel-wise effect (sigma_u=0);
2) your second (fe) regresssion shows a decent R_sq within, but you should investigate the functional from of the regressand;
3) in both models, I would invoke cluster-robust standad errors;
4) as per FAQ, please do not post screenshits, but use CODE delimiters to share what you typed and what Stata gave you back. Thanks.