I am currently sitting on an understanding issue. I did some regressions (pooled OLS) on my panel data and hold for time fixed effects.

However, I am confused about the interpretation of the time fixed effect as my results are completely contradicting each other. I tried to find some literature about time-fixed effects but I could not find the answer I was looking for.

My scenario: I did a pooled regression by holding for time-fixed effects (monthly) as I try to investigate the impact of Fintech Funding on banks stock return. When I run the pooled OLS without holding time fixed my result is slightly positive (0.30) while running the regression with monthly time-fixed effects my result is negative (-1.12). I am confused on how to interpret the results. Why does the time-fixed effect leads to a different result? And also what is the theory behind this?

Additional: I did not use a fixed or random effect model because my Breusch-Pagan LM test indicate to use pooled OLS. However, when using the Stata command testparm my result show that time should be considered.


1 Answer 1


The time-fixed effect allows to eliminate bias from unobservables that change over time but are constant over entities and it controls for factors that differ across entities but are constant over time. So, with and without time-fixed effect result can be significantly different.

This might be a helpful source for theoretical understanding.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.