Balanced firm level panel

[Taiba Chau]

Hi, I am working with balanced firm-level panel data. I have run some regressions and know that I can't test the parallel trends because there is no information for the firm before the treatment.
That is observing 668 firms across 96 days. Over that 96 days, firms are either treated or not for the whole period.

Something I have across is a state-specific time trend. How would that relate to this? Thank you

[Taiba Chau]

Hi, I just wanted to add i wanted to analyse this running a two way fixed effects. So far I have run the regression:

regress movement i.date##i.treated,cluster(firm).

I also tried:
xtreg movement i.date##i.treated,fe

Both of these I similar estimates, except the constants are different. Also in the second regression the estimate for treated is omitted. I do not know why there is a difference between the two.

From here where could I go in terms of running fixed effects.

[Carlo Lazzaro]

Taiba:
please do not duplicate your posts (Two-way fixed effect).