Fixed effects coefficient interpretation and binary dependent variable

[Magnus Justad]

Dear all,
My question is two-fold:

First, I have currently run a fixed effects model using the xtreg, fe command. I wonder if anyone can tell me a bit more about the coefficient interpretations (no bindary dependent variable here)?
From what I understand it goes as follows: a 1 unit increase in X leads to a bX unit increase/decrease in Y. Hence, if my independent variable is -0.5, a 1 unit increase in X leads to a -0.5 decrease in Y. Is this the right intuition?

Secondly, how would you interpret the coefficients if I would use a binary (1,0) independent variable in the fixed effects model?

I hope anyone could shed some light on this as I have found very little information on this.

Thanks in advance!

Best,
Magnus Justad

[Carlo Lazzaro]

Magnus:
helpful replies may stem from seeing what you typed and what Stata gave you back (as per FAQ).
In this respect, pasting codes and results of your Stata session via code delimiters (there's a FAQ on this topic) is an advisable approach. Thanks.

[Magnus Justad]

Dear Carlo,

Thanks for the comment. So to begin with, I have set firm identifiers as the entities and year as the time variable: "xtset firmidentifier year"
I have then run my fixed effect regression: My dependent variable is transaction value of a seasoned equity offering, and my independent variable (of interest) is common shares outstanding.

"xtreg tv_seo commonshares, fe vce(robust)"

This gives me a coefficient of commonshares of -0.02. Hence, does this means that when common shares outstanding increases by one unit, deacrease transaction value of a seasoned equity by -0.02?

Then back to my second question earlier: Could I run the fixed effects regression with a binary dependent value (1=engages in a seasoned equity offering, 0=do not engage), and if so how would this coefficient be interpreted in that case?

Thank you in advance!

Best,
Magnus