Time Fixed Effects in a Regression Model

[Paul Baer]

Hello all,

I have an urgent problem. I have to do a linear Regression Analysis for my Thesis. I have performance Data for many Years but I need to use time fixed effects. I have like 30 years in total. Can someone explain me whats the point of the time fixed effects? Do I get like 30 different Values or do I get one time fixed effect which I can use for my regression?

Help would be really much appreciated.

[Carlo Lazzaro]

Paul:
welcome to this forum.
You have a panel dataset with N unreported and T=30.
I'm really surprised that your supervisor/teacher/mentor did not point you to any decent econometrics textbook, which covers fixed and random effect specification.
That said, you should consider -xtreg- or -xtgls- (possibly the latter is what you're looking for).
As per time fixed-effect, you may want to take a look at -fvvarlist-.
As an aside, it's often reminded on this list that crying for urgency is not the right appraoch to increase your chances of getting helpful replies.

[Paul Baer]

Thanks for the quick answer and I'm sorry writing urgent.
So I have 650 Deals and around 30 Years where the deals happened. Now I want to filter all the yearly effects out but I'm completely new to Stata and the whole vocabulary. I tried to understand it via econometric books but I don't understand it. The methods you wrote are they for eliminating all the effects and if yes how do I do it that I will eliminate all time effects at once. If this isn't possible how do I do it manually?

Many Thanks.

[Jingran Sun]

Paul,
I suppose what you have is actually time-series data rather than a panel data. When you say you have 650 deals, I would guess you there are 650 transactions occurred cross 30 years. If this is the case, then what you have is time-series data.

[Paul Baer]

But how do I filter the effects of each year out? I have to do a regression analysis but I have to filter the macroeconomic effects out.

[Paul Baer]

I could only find some information if I have one entry per year for a time series but not if i have sometimes 10 in the next year 20 and in the next only 5. Neither could I find how to eliminate the year effect as mentioned before.

[Phil Bromiley]

Paul - if you have multiple firms with multiple years per firm and want to have fixed effects for each firm, then you're in a panel world. If you have only one time series, then you're in a time series world. i.year in your xtreg will put in dummies for each year.

You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex.