Page 1 of 1

Dummy Variables for testing January effect

Posted: Fri Mar 20, 2009 10:19 am
by walajee
Hi All,

I am new to Eviews and am begining to learn what this software can do.
I'm sorry if I ask a dumb question.
Currently I need to test january effect on the stock market and so far I've gathered FTSE100 index data from 1984 to 2008.
From the data, I've sorted it and stored the daily return in an eview work file.
Daily return calculation: log(P/Pt-1)

My model to test January effect:
Return = B1D1 + B2D2 ... B12D12

D1-D12 are dumy variables representing january to december.
B1-B12 are representing the estimate of a mean return from Jan to December.

The problem is how do I put this into eview. I believe I need to create 12 dumy variables for each month.
It would be good to have some guidance.

Thanks,
walajee

Re: Dummy Variables for testing January effect

Posted: Fri Mar 20, 2009 10:28 am
by EViews Gareth
Assuming you have your data in a dated workfile (daily, weekly or monthly), then you can use the @month keyword along with @expand to automatically generate the dummies.


For example, say I want a dummy for each month and I want to see the effect each month has on the coefficient of X, then I could do the following regression:

LS Y C (@expand(@month))*X

Re: Dummy Variables for testing January effect

Posted: Sat Apr 18, 2009 6:16 am
by walajee
A long study free holiday due to my full time job.

Thanks for your comment and let me look into it.

Again, thanks,

walajee

Re: Dummy Variables for testing January effect

Posted: Tue Mar 26, 2019 3:52 pm
by pavankataria
What is this x you speak of?