EViews.com

Hey everybody,
i have an important question:
i have the spreads of 3 ETFs for each second for the time period from 1st of January till 31st of January 2005.
I also have the trading volume for each of the 3 ETFs for each second for the time period from 1st of January till 31st of January 2005.
I would like to investigate if the trading volume has an impact on the spread.
I already analyzed each ETF for itself. I uploaded for each ETF one workfile in eviews. It worked. But now i want to regress all ETFs together but i don't know how i can do it.
Could anybody describe me the way how i can do this please? Do I need to run an panel-regression?

What I have so far is an excel-sheet looking like this:
Date/Time ETF Spread trading volumen
1/1/2005 09:00:00 ETF1 0,5 5000
1/1/2005 09:00:01 ETF1 0,3 300
.
31/1/2005 17:30:00 ETF1 0,002 0
1/1/2005 09:00:00 ETF2 0,01 3000
.
.
31/1/2005 17:30 ETF3 0,1 5000

Thank you a lot.

Assuming you have an EViews workfile for each ETF (i.e. 3 workfiles), and now you want to do a joint estimation over all 3, just append the files together (proc->append to current page) to make one big workfile.

Thanks a lot for your answer. I already appended the files together but now i am not sure if it is ok when i just do a normal OLS estimation over all 3 (like i did it with one single ETF) or do i need to run a panel regression (because i have 3 ETFs and not just one)?
I very appreciate your effort.

That's an econometrics question, not an EViews question.

You almost certainly want a different constant for each ETF.