## CAPM

For requesting general information about EViews, sharing your own tips and tricks, and information on EViews training or guides.

Moderators: EViews Gareth, EViews Moderator

halloduda
Posts: 3
Joined: Wed Jun 02, 2010 10:46 am

### CAPM

Hello,

I working with Eviews for a paper I write in my Econometrics class.. And I don't know, somehow my CAPM regression is not working! Maybe somebody could help me out with that! I've been studying Eviews by myself for 2 weeks now, but I must have overseen one thing..
Ok, here is my problem:

I have got data from a market return (e.g. S&P500), the data is like 0,001 or -0,0002 or so..
I also have got data from a fund (e.g. hedge or mutual fund), the data is also plus minus 0
And i have got risk free data (e.g. treasury bill), the data is also around 0 (like +- 3 percent)

... ok, so if I plug that data in eviews, it just gives me crap! Like the R^2 is aroun 0,0001, all coeffients are incignificant! I tried changing the data but it just does not work! Well, maybe somebody could give me a suggestion!

Thanks,
halloduda

halloduda
Posts: 3
Joined: Wed Jun 02, 2010 10:46 am

### Re: CAPM

I forgot to mention: I am an Eview7 user

JimForest
Posts: 80
Joined: Thu Oct 16, 2008 7:53 pm
Location: MA

### Re: CAPM

A low R-squared in a CAPM regression can be an indicator that your market index may not be appropriate as a benchmark for systematic risk. For example, if you are calculating beta for a bond fund you might want to use a bond index rather than a stock index.

halloduda
Posts: 3
Joined: Wed Jun 02, 2010 10:46 am

### Re: CAPM

Thanks for your answer! That is what I tried during the last days.

For example: I have got S&P500 daily returns and than I try to regress a American stock fund (with daily returns)! And that's where the problems start.. The correlation coefficient is like -0,01 and R^2 like 0,000000001

Do you might have another idea?

startz
Non-normality and collinearity are NOT problems!
Posts: 3520
Joined: Wed Sep 17, 2008 2:25 pm

### Re: CAPM

halloduda wrote:Thanks for your answer! That is what I tried during the last days.

For example: I have got S&P500 daily returns and than I try to regress a American stock fund (with daily returns)! And that's where the problems start.. The correlation coefficient is like -0,01 and R^2 like 0,000000001

Do you might have another idea?

How long is your sample? Something like this might happen over a very short sample, I suppose.
Less helpfully, I wonder if you've just made a mistake somewhere. The R^2 should be the square of the correlation coefficient and what you report isn't close.

### Who is online

Users browsing this forum: No registered users and 1 guest