Page 1 of 1

near singular matrix erro

Posted: Sun Sep 07, 2014 12:36 am
by daan
Hello everyone,

I’m a student writing my thesis and in need of some assistance regarding Eviews 8.
When i’m running a regression my output is not correct (the coefficient is too high). When I try to fix it by adding a constant (c) I get a near “singular matrix error”. The same happens when I add a dummy for firms industry (eg dfinancials) but it think that is due to too few data points (I’m not completely sure).

Can some one see what I’m doing wrong?
In my equation and in my dataset (why are my coefficients so high?)

Thanks in advance,

ps i have added my dataset and my output.
in the output file i have added the equation below each output.

Re: near singular matrix erro

Posted: Sun Sep 07, 2014 3:12 am
by daan
Oke i found out that one of my dummies can't be in the same equation as the constant and when i rerun it, it works.
But, i still get these high coefficients does anyone know why that is?

Can it be that my data is scaled wrong? even though some of them are dummies and my dependent is a log value?

thanks