I am trying to estimate a univariate TARCH (-X) model by Maximum Likelihood
from about 4000 data points using Excel and its Solver. And... It seems to
work !
Still, I would like to know :
1) Is Excel solver really trustworthy on such an issue (or more generally
for quite complex estimation problem) ? The reason I am asking is that I
have many times heard concerning similar issues "Yeah, true, Excel can do it
*sniggering* , but it was not designed for it, so you better not use it for
this".
My problem is I have weak numerical analysis and programming skills.
Besides, Excel is extremely convenient for manipulating data extensively and
quickly. So i'd prefer to stick with Excel unless it really cannot handle
such problems (I don't care for waiting 5 seconds).
However, I could deal with a little bit of numerical analysis if I had some
clear guide (say with matlab, or maybe C++...or Excel) which brings me to my
second question :
2) Would you know of some link(s) with really synthetic papers/courses
covering *estimation*, *testing*, and *goodness of fit*, including
computational aspects, on univariate TARCH-X model (or maybe just GARCH) ?
(I am really not looking for very theoretical papers on precisely one kind
of aspect of GARCH models)
Thank you for reading.
David B
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
