S+ Finmetrics and STAMP comparison for multivariate models in state space form

[Poppy Akantziliotou <poppy.akantziliotou@initiativemedia.com>]

Dear S+ users,

I'm having some problems with S+ Finmetrics and particularly with SsfFit function. 

I wrote a piece of code (a small function) that fits a multivariate model in state space form, using the SSfit function and I wanted to compare my results with STAMP. For some datasets I get exactly the same results as the ones derived from STAMP but for some other datasets (and the same set of options in STAMP), I get different results.

Has anyone experienced the same problem before?

Any help would be much appreciated.


Poppy Akantziliotou