Message-ID: <d0f55a670704270000u53a12a06o5ee4b617ff1b135c@mail.gmail.com>
Date: 2007-04-27T07:00:25Z
From: Thomas Steiner
Subject: Cox, Ingersoll, Ross/Vasicek parameter estimation via Kalman-Filter (SSPIR)
In-Reply-To: <E1HhC2F-000072-00@smtp06.web.de>
Arne,
together with Martin Keller-Ressel we did estimations of the Vasicek
and CIR termstructure, based on today's yield curve (so this is
different from your approach). We used R for the estimation. Let me
know if you are further interested (eg in some code).
We recognized that the yield curves of these (and every other affine
model) models do just allow for three different shapes, even if you
allow for jumps. You can get the preprint at
http://www.fam.tuwien.ac.at/~thomas/getfile.php?file=yieldcurves
Thomas