BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Sequential Monte Carlo methods for estimating large scale volatili
 ty  matrices - CANCELLED (to be rearranged in Lent term) - Dr Kostas Trian
 tafyllopoulos\, University of Sheffield
DTSTART:20101201T141500Z
DTEND:20101201T151500Z
UID:TALK26960@talks.cam.ac.uk
CONTACT:Rachel Fogg
DESCRIPTION:During the recent credit crunch and the aftermath of the finan
 cial crisis it has become even more evident than before\, that there is a 
 need for accurate estimation of financial uncertainty. This is usually \nt
 ranslated into estimating the volatility or variance of associated financi
 al instruments\, such as asset returns. In particular\, there is increasin
 g interest in large data sets with the objective of identifying similariti
 es between asset returns\, pinpointing highly volatile assets and thus ide
 ntifying and measuring financial risk.\n\nIn this talk\, I will define a s
 uitable stochastic volatility model based on Wishart autoregressive proces
 ses. Such processes are natural models for estimating covariance matrices 
 and their theoretical \nbackground has only been recently developed. We pr
 opose a sequential Monte Carlo algorithm\, based on an auxiliary particle 
 filter. We discuss the problem of unknown parameter estimation in the part
 icle \nfilter\, and we provide modifications to existing procedures to dea
 l with the large number of such parameters. We address the problem of mode
 l comparison in this particular application by considering several alterna
 tive models. Our findings suggest that particle filters are very useful in
  this context\, overcoming most of the shortcomings of \nalternative model
 s.
LOCATION:LR10\, Engineering\, Department of
END:VEVENT
END:VCALENDAR