BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Signal Processing and Communications Lab Seminars
SUMMARY:Scalable inference for a full multivariate stochas
tic volatility model - Prof. Petros Dellaportas\,
Dept. of Statistical Science\, UCL
DTSTART;TZID=Europe/London:20160421T140000
DTEND;TZID=Europe/London:20160421T150000
UID:TALK65760AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/65760
DESCRIPTION:We introduce a multivariate stochastic volatility
model for asset returns that imposes no restrictio
ns to the structure of the volatility matrix and t
reats all its elements as functions of latent stoc
hastic processes. When the number of assets is pro
hibitively large\, we propose a factor multivariat
e stochastic volatility model in which the varianc
es and correlations of the factors evolve stochast
ically over time. Inference is achieved via a care
fully designed feasible andscalable Markov chain M
onte Carlo algorithm that combines two computation
ally important ingredients: it utilizes invariant
to the prior Metropolis proposal densities for sim
ultaneously updating all latent paths and has quad
ratic\, rather than cubic\, computational complexi
ty when evaluating the multivariate normal densiti
es required. We apply our modelling and computatio
nal methodology to 571 stock daily returns of Euro
STOXX index for data over a period of 10 years.
LOCATION:LR11\, Department of Engineering
CONTACT:Prof. Ramji Venkataramanan
END:VEVENT
END:VCALENDAR