School Seminars and Colloquia

Lévy-driven Stochastic Volatility and COGARCH Models

Joint Seminar Series on Stochastic Processes and Financial Mathematics

by Professor Peter Brockwell


Institution: Department of Statistics, Colorado State University
Date: Thu 21st April 2005
Time: 1:15 PM
Location: Room 213, Richard Berry Building

Abstract: In recent years various attempts have been made to capture the so-called
stylized features (e.g. tail heaviness, volatility clustering and
dependence without correlation) of financial time series using
continuous-time models. The interest in continuous-time models stems from
their use in modeling irregularly spaced data, their use in financial
applications such as option-pricing and the current wide-spread
availability of high-frequency data.

The applications of Lévy-driven continuous-time ARMA (CARNA) processes in
this context are discussed, with particular reference to their use in
extending the stochastic volatility model of Barndorff-Nielsen and
Shephard (J.R.S.S. (B), 2001) and constructing continuous-time GARCH(p,q)
processes which generalize the COGARCH(1,1) process of Klüppelberg,
Lindner and Maller (J.Appl.Prob., 2004).

For More Information: Prof Daniel Dufresne at dufresne@unimelb.edu.au, Dr Aihua Xia at xia@ms.unimelb.edu.au