Efficient simulation of gamma and variance-gamma processes

Avramidis, Athanassios.N., L'Ecuyer, Pierre and Tremblay, Pierre-Alexandre (2004) Efficient simulation of gamma and variance-gamma processes. 2003 Winter Simulation Conference Dec 2003. Institute of Electrical and Electronics Engineers Inc, 319-326.


[img] PDF
Download (179Kb)


We study algorithms for sampling discrete-time paths of a gamma process
and a variance gamma process, defined as a Brownian process with random
time change obeying a gamma process.
The attractive feature of the algorithms is that increments of the processes
over longer time scales are assigned to the first sampling coordinates.
The algorithms
are based on having in explicit form the process' conditional distributions,
are similar in spirit to the Brownian bridge sampling algorithms proposed
for financial Monte Carlo,
and synergize with quasi-Monte Carlo techniques for efficiency improvement.
We compare the variance and efficiency of ordinary Monte Carlo and
quasi-Monte Carlo for an example of
financial option pricing with the variance-gamma model,
taken from \cite{fMAD98a}.

Item Type: Conference or Workshop Item (UNSPECIFIED)
ISBNs: 0780381319 (hardback)
Related URLs:
Keywords: Brownian bridge sampling algorithms Brownian process conditional distributions discrete-time path sampling efficiency improvement financial Monte Carlo financial option pricing gamma processes numerical integration process increments quasiMonte Carlo techniques, random time change, sampling coordinates, simulation, time scales, variance-gamma model, variance-gamma processes
Subjects: Q Science > QA Mathematics
Q Science > QC Physics
Divisions : University Structure - Pre August 2011 > School of Mathematics > Operational Research
ePrint ID: 55793
Accepted Date and Publication Date:
1 January 2004Published
Date Deposited: 06 Aug 2008
Last Modified: 31 Mar 2016 12:36
URI: http://eprints.soton.ac.uk/id/eprint/55793

Actions (login required)

View Item View Item

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics