VaR for Heavy Tailed and FATGBM Models.
by
Khanhav Au
Centre for Financial Mathematics, Australian National University, Canberra.

Considerable empirical evidence suggests that tails of returns distributions are commonly of power type. Certainly they are much heavier than normal and results which have been calculated on this assumption (such as a notable letter written to the shareholders of the ill-fated company Long Term Capital Management) need significant adjustment.

In this talk we will discuss the effect of tail index, and other features, such as self-similarity scaling in the FATGBM model of C.C. Heyde, on the magnitude of VaR estimates. Contrasts will be made with results based on light tailed distributions, and also those based on the assumption that stable distributions provide an appropriate model as is propounded in the recent book "Stable Paretian Models in Finance" by S. Mittnik and S. Rachev, Wiley (2000). This is joint work with C.C. Heyde.

Date received: September 14, 2001

Copyright © 2001 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cagd-97.