Rapidly bounding the exceedance probabilities of high aggregate losses

Isabella Gollini*, Jonathan Rougier

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)

1 Citation (Scopus)
252 Downloads (Pure)

Abstract

We consider the task of assessing the right-hand tail of an insurer’s loss distribution for some specified period, such as a year. We present and analyze six different approaches: four upper bounds and two approximations. We examine these approaches under a variety of conditions, using a large event loss table for US hurricanes. For its combination of tightness and computational speed, we favor the moment bound. We also consider the appropriate size of Monte Carlo simulations and the imposition of a cap on single-event losses. We strongly favor the Gamma distribution as a flexible model for single-event losses, because of its tractable form in all of the methods we analyze, its generalizability and the ease with which a cap on losses can be incorporated into it.

Original languageEnglish
Pages (from-to)97-116
Number of pages20
JournalJournal of Operational Risk
Volume11
Issue number3
DOIs
Publication statusPublished - 1 Sep 2016

Keywords

  • Catastrophe modeling
  • Compound Poisson process
  • Event loss table (ELT)
  • Moment bound
  • Monte Carlo simulation

Fingerprint Dive into the research topics of 'Rapidly bounding the exceedance probabilities of high aggregate losses'. Together they form a unique fingerprint.

Cite this