Actuarial Notes

Catastrophe Modelling: Guidance for Non-Catastrophe Modellers

By Lloyd's Market Association Exposure Management Working Group · Lloyd's Market Association · 2013 · CP312-106-25

Study guide: Exam CP-312

A June 2013 guide from the Lloyd's Market Association (LMA) Exposure Management Working Group for colleagues who do not work on catastrophe ("cat") modelling. It provides background on what a cat model is, explains the terminology frequently used by modellers and answers many frequently asked questions. It is not intended to be a comprehensive guide to cat modelling, but a quick reference giving non-cat modellers a basic understanding of the work of cat modellers.

On the syllabus

  • Exam CP-312 — the whole guide, for outcome 2a (Topic 2).

Introduction

  • Uses of Catastrophe Modelling
    • A cat model is a computerised system that generates a set of simulated events and estimates the magnitude, intensity and location of each event to determine the damage and the insured loss from a catastrophic event such as a hurricane or an earthquake, and the first core use the guide lists within the insurance industry is risk pricing.

Catastrophe Modelling practice

  • Different components of a Catastrophe model
    • A cat model is built up of a number of modules that must all operate in coordination, among them an exposure data module (the risks to be assessed, with at least a geo-location and an insured value) and the hazard and vulnerability modules, each of which could be considered a model in its own right.

Catastrophe Modelling Terminology

  • EP curve
  • OEP and an AEP curves
  • VaR and TVaR
  • Event Loss Table
  • Coefficient of Variation (CoV)
  • Difference between Near Term, Long Term and Historical rates
  • Difference between Ground Up, Gross, Net and Final Net losses

Frequently Asked Questions

  • Why is it that every time an event occurs I hear that it was not covered properly by the catastrophe models?
  • What is the impact of poor quality data on results?
  • Why do I need aggregates if I have a model?
  • What is a 1-in-250 return period?
  • Why do 1 in 100 year losses happen every few years?
  • What is an n-year event?
  • Why can you not add up return period losses?
  • What is Pure Premium?
  • How can I still get a loss to a layer when the mean loss is less than the attachment point?
  • Why is the 10,000 year loss in RMS not the worst case loss for this account or portfolio?

Sources