Determining Extreme Values for the Insurance Sector

Several industrial sectors require knowledge of climate extremes above and beyond those experienced during the satellite era. Examples include the insurance industry where regulators will soon require adequate capital to cover a 1 in 200 year event; and the nuclear industry which is required to prepare against a 1 in 10,000 year event. The common approach to determine these is to extrapolate from observations over the past 50 years or so, either directly or through the use of a catastrophe model. Often these estimates do not include a quantification of the uncertainties arising from either natural climate variability or climate change. This project aims to determine the risks that are missed by these standard practices.

The direct results of this work will be a short report distributed freely within the insurance industry (either through Standard & Poors or Lloyds). This report will be drafted with the assistance of the commercial partners, to prevent it being overly academic. It will summarise the research performed during this feasibility and highlight the need to better quantify the uncertainty in catastrophe modelling.