The insurance sector is quite transversal and interacts with many other economic sectors such as agriculture, energy, health, transportation, etc. Risks managed by the insurance sector are related to weather conditions, particularly extreme climate events, which are the main drivers of important economic losses. Applying High Performance Computing (HPC), we produce climate predictions that give the probability of occurrence of extreme rainfalls, droughts or on the number of hurricanes expected for the next season. This information helps evaluating the risks underwritten by the insurance sector and is useful as early-warning system for society and the business sector in general.
Floods, droughts and cyclones are weather phenomena that cause huge social and economic damage, and the frequency and magnitude of these events are likely to be impacted by increasing concentration of green-house gases in the atmosphere. Climate models can improve our understanding of future extreme events and this information can be very useful for the insurance sector regarding the following applications:
- Future loss estimation
- Pricing determination and capital allocation
Hurricane activity varies significantly from year-to-year due to the climate controls (i.e., ENSO) that suppress or enhance hurricane activity over a given season. The insurance and re/insurance sectors are therefore exposed to a risk that varies considerably from one year to the other. Due to regulatory constraints, large price fluctuations in prices, which could reflect the underlying risk, are not allowed. Thus, applications of multi-annual predictions (typically 5 years) to the insurance sector allow the spreading of risk over multiple seasons and produce stable prices for the coverage offered.