Best catastrophe risk modelling solution: RMS
Flooding presents a huge risk to re/insurers but has received comparatively little attention from modellers, compared with other perils such as hurricanes and earthquake. This is partly due to the technical challenges of understanding which properties are at risk: for instance, half a metre of difference in flood depth can make the difference between no loss and a disaster.
With the development of new technologies, the industry's approach to flood risk was rife for innovation and RMS' Europe Inland Flood HD (high definition) Model has provided something of a breakthrough to the industry.
The model, which has already been used by the industry for the 1 January European renewal, stood out for two reasons: first, its use of cloud computing which enables flood risk to be represented at a highly granular level, down to individual properties. Second, it is reportedly the first ever pan-European fully probabilistic flood model that covering 13 countries from the UK to Hungary, accounting for all sources of inland flooding. This is notable since river and weather systems span the continent, and floods do not respect national boundaries, so it allows reinsurers to analyse the accumulation of risks across the continent.
Furthermore, while primary models usually have access to high-resolution data, the resolution decreases as it is ceded to the reinsurer and along the risk management chain. However, the HD model includes an innovative exposure disaggregation component that delivers both expected loss at any location, and also the uncertainty associated with it.
Eager not to increase any technological burden, the RMS model runs on Microsoft Azure, meaning little or no local IT resource implications, scalable capacity and state-of-the-art, robust backup recovery.
New settings allow for sensitivity analysis and validation, and thus greater transparency. These include a user-defined hours clause, bespoke flood defences, and the ability to exclude pluvial risk. In addition, underwriting solutions are derived from the probabilistic model, ensuring consistency across the whole risk management chain. All these features are key to meet the requirements of Solvency II.
In the course of 2018, RMS will release two new probabilistic flood models, for the whole of US and whole of India.