Impact Forecasting, the catastrophe model development center of excellence at Aon Benfield, has launched its new European windstorm model. The model, which has been in development for the last three years, “offers an alternative view on this peak zone peril and enables insurers to gain a better understanding of the possible losses, including the effects of storm clustering,” said the announcement.
“Windstorm is the number one peril in Europe, causing the highest annual insured losses. The potential for windstorm footprints to extend into multiple countries can result in high aggregate losses for insurers with multi-country portfolios and shows the need to model this hazard consistently on a Europe-wide basis. For example, Xynthia, which caused €1.6 billion [$2.195 billion] in insured losses, impacted eight countries and the model suggests the potential for more extreme events many multiples larger.”
In addition, Impact Forecasting pointed out that “Solvency II is emphasizing the need for a deeper understanding of the model solutions that help drive insurers’ views of risk. For some firms this increases a desire to own a catastrophe model in-house. Impact Forecasting’s suite of models – including the new European windstorm tool – can provide a fully transparent and documented approach that delivers on this requirement.”
The windstorm model currently covers Belgium, France, Denmark, Germany, Ireland, Luxemburg, Netherlands and the UK. The development involved three years of collaborative research with the department of meteorology and climatology at the University of Cologne, an Aon Benfield Research partner, to incorporate the latest peer reviewed scientific research into the model.
Differentiating features of the model and benefits for insurers include the following:
– Incorporating the latest insights from climate research to better understand the maximum possible loss from a single windstorm event or a clustering of events on an annual basis that may impact reinsurance protection under the treaty.
– Allows customized modelling based on an individual insurer’s portfolio, for example using insurance claims to modify the damage component of the model. This achieves a better representation of the insurer’s risk.
– New events can be added to the model to forecast losses for ongoing or hypothetical events.
Adam Podlaha, global head of Impact Forecasting, said: “Windstorm is the most damaging peril in Europe so bringing an additional view on the risk – and new insight into the financial impact of clustering – is invaluable to the market.
“In line with the proposed Solvency II regulatory framework, Impact Forecasting’s transparent approach to catastrophe modelling means insurers can develop and truly own their view of risk. The model runs on our open architecture modelling platform ELEMENTSwhich, in response to insurers’ evolving needs, delivers the capabilities to customize individual model components or quantify uncertainty.”
Alexandros Georgiadis, climatologist and catastrophe model developer at Impact Forecasting, explained: “Due to short historical records and the comparatively smaller number of annual significant European windstorms – 2.5 per year on long term average – it remains challenging to quantify the frequency of historical extreme events, and thus of seasonal clustering, from observations alone. In order to solve this, a large set of global climate models (GCM) simulations, run with set ups consistent with current climate conditions, are explored to strongly enlarge the data sample to over 4600 years and thus to provide better estimates of event-based and cumulative losses. Our approach preserves the clustering as simulated by the GCM, thus enabling a more accurate and physically-based definition of loss clustering and frequency.”
The next phase of model development will see the addition of Central Europe and Scandinavia. The windstorm model builds on Impact Forecasting’s European model suite including flood and earthquake.
Source: Aon Benfield/Impact Forecasting