New flood risk model for Germany
Swiss Re uses ZÜRS zones to model flood risk in Germany
Just in time for the 2006/07 renewals
In 2001, the ZÜRS zoning system for floods, backwater and heavy rainfall was made available to German insurers for the first time. It enables insurers to identify the flood risk of individual locations in detail and to substantially improve underwriting for flood risk. Previously, estimating the accumulation risk of entire portfolios presented major uncertainties. There were no data records to indicate which locations were likely to be hit by floods at the same time.
To improve this situation, leading reinsurers joined to commission a computation of so-called probabilistic flood events. This is done by modelling events featuring the same statistical characteristics as the catalogue of historical flood events, but projected to span a period of several thousand years. This includes a direct link between the ZÜRS zones and probabilistic events.
The event database which has been available since spring 2006 is now integrated in MultiSNAP – Swiss Re's own platform for assessing natural perils exposures. Swiss Re is currently the only reinsurer to have adopted this new approach for calculating flood accumulation events available for the 2006/2007 renewal season.
Input data can be aggregated into five-digit postal code areas and correlated with high-definition geographic coordinates. Data aggregated on the five-digit postal codes should be used for modelling large-scale retail business with personal lines and commercial risks. The distribution of the risks over the various ZÜRS zones within the postal code areas can be considered, if they are known. Industrial business and smaller portfolios can be modelled in detail based on the precise geographic coordinates of each individual risk. If the geographic coordinates are not known, they can be determined from the full address of the risk in a preceding geocoding process.
This new flood model is a major step toward a probabilistic tool containing data on more than 10 000 events and away from conventional scenario-based approaches. It is now possible to estimate the extent and frequency of losses due to flood events throughout Germany. This makes it easier to identify suitable reinsurance structures for flood risks and to calculate the commensurate risk premium more reliably.
It is now in the insurers' own interest to provide high-quality data as the basis for tailoring their own customised reinsurance solution for the flood peril in Germany in cooperation with Swiss Re.
For more information on the probabilistic flooding model for Germany, refer to:
Contacts
Swiss Re Germany
Swiss Re Zürich
German version
Überschwemmungsmodell für Deutschland
Related links
