Public Services

Catastrophe models: Only as good as weakest link

<< Back to overview

Natural disasters wreak havoc on the communities they strike. I know this firsthand from growing up near New Orleans. I watched as trees were snapped like toothpicks when Hurricane Andrew hit in 1992. Years later, my uncle lost his house when Hurricane Katrina ravaged the Gulf Coast in 2005.

Over the years I’ve watched with great interest as our predictive technologies have grown more sophisticated. Minimizing loss of life and property requires more accurately forecasting natural catastrophes and responding to them more effectively. The reinsurance industry in particular invests enormous resources in devising predictive algorithms and ever more complex models to determine the likelihood and extent of damages to a given region. GfK’s Geomarketing solution area supplies this industry with the global boundary data at the heart of these models.

Transforming information into insights
Natural catastrophe modelers are incorporating ever greater quantities of data into their models. The goal is to take into account all variables that relate to the behavior and repercussions of disaster events. In this era of big data, enormous amounts of information can now be compiled and funneled into these models. But even the most all-encompassing models with the most eloquent algorithms are useless without good foundational boundary data.

FIG 2_postcode

For years now, GfK’s Geomarketing solution area has been supplying the industry with this boundary data, which is usually in the form of CRESTA zones and postcode boundaries. Why is this boundary data so important? Because it is the means by which modelers and others in the reinsurance industry geographically link, analyze, aggregate and display the mountains of information with which they are working. Almost all exposure data—asset locations, premiums, loss sums, past claims, flood plain distributions, etc.—has a geographic component such as coordinates or a postcode. This makes it possible to link this information to the polygons formed by boundary units such as CRESTA zones.

These variables can then be analyzed and displayed cartographically. These visualizations quickly reveal trends and relationships in the data that would not otherwise be apparent. This allows modelers and re-insurers to make more accurate predictions and understand the unique regional characteristics of their markets.

FIG 3_Typhoon Haiyan

Managing risk with GfK boundary data

GfK continues to build its partnerships with key players in the reinsurance industry, such as Aon Benfield. The Impact Forecasting division of Aon Benfield develops risk models for catastrophes related to earthquakes, hurricanes, wind storms, floods and other disaster events. These models rely heavily on CRESTA zones and postcode boundaries from GfK’s Geomarketing solution area. Two recent examples are Aon Benfield’s South African hail model and pan-European windstorm model. GfK offers the largest collection of worldwide boundary data available on the market, and is also the official provider of the CRESTA zones. GfK’s boundary data is regularly updated and features detailed, pristinely rendered coastlines, which is an essential feature in the reinsurance industry, as many high-value assets are concentrated in these areas. Additionally, the boundaries offer comprehensive coverage, without any gaps or overlaps, and the nested structure of postcodes and CRESTA zones makes it possible to aggregate data at varying levels of detail. This gives modelers and other users in the reinsurance industry a robust and flexible structure for linking, analyzing and sharing information.

Natural disasters will continue to challenge communities around the globe, but GfK’s boundary data helps the insurance and reinsurance industries predict and manage these events more effectively.

Get more insights

GfK will showcase how its boundary data benefits the reinsurance industry in an Intelligent Insurer-hosted webinar on September 10, 2015, from 3-4pm London time (BST). Titled “Improving exposure data quality and usability with GfK boundary data,” the webinar features a use case by ImageCat, a GfK partner and leading provider of risk and disaster management technologies. You can register for the webinar here.

For more on our offerings, visit our Geomarketing page or contact christopher.guider@gfk.com.

 

 

<< Back to overview

Share your thoughts

Leave a Reply

Your email address will not be published. Required fields are marked *