Willis Re has come up with a Japan tsunami model, something it asserts is a first-time milestone for an extreme weather event whose resulting losses have been hard for the industry to understand.
In the making since 2010, the model is the culmination of work by Willis Re, Willis Research Network, Tohuko University in Japan and University College London in the U.K, according to the announcement.
Willis Re’s model blends tsunami loss information with earthquake shaking damage output. The reinsurer asserts that it will help it, and clients, both quantify and manage risk from tsunami events.
“By combining academic excellence with in-house analytics and market expertise we have developed an independent model [while] also being in a unique position to analyze the new tools that are now emerging from model vendors,” Rowan Douglas, chairman of the Willis Research Network, said in a prepared statement.
William Thompson, regional director for Willis Re Japan, said the new tsunami model should fill a gap in the cat modeling industry that has existed for a long time.
“Japan earthquake risk has been rigorously investigated and modeled but the complexity of modeling tsunami has led to a significant gap in the industry’s ability to quantify risk for severe earthquake events,” Thomson said in a statement. “The Willis Japan Tsunami Model is another vital step towards closing this gap and providing Willis Re clients with the most comprehensive view of risk.”
Tsunamis can cause severe losses and are touched off by earthquakes, often under water. Japan’s Tohuku region, for example, suffered a devastating tsunami in early 2011, triggered by a submarine earthquake. Similar factors set off a devastating tsunami that struck Indonesia in 2004.
Willis Re bills the new model is the first tsunami model of its kind “with probabilistic and deterministic modeling functions to be created by a broker.”
Both Willis Re and Willis Research Network are part of Willis Group Holdings, a global risk adviser, insurance and reinsurance broker.
Source: Willis Re