RMS’ modelling grows broader following Typhoon Jebi, says Asia director

Media reports of differences between modelled losses and actual claims following Typhoon Jebi reflect a misguided expectation that the modelled figures include factors not covered by the models factors that the re/insurers using the models know they need to consider.

RMS has now taken the opportunity to incorporate broader elements in its post-event industry loss estimate in order to improve understanding of modelled losses, JB Dumerc, North Asia sales director at RMS, told SIRC Today.

“The big learning we had from Jebi was that the media had a different understanding of the losses compared to clients, so after Jebi, when we had Typhoon Faxai, we factored additional non-modelled losses because it included things that we normally include post-event,” Dumerc said.

“We have added elements to the cat modelled losses. These elements are not necessary for the purpose of our customers but are more for the industry in general.”

These elements could include speciality risk, automobile, and business interruption as well as feedback from primary insurers. For US hurricane, RMS uses the real-time analysis provided by its HWind modelling product to generate predictions about the impact of an event while it is happening.

Drawing in information about wind speed, direction and other factors, it can predict the impact of a storm before it makes landfall.

“It means we can put this information into our model and that a few days in advance, we can tell customers what to expect.

“The reinsurer that knows its customers can propose additional capacity if they feel it is good for them,” Dumerc said.

“There is good value for the customer because they have 48 hours’ notice of something the rest of the industry doesn’t know.”

Another new RMS product, Site IQ, also makes use of real-time data this time to give underwriters detailed data on specific properties.

“It’s an underwriting tool to help the underwriters in the field, in real time, by enabling them to enter the address of a property and have the risk analysis of that property,” he said.

“We believe this will revolutionise the cat modelling industry, putting these services in the cloud on a subscription basis and giving re/insurers access to the science and data without needing the traditional cat modelling expertise,” he concluded.