While catastrophe vendor models have an important place in the insurance and reinsurance industry, the influence of climate change suggests it’s essential to leverage more physical exposure based models, according to industry experts.
On the opening day of the virtually held 2020 ILS Bermuda Convergence event, climate change took centre stage as speakers from across the insurance, reinsurance and insurance-linked securities (ILS) sectors discussed its influence on various parts of the market.
The second panel of the day was moderated by David Flandro, Managing Director, Analytics at Hyperion X, and featured Peter Dailey, Vice President of RMS; Craig Tillman, President, Natural Hazards Risk, Portfolio Management, Risk Mitigation, RenRe Risk Sciences; and Lixin Zeng, Managing Partner at Integral ILS.
During the session, vendor models were discussed at length and while it was agreed that they are getting better, the audience heard how the effects of climate change intensify the need for change and more transparency.
“Our industry needs to recognise that global climate change is now changing our view of the risk, and it’s happening at a very rapid pace, much more rapid than we would have expected 10 years ago,” said Tillman. “So, we’re trying to understand and represent this very important non-stationary driver as a critical parameterisation of the models that we use, forward thinking models.
“Where models only look at historical data to guide, it needs to be recognised that it’s much less relevant to our work now. We’re in a new and changing regime and that makes the physical models that we have access to now much more important characterising the future risks.
“And, importantly, that characterisation of how much increased frequency or how much disparity we might see in the future.”
As noted by Tillman, the climate variability and the weather conditions witnessed over the past decade or even the next 50 years, are very different from what was seen in the 1900s. And, in his view, ILS managers and reinsurers need to understand this issue on a regional peril basis.
“We need to evolve an understanding that had a reasonable timeframe, 10 years out, 30 years out, 40 years out, to create this ramp up on a region peril specific basis to understand how climate change is going to affect that peril.
“That gives us all very good intelligence as to how to augment vendor models in a way that makes our view of the now, a little bit more robust.
“To me, that’s the only way forward to do this sensibly, and still end up with a rational price, not one that is looking 50 years out and not really being relevant to today’s risk.”
The panel agreed that this move is required, while the definite and important place of models within the industry was also highlighted. Additionally, speakers noted the need for continued disruption and innovation in the modelling space, alongside the benefits of an independent peer review to ensure the market is getting the best out of the models available.
“I think that a very important criteria as we go forward, is to make sure that we’re using models, wherever they come from, that allow us to take our own science views of how the risk is changing due to the climate change and kind of input those into the models and see what the results are.
“There’s so many good things about the models, you just want them to be a little bit more open and useful,” said Tillman.
