Garry Booth recently spoke with Karen Clark, chief executive of cat risk consultants Karen Clark & Co, about the launch of the industry-backed cat model framework Oasis. This is the full transcript of the interview
GB: Is the Oasis launch significant? Why does it matter?
KC: The Oasis initiative is very significant. They have 22 sponsors, which demonstrates strong support for the idea. That’s because it promises three important things that most insurers and reinsurers are looking for in cat models: greater transparency, efficiency, and cost effectiveness.
Oasis isn’t the only platform that’s designed to deliver these benefits of course. KCC’s RiskInsight® is already being used by a dozen companies. And it is already delivering on all three counts.
So we’re well aware that there is great demand for alternatives to the status quo.
Is it about cost or transparency or the hegemony of the Big Three?
The established vendor models are getting more and more expensive and everyone knows that these models only provide very rough estimates of losses. Insurers and reinsurers are questioning why they are now paying millions of dollars in licence fees for the subjective judgments of scientists at only three firms! They know that they can now obtain more, arguably better, knowledge more transparently, at a lower cost.
It just isn’t healthy to have all the decisions based on three modelling companies – especially when they are black boxes. You can’t look at the assumptions. There is systemic risk in the industry relying on a small number of black box models.
The Big Three do not have a lock on knowledge: most of the scientific information that underlies their models is publicly available.
The benefit/cost ratio for cat models has gone out of whack. That’s why the time is right for open platforms like Oasis and RiskInsight. Greater choice will lead to lower cost.
Do insurers and reinsurers have the resources to exploit the claims made for Oasis?
The Oasis strategy is to provide an open marketplace for models from different agencies. But users need to be sufficiently well resourced to examine and choose from the available models in Oasis. So in that sense, looking at say 30 models is more complicated than looking at three models.
RiskInsight’s strategy is different. We start users on baseline models where it is possible to see all the assumptions, all the events and intensities as well as the damage functions. Because all the fundamental assumptions are ‘visible’ users can customise the assumptions in order to obtain their own view of risk, without devoting a lot more resources to the task.
(We are discussing integrating some of RiskInsight’s tools into Oasis to help make it easier for companies to use that platform.)
Sophisticated users can customize their own models using RiskInsight, their own events and their own damage functions. Very sophisticated users can use it to build models for ‘non-modelled’ perils--those not covered by available cat models.
RiskInsight currently has US wind and we’re about to release US earthquake and storm surge flooding baseline models. We will soon have European windstorm and Japan windstorm and earthquake baseline models.
What’s your reaction to RMS(One)
RiskInsight, Oasis, and RMS(One) are all multi-model platforms—the differences are in the implementation strategies and the cost structures. RMS(One) is likely to be the most costly by far so insurers and reinsurers using that platform could be at a significant expense disadvantage relative to companies adopting the alternative, open platforms. It’s also not clear how RMS(One) will deliver greater transparency. Because of the growing concerns over the cost and business risk of locking important risk management processes into one platform, many companies are actively evaluating all of the new options and will no doubt ultimately adopt the most transparent and cost effective solutions.