Risk modellers believe insurers, reinsurers and rating agencies all need to be more sophisticated in their use of risk models. They also complain that people do not appreciate the differences between the models as well as they should.
At the World Insurance Forum, Karen Clark, CEO of AIR Worldwide, admitted her frustration about people lumping the three main risk modelling firms together when discussing how models performed last year. She pointed out that the different models give widely different outcomes.
Clark also thinks risk modellers should not be blamed for pricing of business. “Everyone has a quaint notion that price is based on the models, but the market sets the prices,” she said.
Rick Clinton, president of Eqecat, agreed. And he raised concerns that the rating agencies do not make a distinction between the different models. He said this may lead some companies to choose those models that give the lowest loss estimates for a particular exposure.
“There is a lot of naïve use of the models out there,” he said. “We would like to see rating agencies having a much greater understanding of models. They are different. But rating agencies don’t take this into account. Whatever model you use is okay with them. That leaves the potential for gaming the models.”
Clark also said that AIR will be providing its clients with shorter-term assessments of risk to take into account for this period of greater frequency and severity of hurricanes. But this is only a supplement to the longer-term view AIR’s models take.
“We are going to stick with providing long term but also give another catalogue that is short term,” she said. “You can decide how to do it. Basing pricing on the short term is going to be very difficult to get through the regulators. But a primary company may think: ‘I’m going to buy more reinsurance this year.’ Both approaches have uses. We don’t want to throw the baby out with the bathwater.”