Here to stay: Amwins doesn’t expect a soft market any time soon

Even without significant losses from storms, the hard market is likely to persist in the near time. What might change is re/insurers’ assumptions—and their reliance on models.

When Amwins published its Q2/Q3 2022 State of the Market report in April, one theme impacting the property market was consistent: natural catastrophe losses. Last year was the second-costliest on record after 2017, it noted.

“Although the severe storms cannot be directly attributed to climate change, climatologists at the large insurers believe this is a pattern we are likely to see continue,” said the report.

That’s undoubtedly how it feels to Mark Bernacki, the global specialty insurance distributor’s first chief underwriting officer, appointed in September 2021 to oversee the underwriting performance of its delegated authority business. He joined the company’s senior vice president of actuarial services, Chris Platania, to discuss its view of nat cat risk with Intelligent Insurer.

As Bernacki noted, losses in recent years have been extraordinary in both frequency and severity.

“It seems that every year there’s an unexpected event,” he said. “In the industry now, we almost have to expect the unexpected.”

That’s at least partly driven by climate change. According to Platania, more frequent and severe hurricanes such as Harvey, Irma and most recently Ida, and other risks, have come to the fore. California and Texas are becoming wildfire “hotspots”, he notes, and the threat has rapidly risen up the agenda.

“If you look at wildfire claims over the last 15 years, eight of the 10 largest wildfires have been since 2017,” he said.

Established risks are also impacting new areas. Oklahoma’s Tornado Alley, for example, seems to be reaching over to Tennessee and Nashville.

“There could have been a bunch of tornadoes before, but in the middle of a cornfield they didn’t hit anything,” said Platania. “Now they’re starting to shift more in urban centres where there is a much higher propensity for loss.”

In the same way, it’s not just severity or frequency causing greater losses from hurricanes. The changing activity has been accompanied by social changes exacerbating losses. In recent years, Florida and Texas have had some of the highest population growth rates, for example, increasing exposure in those areas.

According to Bernacki, in the Florida market, regulatory challenges around valuations, building ordinances and follow-ups and inspections have added to the difficulties for re/insurers.

“In Louisiana, we’ve seen a firming and hardening market as a direct result of loss activity in the region from the frequency and severity perspective,” he said. “Florida, on the other hand, hasn’t had as much direct loss activity, but is probably harder, even without it.”

A renewed focus on models

Whatever the causes, however, there are some consequences. One is a hard market that, for Bernacki, is unprecedented.

“In my 30-year career, this is the longest, most sustainable, most stable hard market I’ve witnessed,” he said. It will continue, he predicts, for another 12 to 18 months at least: “until we see a sustained reasonable margin made by the risk-takers”. And that’s without a big storm in the next couple of years.

“If there is a storm, then all bets are off, and it will stay this way for a long time,” added Platania. Even without it, ongoing concern over climate change, regulatory and other issues will mean rates are more likely to flatten than fall. Capacity is unlikely to rush back.

“It’s not going to be everybody’s desire to write a bunch of Florida business next year just because we didn’t have a storm this year,” said Platania.

“I don’t think we will go into a soft market. It will just level.”

The other may be a re-evaluation of the dependence on models. According to Bernacki, if cat losses keep surprising on the upside, it’s worth reviewing how expectations are being set.

“To a certain degree, we become overly reliant on the cat models as a perfect predictor of what the storm activity will be,” he said. “When these models were originally rolled out, they were used as one of many tools from an underwriting risk quantification perspective.”

Now rating agencies, boards of directors and others rely on them for their view of nat cat risk. That’s also perhaps due to the entry of investors and others without underwriting expertise through insurance-linked securities.

That’s not going to mean models will be abandoned. The models are imperfect, said Platania, but they are “the best we have”. However, a couple of things are likely to happen in addition to a continuing hard market.

First, there will be continuing refinement of the cat models as the modelling firms try to reflect the changes being seen. There is also likely to be a renewed focus on the data used with the models. Ensuring accurate valuations, for instance, to avoid underpricing is crucial.

“We’re trying to get the best data because while the models are imperfect, the better data you have, the more confidence you can have in the output,” said Platania.

As Bernacki put it: “Unless we have proper valuation, the model breaks before it even begins.”