The update to Verisk’s severe convective storm (SCS) model that was released in June has offered a fresh view of risk to the ILS market, bringing modelled outputs more in line with actual loss experience.
It is a development welcomed by market participants, with the expectation that it will enable more accurate underwriting and pricing, including easing placements for SCS-exposed cat bonds amongst a wary investor base.
The convergence of model expectations and historic loss experience is expected to enable more accurate underwriting and pricing, including easing placements for SCS-exposed cat bonds amongst a wary investor base.
“What we’ve seen in the last 12 to 24 months is bridging the gap between historical experience and modelled output – this often has been the catalyst to allow risk to trade in the capital markets,” said Mitchell Rosenberg, MD, co-global head of ILS, at Howden Tiger Capital Markets & Advisory.
One cat bond manager told Insurance Insider ILS that they had significantly underweighted SCS risk, given that the actual loss experience was considerably higher than modelled loss expectations.
The lack of model clarity, in addition to generally harder market conditions, has also led to greater use of per-event deductibles over franchise deductibles in cat bond structures, particularly for multi-peril aggregate bonds.
One source told this publication that the model update would support lower multiples on SCS-exposed cat bond deals on the one hand, and on the other would give investors more confidence in the risk they are buying.
The cat bond market has absorbed around $1.4bn of SCS-exposed aggregate bonds so far this year.
William Dubinsky, chairman at Gallagher Securities, said: “The cat bond and sidecar markets have an important role in supporting cedants and complementing reinsurance capacity in handling frequency risk.”
Andrew Siffert, SVP and senior meteorologist at BMS Group, added: “Verisk’s updated model arrives at a pivotal time, offering insurers enhanced tools to better capture the frequency, intensity and geographic spread of these events.”
Insurance industry losses from SCS tallied to around $60bn in 2024, according to an average of estimates from Aon, Gallagher Re and Swiss Re.
The 2024 average was slightly lower than in 2023 but higher than around $40bn in 2022, with the industry loss trend displaying some volatility but with a growth trajectory over time, sources said.
The top 10 most costly SCS events since 2020 have occurred in the past two years, according to data collated by Gallagher Re.
Dr Jay Guin, EVP and chief research officer at Verisk, said the firm’s model update elevates its view of SCS risk on a present-climate basis (over 10 years), pushing up annual aggregate industry losses by 50% to $55bn a year, from $30bn previously.
Verisk’s definition of SCS is of a “frequency peril”. It comprises the three sub-categories of straight-line winds, hailstorms and tornadoes.
A significant proportion of the increase is from hailstorms, which are showing frequency uplift, and where some evidence also indicates severity rises.
Hailstorms now contribute 75%-80% of the ~$55bn annual aggregate industry loss from SCS and the rest is straight-line winds and tornadoes. This compares to 50% of the total per the old model.
SCS also now contributes the largest component of annualised cat losses across all perils, according to Verisk. “Given all other perils – hurricane, earthquake, flood, wildfires – SCS is up to about 60% now. Every year, there are 30-40 big SCS events,” Dr Guin said.
The change in the view of SCS risk differs by state, with Texas more prone to hail, and the plains east of the Rocky Mountains also very susceptible.
There were also decreases in SCS modelled losses, for east coast states California, Oregon and Washington, and for some Northeast states.
There was little change in climatology of tornadoes; however, tornadic activity has expanded into the Southeast states of Louisiana, Mississippi and Alabama.
“The bullseye of tornadic activity has shifted over time from Oklahoma, so-called ‘Tornado Alley’, to the ‘Dixie Alley,’” Dr Guin said.
Hailstorms now contribute 75%-80% of the ~$55bn annual aggregate industry loss from SCS and the rest is straight-line winds and tornadoes. This compares to 50% of the total per the old model.
Separately, Moody’s RMS is also working to update its SCS model, rebuilding hazard components, event set and the vulnerability components for each of the sub-perils straight-line winds, tornadoes and hailstorms.
This will change the view of risk from an exposure standpoint and a hazard standpoint.
The update will take in new radar data that has become available, as well as trends and changes in size of property, property valuations, exposure growth and social inflation. It will draw on more than $50bn of company claims data from the past 13 years.
High-definition radar methodology is now available that can “express uncertainty characteristics of tornadoes”, said Julie Serakos, MD, Moody's model product management.
“Tornadoes being self-contained, super strong, very destructive, smaller footprints. That high-definition methodology really helps with decision making about where to allocate capital,” Serakos said.
Increased SCS frequency had made cat aggregate products difficult to place over the last several years, she noted.
“There is now more information about frequency and severity, helping reinsurers and ILS investors to make more informed decisions about how much capital to put against the risk,” she said.