This article was created in partnership with ICAT.
Growing volatility in hurricane-related losses is forcing insurers and MGAs to evolve their catastrophe risk strategies. According to the Office for Coastal Management, the US experienced 27 billion-dollar disasters in 2024, resulting in $182.7 billion in damages. With no relief expected in 2025, insurance stakeholders are sharpening strategies to contain exposure and protect capital.
In the world of risk management, few bring the breadth of perspective that Kevin Sharp (pictured) does. As chief risk officer at ICAT, Sharp’s background in hurricane meteorology, catastrophe modeling, and reinsurance broking positions him at the center of the industry's evolving risk strategies. In a recent interview with Insurance Business, Sharp explained how ICAT is advancing its catastrophe modeling and exposure management strategies to keep pace with escalating volatility and rising industry losses.
“When it comes to hurricanes, the biggest thing is recognizing how much uncertainty exists, not just in seasonal variability but in where storms will actually make landfall,” Sharp told IB. “Last year, Hurricane Beryl became the earliest Category 5 storm on record in the Atlantic. This year, while the Pacific season got off to a fast start, the Atlantic had its latest start in over a decade. A late start doesn’t mean a quiet season. Hurricane Andrew was the first named storm of 1992 and one of Florida’s most devastating. It didn’t form until August. That’s why forecasting impact is so challenging, and why we stay nimble in how we model and manage risk.”
And it’s that kind of historical whiplash that has shaped ICAT’s approach. “You can have seasons with well above-average storm activity, and yet none make landfall,” added Sharp. “It’s really hard to predict what any given year will do. After the record-breaking 2004 and 2005 seasons that brought major US landfalls and storms like Katrina, many thought this was the new normal. But then we went more than a decade without a single major hurricane making landfall in the US.”
While catastrophe models have advanced significantly, Sharp emphasized the importance of maintaining an independent view of risk. “I started working with catastrophe models during my master’s research on hurricane impacts and early in my industry career,” he said. “They’ve come a long way, but even now it’s critical to ground them in real-world experience and claims data.”
Instead, ICAT has developed a proprietary risk framework called PEAK, built by its experienced in-house catastrophe modeling team. “We’re very proud of our modeling team – they’re the brains behind PEAK,” Sharp said. He explained that the team regularly compares model output to actual loss experience, asking what the model predicted, what actually happened, and how that insight can inform adjustments going forward.
“It incorporates over 25 years of ICAT’s claims history and helps us continuously refine our expectations based on what we’ve learned,” Sharp told IB. This approach allows ICAT to align external probabilistic models with real-world data – particularly vital in high-risk areas like the Gulf Coast and Southeastern US.
Modeling aside, new and innovative technology has also accelerated ICAT’s ability to assess property vulnerability, and artificial intelligence is now embedded in their underwriting process.
“We've started utilizing AI through aerial imagery,” Sharp revealed. “For all the properties that we underwrite, we can access recent aerial imagery to get high-resolution views of what the property looks like.”
But ICAT isn’t alone in this sector’s fervor for AI. Data from ZestyAI found that one in four insurers are now using AI to assess risks associated with severe storms, with 20% of executives adding that they believe AI and machine learning models are the most accurate, and 27% believing that a combination of different modelling approaches offers the best predictions for managing risk.
ICAT continues to inspect every property it underwrites, but the integration of aerial AI has enhanced the process. AI scans recent imagery to assess not only roof condition but also surrounding hazards. It has been a gamechanger.
“Instead of sorting through thousands of images manually, we can use AI to flag the condition of a property,” Sharp added. “It helps us direct underwriter attention where it’s most needed.”
With reinsurance rates near all-time highs, particularly in catastrophe-prone regions, the economic pressure on the market is intensifying.
“It’s about staying in tune with where the reinsurance market is,” Sharp said. “Rates have surged in recent years, largely driven by rising claims – especially from secondary perils like severe convective storms, which include tornadoes and hail.”
“Wildfires have also been a contributing factor,” Sharp said. “Even when you’re writing properties in hurricane-prone regions, you must account for the broader reinsurance environment. With pressure coming from hurricanes and other perils, reinsurance costs affect pricing across all catastrophe-prone areas.”
ICAT’s approach emphasizes disciplined risk management, appropriate pricing, and pursuing opportunities in segments the company understands well. “It’s about maintaining underwriting discipline while staying aware of market dynamics,” Sharp said. “We focus on writing where we have strong insight into risk and can diversify intelligently.”
Legal and regulatory changes also play a role – something ICAT must account for when assessing risk. Sharp points to Florida’s legislative overhaul following Hurricane Irma in 2017, when industry losses came in far higher than expected, driven in part by assignment of benefits issues and increased public adjuster activity. “With Hurricane Ian, there were quite a few improvements,” he said. “Many of the issues we saw during Irma, such as litigation and claims inflation, had improved noticeably, largely due to legislative changes. Still, the potential impact of legal dynamics remains an important factor to estimate.
The growing impact of compound events – where wind, storm surge, and inland flooding converge – has led ICAT to enforce a more holistic approach to coverage.
“You have these compounding perils,” said Sharp. “When people think about hurricanes, they tend to focus on wind, but storm surge accounts for a significant share of losses. We’ve put a stipulation in place: if someone wants to buy wind coverage from us and we can’t offer storm surge, they’re required to carry storm surge coverage from another provider.”
“We don’t want situations where insureds are left without flood protection. Requiring both wind and flood coverage isn’t just about protecting our portfolio – it helps avoid confusion and disputes during the claims process and ensures the insured isn’t caught off guard when it matters most. It reduces the risk of coverage gaps that can complicate settlement and delay recovery.”
Looking ahead, Sharp sees a growing role for alternative risk transfer, particularly parametric solutions.
“Alternative forms of capacity are at an all-time high and continue to grow alongside traditional markets,” he told IB. “Traditional reinsurance can tighten after large loss years, so parametric structures are becoming more relevant. One of the most important things is understanding the basis risk – what the trigger represents and how it aligns with the actual impact to your portfolio.”
For these solutions to gain wider traction, especially with rating agencies, transparency is critical.
“You also need to be able to clearly communicate how the structure works,” Sharp added. “Carriers often rely on reinsurance to support their financial strength ratings, and rating agencies tend to be more skeptical of parametric solutions. That’s why it’s critical to quantify the expected recovery based on the trigger and translate that into how it will perform relative to your portfolio. When structured thoughtfully, parametric solutions can effectively supplement traditional reinsurance and offer meaningful pricing efficiencies.”
In an era of escalating catastrophe losses, ICAT’s adaptive risk strategy under Sharp’s leadership stands out as a model of discipline over guesswork. By refining internal models, integrating AI into underwriting, and insisting on comprehensive coverage frameworks, Sharp and his team are helping redefine how the industry approaches catastrophe risk for the long term.
