Severe Thunderstorms: A Frequency, Accumulation, and Volatility Peril

April 21, 2026 / 5 min read

Severe thunderstorms aren’t catching the insurance industry off guard. The scale and volatility of losses are. For many insurers and reinsurers, the question is no longer whether severe thunderstorms matter. It’s whether their current view of risk can adequately explain earnings volatility, support pricing decisions, and withstand scrutiny in reinsurance and capital discussions.

The cause is often not a single, outsized catastrophe. It is accumulation—repeated storms, localized damage, and frequency-driven loss that compounds across the year before it becomes visible in underwriting results.

clouds-storm-horizon-homes

Severe thunderstorms are now among the most consistent drivers of insured catastrophe losses in the United States, fueled by both headline events and frequent attritional activity across portfolios. In 2025, they contributed 42% of global annual average loss, according to Verisk’s Global Modeled Catastrophe Losses report. Increasingly, they represent a frequency-driven volatility challenge shaped by evolving hazard patterns, growing exposure, and portfolio concentration.

This context underpins Verisk’s current view of U.S. severe thunderstorm risk. The model reflects near-present climate conditions, regenerated event catalogs, a component-based vulnerability framework, and explicit treatment of hail, tornado, and straight-line wind so insurers can make more reliable underwriting, reinsurance, and capital decisions.

One year after release, the key question is not how the model changed, but how well this view of risk holds up in practice. Across the market, insurers and reinsurers have spent the past year evaluating the updated severe thunderstorm model for the U.S. against their portfolios—examining hazard behavior, vulnerability assumptions, and loss distributions to understand how the updated view of risk compares with observed experience.

Sub-peril separation matters because financial outcomes depend on it

Hail, tornado, and straight-line wind may sit under the same severe thunderstorm umbrella, but they don’t behave the same financially. Policy terms often differentiate by sub-peril. Treaty structures can also respond differently depending on whether hail accumulation, tornado outbreaks, or broader wind activity drive losses. When output is blended into a single severe thunderstorm number, pricing precision can weaken, alignment with contract terms becomes harder, and post-event attribution becomes more difficult.

Separating sub-perils makes the output more useful in practice. When a portfolio underperforms, carriers need to know more than how much loss occurred. They need to know what drove it, where it happened, and which segments were affected. The model treats hail, tornado, and straight-line wind independently and carries those distinctions through the financial framework. This allows contract terms, deductibles, and limits to respond realistically, improving pricing, reinsurance structuring, and loss attribution.

Hurricane and earthquake models are often judged primarily by how they represent the tail. Severe thunderstorms demand equal attention to the middle of the loss distribution. This peril produces losses through a continuum: repeated moderate hail events, regional outbreak sequences, and multiday activity across portfolios.

The challenge is not only event severity. A catalog can produce plausible tail metrics while underrepresenting the moderate-frequency activity that drives quarterly volatility. When that happens, a portfolio can appear more stable than it actually is.

The Verisk Severe Thunderstorm Model for the U.S. was designed to preserve this full distribution. The regenerated catalogs simulate up to 100,000 years of activity, representing millions of severe weather events. The objective is not simply more simulation—it is a more complete representation of the continuum of severe thunderstorm activity that drives both attritional and catastrophic loss.

Vulnerability must reflect what claims actually show

Hazard becomes insured loss through vulnerability. In hail, that translation is especially sensitive to building-level characteristics.

Recent claims experience makes that clear. Roof condition and remaining useful life materially influence loss severity. Verisk’s forthcoming 2026 roof report show that roofs in moderate-to-poor condition can generate up to 60% higher loss costs than roofs in good or excellent condition, while roofs nearing the end of their useful life can sustain substantially more damage under comparable storm conditions. At the same time, replacement costs have increased, amplifying the severity of already elevated loss activity.

2437700_USST_Blog_image_Map_wLabels — ^{In 2025, 16 U.S. states had at least 20% of roofs affected by severe hail, with several Plains states exceeding 40%. These concentrations highlight how localized severe thunderstorm risk can be across regions and years.}

This is exactly why vulnerability calibration matters. If building-level characteristics materially alter damage outcomes, they need to be represented inside the model rather than left to downstream adjustments or broad assumptions.

The model incorporates detailed roof attributes—including roof age and roof covering—alongside other secondary characteristics. These vulnerability functions are calibrated using Verisk’s extensive claims datasets, including more than $130 billion in detailed property claims and over $370 billion in aggregate loss experience, helping ensure modeled damage relationships reflect how buildings actually perform during severe thunderstorms.

The model captures severe thunderstorm risk not only to personal and commercial property, but also includes support across additional assets, such as industrial facilities and solar-related risks.

Solar: a distinct vulnerability that requires explicit modeling

Solar installations experience distinct failure modes under hail and straight-line wind—from panel fracture and inverter damage to racking deformation and array-level faults. These failure modes produce a different loss profile: repair or partial-replacement costs combined with potential business-interruption or grid-connection expenses.

Verisk’s severe thunderstorm model incorporates more than 70 new vulnerability functions for rooftop and utility-scale solar installations, calibrated against observed claims patterns and Verisk’s broader loss database.

Representing solar explicitly improves loss attribution, reduces the need for post-event adjustments, and supports more accurate pricing and accumulation management where solar penetration continues to grow.

One year later: how insurers are validating this view of risk

In today’s environment, insurers and reinsurers do not need a severe thunderstorm model that simply looks more sophisticated. They need one that performs in real decision contexts. Over the past year, validation efforts have focused on:

Regional hail frequency and severity
Vulnerability impacts (roof age, materials)
Loss distribution across the curve
Alignment with recent loss experience

Across many evaluations, insurers report that the model better reflects observed volatility, explains accumulation patterns more clearly, and aligns more closely with recent loss behavior.

Severe thunderstorms will continue to generate frequency-driven volatility. The role of a catastrophe model is not to smooth that volatility but to represent it clearly enough that decisions remain defensible when compared with real outcomes.

Verisk’s current view of severe thunderstorm risk is designed with that objective:

Reflecting near-present climate conditions
Aligning with observed loss behavior
Supporting real-world underwriting and capital decisions

One year after release, the model’s value is increasingly visible not in theory, but in how insurers validate results, refine portfolio assumptions, and make more confident decisions.

Request a validation briefing to see how our view of severe thunderstorm risk performs against your portfolio—and how it can support underwriting, reinsurance, and capital decisions.

Learn more