Modeling the Unmodeled: Secondary Perils and the Nat Cat Reinsurance Reset

By Hitul Mistry | Last reviewed: November 2025

For a generation, catastrophe reinsurance was built around a handful of peak perils — Atlantic hurricanes, Japanese and Californian earthquakes, European windstorm. Yet the losses that now define the market rarely make those lists. So-called secondary perils — severe convective storms (SCS), inland flood, wildfire, hail, and drought — accounted for the majority of global insured catastrophe losses in recent years, with SCS alone driving more than USD 60 billion of insured losses in 2023 (Swiss Re Sigma, 2024). Global insured natural catastrophe losses have now exceeded USD 100 billion for several consecutive years, and roughly 70–80% of that total has come from secondary rather than peak perils (Aon, Climate and Catastrophe Report, 2024). The problem is that the industry's modeling, capital, and treaty architecture were engineered for the tail, not for a relentless drumbeat of mid-sized events. The 2023 renewal reset was, in large part, the market's attempt to correct that mismatch.

Talk to Our Specialists

Why have secondary perils overtaken peak perils in insured losses?

Secondary perils have risen because exposure growth, inflation, and a climate signal have combined to make frequency — not a single mega-event — the dominant loss driver in most calendar years.

1. Exposure accumulation in hazard-prone geographies

• Population and property values have grown fastest in convective-storm and wildland-urban-interface zones such as the U.S. Southeast, Texas, and the mountain West.
• Rising replacement costs and post-event demand surge inflate every hail and wind claim beyond historical assumptions.

2. A measurable but uneven climate signal

• Warmer, moister atmospheres raise convective available potential energy, lengthening hail and tornado seasons in several basins.
• Prolonged drought and heat extend wildfire windows, while intense short-duration rainfall drives pluvial flood outside mapped floodplains.

3. Frequency erodes retentions faster than severity

• Multiple sub-catastrophe events in one year can breach an annual aggregate deductible that a single event never would.
• Cedents historically priced these as attritional losses, only to see them accumulate into treaty-threatening totals.

4. Underinsurance and data gaps compound the trend

• Flood and wildfire exposure is frequently mislabeled or absent in cedent bordereaux, understating true accumulation.
• Secondary-peril losses often surprise on the downside because the exposure was never fully captured in the first place.

How did the 2023 reset change attachment points and structures?

The reset re-drew the line between what cedents retain and what reinsurers pay by lifting attachment points sharply and curtailing the aggregate covers that had been absorbing frequency losses.

1. Higher attachment points across the board

• Property-catastrophe attachment points rose by 50–100% or more at loss-affected accounts in the January 2023 renewals (Gallagher Re, 1st View, 2023).
• The intent was explicit: return treaties to genuine tail protection and stop subsidizing earnings volatility.

2. Retrenchment from aggregate covers

• Reinsurers pulled back from annual aggregate and low-attaching per-occurrence layers most exposed to secondary-peril frequency.
• Where aggregate capacity remained, it came with franchise deductibles, event caps, and named-peril restrictions.

3. Tighter terms and narrower wordings

• Hours clauses, occurrence definitions, and cascading provisions were tightened to prevent frequency events from aggregating into single large recoveries.
• Named-peril and per-occurrence structures were favored over broad all-natural-perils aggregate protections.

4. A durable repricing, not a one-year spike

• Risk-adjusted property-catastrophe rate increases persisted into the 2024 and 2025 renewals before stabilizing (Guy Carpenter, 2025).
• The market signaled that the higher-attachment regime is structural, leaving cedents to manage secondary-peril volatility net or through alternative capital.

Talk to Our Specialists

Where do vendor catastrophe models fall short on secondary perils?

Vendor models remain far more mature for peak perils than for secondary ones, leaving reinsurers exposed to completeness, resolution, and trend gaps precisely where losses now concentrate.

1. Thin and non-stationary event catalogs

• SCS, wildfire, and flood have shorter reliable historical records than hurricane or earthquake, weakening statistical stability.
• Non-stationarity from climate change means the historical catalog may understate today's frequency and severity.

2. Coarse resolution versus localized hazard

• Hail swaths, flood footprints, and wildfire perimeters vary block-by-block, but many models resolve hazard too coarsely to capture that granularity.
• Defensible-space, roof-age, and drainage attributes that drive damage are often missing from the exposure schema.

3. Uncertain vulnerability and damage functions

• Damage curves for hail and wind-driven roof loss are highly sensitive to construction, age, and material assumptions that vary by region.
• Post-event inflation and litigation (for example, roofing assignment-of-benefits disputes) distort observed severity in ways models rarely anticipate.

4. Divergent views demand blending

• Different vendors can produce materially different loss estimates for the same SCS or wildfire portfolio.
• Relying on a single model view creates hidden basis risk; a blended, expert-adjusted view is now best practice.

What treaty structures work best for secondary-peril volatility?

No single structure solves secondary-peril risk; effective programs layer occurrence, aggregate, and alternative-capital covers so each responds to a distinct slice of the loss distribution.

1. Occurrence excess-of-loss for the tail

• Per-occurrence cat XL, attaching high, protects against the rare severe convective outbreak or major wildfire complex.
• Higher post-reset attachment points concentrate this cover on genuine tail events.

2. Aggregate covers for frequency — carefully structured

• Annual aggregate XL can protect against an accumulation of medium events, but requires franchise deductibles and event caps to remain sustainable.
• Pricing must load explicitly for frequency trend and climate uncertainty.

3. Parametric and ILS complements

• Parametric triggers (hail size, wind speed, burned area) pay quickly and reduce basis on well-defined hazards.
• Aggregate cat bonds and collateralized reinsurance add capacity but demand transparent terms given attritional erosion risk.

The table below contrasts how the main structures respond to secondary-peril dynamics.

Structure	Trigger basis	Secondary-peril fit	Key limitation
Occurrence cat XL	Single event above attachment	Strong for tail SCS/wildfire	Misses frequency accumulation
Annual aggregate XL	Sum of qualifying losses vs. annual deductible	Good for frequency	Erosion risk; harder to price
Parametric	Physical index (hail size, burned acres)	Fast, low-dispute	Basis risk if index diverges
Quota share	Proportional share of losses	Smooths volatility	Cedes profit; no tail focus
Aggregate cat bond	Modeled or indemnity aggregate	Adds capacity	Attritional erosion, tight terms

Talk to Our Specialists

How can data and AI close the secondary-peril modeling gap?

Data enrichment and AI let reinsurers see exposure and accumulation that vendor models miss, turning incomplete bordereaux into decision-grade views of secondary-peril risk.

1. Geospatial and structural exposure enrichment

• Satellite imagery, LiDAR, and property databases fill missing roof-age, construction, and defensible-space attributes.
• Vegetation and terrain layers refine wildfire and pluvial-flood exposure beyond mapped zones.

2. Multi-model blending and adjustment

• Machine-learning ensembles reconcile divergent vendor views and calibrate them to a cedent's actual loss experience.
• Expert-in-the-loop adjustment keeps blended outputs defensible for rating agencies and capital providers.

3. Accumulation and drift detection

• Analytics flag creeping concentration in convective or wildfire corridors before it breaches appetite.
• Near-real-time monitoring replaces quarterly manual roll-ups, shortening the reaction time to portfolio drift.

4. Faster, evidence-based renewals

• Submission triage prioritizes accounts where enriched data changes the risk view most.
• Transparent, data-backed pricing narratives strengthen the cedent-reinsurer conversation at renewal.

What does the outlook hold for nat cat capacity and pricing?

The market is likely to hold its higher-attachment discipline while gradually improving secondary-peril modeling and welcoming more alternative capital.

1. Structural, not cyclical, discipline

• Reinsurers have signaled they will not return to low-attaching aggregate covers absent a material change in loss trend.
• Cedents will continue to invest in net retention management and internal analytics.

2. Growing but selective alternative capital

• ILS capacity, including catastrophe bonds, reached record levels but remains selective on frequency-exposed aggregate risk (Artemis, 2025).
• Sidecars and collateralized structures will increasingly share secondary-peril tail risk on clear terms.

3. Modeling investment accelerates

• Vendors and reinsurers are expanding SCS, wildfire, and flood model coverage and climate conditioning.
• The competitive edge shifts to those who blend, enrich, and adjust models rather than accept them at face value.

4. Protection-gap pressure persists

• A large share of secondary-peril losses remains uninsured, especially in flood and emerging markets (Swiss Re Institute, 2024).
• Closing that gap sustainably requires better data, parametric innovation, and disciplined capacity.

Talk to Our Specialists

Frequently Asked Questions

What are secondary perils in reinsurance?

Secondary perils are high-frequency, lower-individual-severity catastrophes such as severe convective storms, floods, wildfires, hail, and drought. In aggregate they now drive the majority of annual insured catastrophe losses, even though no single event rivals a peak hurricane or earthquake.

Why did nat cat attachment points rise after 2023?

The 2023 renewal reset pushed reinsurers to raise attachment points and restrict aggregate covers so that treaties responded to true tail events rather than absorbing frequency losses. Cedents retained more secondary-peril volatility net.

How well do vendor cat models capture secondary perils?

Historically poorly. Peak-peril models for hurricane and earthquake are mature, but SCS, wildfire, and flood models have thinner event catalogs, coarser resolution, and weaker climate conditioning, leaving material basis and completeness gaps.

What is the difference between aggregate and occurrence cover?

Occurrence cover responds to losses from a single event above an attachment point; aggregate cover accumulates qualifying losses across a period against an annual deductible. Secondary perils stress aggregate structures because many small events erode the annual retention.

Is climate change the main driver of rising secondary-peril losses?

Exposure growth, inflation, and urban sprawl into hazard-prone areas are the largest drivers, but a measurable climate signal is amplifying convective and wildfire frequency and severity in several regions, adding trend uncertainty to pricing.

How does AI improve secondary-peril exposure management?

AI enriches exposure data with geospatial, structural, and vegetation attributes, blends multiple vendor and internal views, and detects accumulation drift far faster than manual processes, sharpening PML estimates and treaty pricing.

What is model blending and why does it matter here?

Model blending combines multiple vendor and proprietary views of risk with expert adjustments to reduce dependence on any single flawed catalog. For secondary perils, where models disagree widely, blending materially improves tail estimation.

Can ILS and cat bonds cover secondary perils?

Increasingly yes, through aggregate and per-occurrence structures, but investors demand clearer parameters and tighter terms because secondary-peril frequency raises attritional erosion risk to their principal.

Editorial note: Figures cited here are drawn from public industry research and reinsurance market commentary and are intended for educational context. Market conditions, model outputs, and loss estimates change frequently. InsurNest does not guarantee specific pricing, capacity, or underwriting outcomes.

Sources

• Swiss Re Sigma — Natural catastrophes and secondary perils
• Aon — Climate and Catastrophe Insight Report
• Gallagher Re — 1st View Renewal Reports
• Guy Carpenter — Reinsurance Market Renewals Briefings
• Verisk — Extreme Event and Wildfire Models
• Artemis — Catastrophe Bond and ILS Market Data
• Swiss Re Institute — Natural Catastrophe Protection Gap

Secondary perils turned frequency into the new tail — the reinsurers who win will be those who model, enrich, and price the losses everyone else left unmodeled.

Talk to Our Specialists

Visit InsurNest to learn more.