Casualty lines posted $15.8 billion in adverse prior-year development in 2024, the highest level on record and equal to 5.9% of prior reserves, as nuclear verdicts jumped 52% in count to 135 cases and 116% in dollar terms to $31.3 billion.

That shock is now sitting inside every carrier's loss triangle, and it does not sit in one accident year. It sits on one diagonal, which is a different and harder problem.

Key Takeaways

  • One calendar diagonal, every accident year. An AY2021 claim maturing 36 to 48 months and an AY2023 claim maturing 12 to 24 months in calendar 2024 both absorb the same verdict shock.
  • Both Bornhuetter-Ferguson inputs move the wrong way at once. Accelerating link ratios understate multiyear-average development factors, and a roughly three-year lag from loss cost to premium biases the a priori loss ratio low.
  • A 43-point loss ratio gap separates the top and bottom quintile of other liability writers over 2021 to 2025, 59% against 102%, and reinsurers priced that dispersion directly at July 1.
  • $10 million average lead umbrella limit, down from $20 million in 2019, against a median top-case verdict of $98 million versus $49.7 million. Severity is migrating into layers with thinner triangle data.

How One Diagonal Contaminates the Triangle

Age-to-age factors are typically selected as an average of the ratios observed across several historical accident years at each maturity, assuming the process generating them is stable. Nuclear verdict activity breaks that assumption in a specific way. It does not raise severity evenly across the triangle; it concentrates in the diagonal for the current calendar period, because that is when verdicts, settlements struck in their shadow, and case reserve strengthenings get booked.

An accident year 2021 claim maturing from 36 to 48 months in calendar 2024 and an accident year 2023 claim maturing from 12 to 24 months in the same period both pick up the same shock, though they sit in different rows and columns. Average historical link ratios to select a 36-to-48 month factor for a fresh accident year and that single contaminated diagonal pulls the selected factor upward for every accident year evaluated afterward, including ones whose own verdict exposure has not yet emerged.

Milliman's year-end 2024 statutory data shows where the bias landed. Other liability occurrence carried $10.0 billion of adverse development against $150.8 billion of prior reserves, a 6.6% miss. Commercial auto added $3.8 billion against $66.7 billion, 5.7%. Non-proportional reinsurance liability contributed $1.7 billion against $37.1 billion, 4.6%.

The largest percentage misses by accident year sit in the 2015-2019 soft-market cohort, with accident year 2017 alone running 12% adverse against its prior carried reserves. The contamination has been compounding since well before the 2024 verdict spike made it visible.

Which Method the Contamination Breaks

The CAS's Jim Lynch and Dave Moore worked the mechanics through for Bornhuetter-Ferguson specifically: when link ratios accelerate from one accident year to the next, multiyear-average cumulative development factors come out understated, and because industry data shows roughly a three-year lag between loss cost increases and the premium increases that follow in commercial auto liability, the a priori expected loss ratio is also biased low. Both inputs move the same wrong direction at once, which is why an unadjusted BF reserve misses low precisely when verdict severity is accelerating.

Method Mechanism Best fit
Bornhuetter-Ferguson Blends an a priori expected loss ratio with reported development; limits how far a single contaminated diagonal can move the ultimate for immature years Commercial general liability, where premium and exposure data support a credible a priori pick
Cape Cod Derives the expected loss ratio from the triangle's own reported experience weighted by used-up exposure, rather than an external pricing assumption Commercial auto liability, where recent rate actions make an external a priori assumption unreliable but exposure data is credible
A-priori-first (frequency × severity) Builds the ultimate from separately trended frequency and severity assumptions rather than aggregate link ratios, isolating the verdict-severity shock as its own trend component Excess casualty and umbrella, where a handful of large claims dominate the triangle and frequency is too thin for chain-ladder credibility

For commercial general liability, BF still caps the influence of a contaminated triangle on immature years, provided the a priori loss ratio is not itself drawn from the same history. The corrective is direct: select link ratios from the most recent development year or an extrapolated trend rather than a multiyear average, and adjust the a priori loss ratio upward explicitly.

For commercial auto, Cape Cod has gained ground because BF's anchor to a pricing assumption becomes a bigger liability when rate actions have moved fast. Deriving the expected loss ratio from used-up-exposure-weighted experience sidesteps whether last year's filed indication was adequate, but it does not escape triangle contamination: applied to an unadjusted triangle it launders the same severity through a different weighting scheme.

The tempting shortcut is to cap extreme verdicts out of the link-ratio calculation entirely, and that is where the actuarial judgment binds. Thermonuclear verdicts above $100 million hit a record 49 cases in 2024 against 27 in 2023, five topped $1 billion against two, and the median verdict climbed to $51 million from $44 million. With 135 verdicts above $10 million in 2024 across 34 states and 77 courts, an actuary capping at $10 million is not removing noise. That actuary is removing the specific tail the excess and umbrella layers were priced to absorb.

The defensible version keeps both pieces visible. Cap or Winsorize for link-ratio selection, document the threshold and its rationale, then reserve the excluded severity through an explicit large-loss load calibrated to the actual frequency and severity distribution in the relevant venue and line. The Statement of Actuarial Opinion should carry the capped ultimate, the load, and the sum, so a reviewer can see what left the triangle and where it came back in.

The Cushion Is Thinner Than the Miss

A reserving miss on accident years 2021 through 2023 forces a correction that flows through current-year loss picks, and the margin available to absorb it is narrower than it was in 2022.

Analysts project a 96% to 97% combined ratio for 2026, reflecting a normalizing catastrophe year and materially lower favorable reserve development than the releases that flattered 2025. Aon's outlook shows general liability rates rising 5.6% in the fourth quarter of 2025 with a forecast climb to 9%, and auto liability up 9.2% with 7% to 15% forecast. Those increases are a lagging response to verdict severity already booked, not a buffer against verdicts still to come.

The tower is also compressing. Average lead umbrella limits have fallen to $10 million from $20 million in 2019 while the median verdict for top US casualty cases climbed to $98 million from $49.7 million over the same span. Primary and lower excess layers are absorbing severity that used to sit higher, so the development-factor problem is migrating into layers with thinner loss experience and less credible triangle data of their own.

Reinsurers are pricing the resulting dispersion rather than averaging across it. Howden Re found ceding commissions flat overall at the July 1, 2026 casualty and financial lines renewals, but with a 43-percentage-point calendar-year loss ratio gap between the top and bottom quintile of other liability writers over 2021 to 2025, 59% against 102%. Cedants able to document credible adjustment for verdict contamination were priced on their own merits; opaque positions were priced to the market's worst case.

Individual filings show the same spread. W.R. Berkley reported $8 million of adverse development in its Insurance segment from umbrella and excess liability claims in accident years 2019 through 2023, citing rising frequency of litigated claims. RLI released $14 million favorable in casualty across accident years 2018, 2019, 2021, 2022 and 2025, while its personal umbrella book developed adversely.

Peer benchmarking will not settle which side a given carrier sits on. Within Milliman's own adverse cohort, 49% of companies reported favorable development, and within the favorable cohort 41% reported adverse. The dispersion inside each grouping is nearly as wide as the dispersion between them, so line-of-business mix does not tell a carrier where its own triangles stand.

Further Reading

Sources