Nuclear Verdicts
How should we reserve for social inflation in long-tail liability lines?
Updated July 2026
Reserve against the current trend, not a historical loss year. Social inflation lifted nuclear verdict frequency from 44 in 2020 to 149 in 2025, a 239 percent rise, so a model calibrated on older years understates the tail. CaseGlide does not set reserve figures. It gives claims teams current case status, exposure trajectory, and settlement posture, the visibility a reserving process needs to re-baseline.
Why does social inflation force long-tail reserves higher?
Because the baseline itself is moving. Social inflation, rising jury distrust of corporations combined with anchoring, reptile-theory framing, and third-party litigation funding, has pushed nuclear verdict frequency up every year since 2020. In 2025 there were 149 verdicts of $10 million or more, totaling $25.1 billion. A long-tail line reserved against last decade's baseline is reserved against a baseline that no longer exists.
| Year | Nuclear verdicts ($10M+) | Combined damages |
|---|---|---|
| 2020 | 44 | $4.2B |
| 2021 | 58 | $7.1B |
| 2022 | 72 | $9.8B |
| 2023 | 89 | $14.5B |
| 2024 | 135 | $31.3B |
| 2025 | 149 | $25.1B |
+239%
Growth in annual nuclear verdict frequency from 2020 to 2025
Litigation Sentinel verdict database
What specifically drives social inflation in jury awards?
No single cause. Anchoring plants an outsized number early so the jury negotiates down from it instead of up from zero. Reptile theory frames the defendant as a danger to the community. Third-party litigation funding gives plaintiffs the capital to hold out for a bigger number. Together they lift what a jury treats as reasonable, which is the mechanism a reserve actually has to price.
- Anchoring: a large early number resets what the jury treats as a reasonable award.
- Reptile theory: framing the defendant as a threat to community safety raises punitive appetite.
- Third-party litigation funding: outside capital lets plaintiffs hold out for larger awards instead of settling early.
- The compounding effect: each mechanism reinforces the others, which is why the baseline moves every year, not just in outlier cases.
What actually improves reserve accuracy on litigated long-tail claims?
Current visibility into the case, not a better formula alone. A reserve set at intake and never revisited misses the file that is drifting: specials climbing, venue turning unfavorable, the plaintiff's anchor unchallenged. Structuring live case status, exposure trajectory, and settlement posture from defense counsel status reports gives a reserving process the update signal an annual loss review cannot supply alone.
CaseGlide is a litigation intelligence platform, not an actuarial consultancy. It does not compute reserve figures or run loss models. What it does is structure the live status reports your defense counsel already file into current case status, exposure trajectory, and settlement posture, by venue and case type, so the people who set reserves are working from this quarter's file instead of last year's assumption.
How does venue concentration change reserve strategy for long-tail lines?
Because nuclear verdicts do not spread evenly across a long-tail book. California, Florida, Texas, Georgia, and New York carry the bulk of tracked activity, and a handful of counties inside them drive most of it. A reserve model that treats every open venue the same underweights the tail sitting in a small number of high-verdict jurisdictions.
- Map every open long-tail matter to its filing state and county.
- Weight reserve scrutiny toward top-tier venues, especially products, auto and trucking, and medical malpractice.
- Re-baseline whenever a venue passes tort reform, since Florida's 2023 reform and Georgia's SB 68 in 2025 both changed the mechanics without eliminating the risk.
Common questions
Does CaseGlide calculate reserves or recommend reserve amounts?
No. CaseGlide is a litigation intelligence platform, not an actuarial or reserving tool, and it does not compute a reserve figure, run a loss triangle, or recommend a number for any claim. What it does is structure the live status reports your defense counsel already file into current case status, exposure trajectory, and settlement posture, organized by venue and case type. That gives the people who actually set reserves, whether an in-house actuary, a reserving committee, or a claims executive, a current read on the file instead of a stale one. The reserving methodology, the judgment, and the number stay entirely with your team. The platform's contribution is visibility into what has changed in the case since the last reserve review.
See the CaseGlide platform→Why do historical loss triangles understate today's nuclear-verdict tail?
Because the frequency of nuclear verdicts has more than tripled in five years, and a triangle built on older loss years assumes a tail that no longer applies. Litigation Sentinel tracking recorded 44 nuclear verdicts of $10 million or more in 2020 and 149 in 2025, a 239 percent increase, with combined damages reaching $25.1 billion in 2025. A model calibrated on 2018 or 2020 loss experience is pricing against a jury environment that has since shifted through anchoring, reptile-theory framing, litigation funding, and broader social inflation. The frequency trend, not any single year's damages total, is the more reliable signal for how far a reserve baseline needs to move.
The nuclear verdict growth rate→What is social inflation, in plain terms?
Social inflation is the broad rise in jury willingness to award large sums against corporations and insured defendants, independent of actual economic damages or general inflation. It shows up as juries treating eight and nine-figure numbers as normal in cases that would have settled for a fraction of that a decade ago. It is not one mechanism but several reinforcing ones: plaintiff attorneys anchoring on outsized early numbers, reptile-theory tactics that frame the defendant as a danger to the community, and third-party litigation funding that lets plaintiffs hold out for bigger awards. Together they have pushed nuclear verdict frequency from 44 in 2020 to 149 in 2025, which is the practical, measurable face of social inflation for a long-tail book.
Why nuclear verdicts happen→Does venue matter more than case type for long-tail reserving?
They work together, but venue is the sharper lever. California, Florida, Texas, Georgia, and New York carried the bulk of tracked nuclear verdict activity in 2025, and within them a handful of counties drive most of the exposure, so two claims with identical facts can carry very different reserve implications depending on where they are venued. Case type still matters: products liability, auto and trucking, and medical malpractice concentrate a disproportionate share of nuclear-range severity. The practical approach is to weight reserve scrutiny by both, with venue as the first filter and case type as the second, rather than reserving every long-tail claim of a given type identically regardless of where it sits.
Which states have the most nuclear verdicts→CaseGlide is the litigation intelligence platform for Fortune 500 legal departments and insurance claims organizations. It structures live litigation data from defense counsel into executive decisions: reducing defense spend, settling the right cases sooner, and shrinking litigated claim volume.
Next step · See it on your docket
See what your litigation portfolio is telling you
A 30 minute walkthrough on your own docket. No slides, no committee.