A contractor’s general liability policy carries two numbers that matter in a bad year: the per-occurrence limit, which caps what the insurer pays for any single loss, and the general aggregate, which caps what it pays for all losses combined across the entire policy term. Most contractors watch the first number and ignore the second. In 2026, with liability verdicts climbing and construction claims getting more expensive to resolve, the aggregate is where a contractor quietly runs out of coverage — and the per-project aggregate endorsement is the instrument that keeps it from happening.
The problem is erosion. A standard commercial general liability policy sets one general aggregate for the whole year. Every claim payment and every defense dollar on every job draws the same shared balance down. A contractor running eight projects at once is protecting all of them with a single pool. One serious completed-operations claim on one project — a water-intrusion defect, an injury suit, a construction-defect action that drags on — can consume a large share of that pool through settlement and legal fees, and the seven other jobs now stand behind a limit that has already been spent. The policy has not been cancelled. It has simply been emptied by a loss somewhere else.
The verdict environment is what makes this urgent. Social inflation — the steady rise in jury awards and litigation costs beyond ordinary economic inflation — is pushing liability losses to levels that erode aggregates fast. Enlyte’s review of the second quarter of 2026 catalogued nine so-called nuclear verdicts above $10 million, led by a $1.6 billion award in a Texas workplace-explosion case. The median nuclear verdict has risen more than 25% over the past decade, and in many of these cases less than 15% of the award is tied to actual economic loss — the rest is the product of a jury’s view of the defendant. One trade analysis pegged the broader social-inflation drag on casualty insurance at roughly $529 billion. For construction specifically, reported insurance costs have surged around 22% in 2026, and carriers increasingly tie that to the cost of defending and settling liability claims rather than to property losses alone.
A per-project aggregate endorsement changes the architecture. Instead of one shared pool, it gives each project its own separate general aggregate limit. A claim on the parking-structure job draws only against that job’s aggregate; the office build, the school addition, and the tenant fit-outs each keep a full, untouched aggregate of their own. The endorsement does not raise the per-occurrence limit — a single catastrophic loss is still capped at the occurrence number — but it ensures that a bad outcome on one site does not strip protection from every other site the contractor is running. In a year when a single claim can run into eight figures, that separation is the difference between a contained loss and an uninsured exposure.
Owners and general contractors are increasingly writing the requirement into the contract. It is now common to see insurance specifications demand a per-project aggregate, often referencing the ISO CG 25 03 endorsement by name, alongside additional-insured and primary-and-noncontributory language. A subcontractor who cannot produce it may lose the bid or, worse, win the work and discover at renewal that the carrier will not attach the endorsement on the terms the contract assumed. The clause is quiet until it is not — and the moment it matters is the moment a certificate is rejected or a claim exhausts a shared limit mid-project.
The discipline, then, is to read the aggregate as carefully as the occurrence limit, because the two fail in different ways. A contractor should know whether its policy carries a per-project aggregate, whether that endorsement actually tracks the projects the way the contract language assumes, and whether the defense costs that erode the limit are inside or outside it. Those are not questions to answer at claim time. They are underwriting questions, and they belong in the submission.
This is where our four-step process does its work. Strategic Discovery maps the contractor’s active and pipeline projects and the insurance requirements buried in each contract. Risk Assessment tests the current CGL structure against a realistic bad year — not one claim, but the concurrent exposure across every open job. Solution Design places the per-project aggregate, aligns the limits with what owners actually require, and confirms how defense costs interact with the aggregate. Ongoing Optimization keeps the structure current as projects close and new ones start, so the coverage schedule never drifts out of step with the work.
A contractor builds in the open, one project at a time, and expects each to stand on its own. The insurance behind the work should be built the same way. A per-project aggregate endorsement is how a contractor keeps a loss on one site from darkening every other — and in the verdict climate of 2026, that separation is no longer a refinement. It is the floor.
— Ryan Mefford, President & Risk Advisor
Sources
- Enlyte — The Largest Nuclear Verdicts in Q2 2026
- Billyfor Insurance — Nuclear Verdicts Are Rewriting Construction Insurance (2026)
- Grit Insurance — 2026 Construction Insurance Market Outlook
- BuilderMuse — Construction Insurance Costs Surge 22% in 2026
- Actuary.info — Social Inflation & Litigation Trends 2026
- Aon — 2026 P&C Outlook: Navigating Volatility, Unlocking Growth
- Zurich — How Social Inflation and Nuclear Verdicts Shape the Liability Landscape