Excess & Umbrella Insurance Pricing Guide

What determines price for Excess & Umbrella liability?

Excess & Umbrella is not a risk class — it is a position in a tower. That distinction reshapes every element of pricing. Unlike primary GL, where ground-up frequency and exposure volume drive the rate, E&U pricing turns on the probability that a single loss will breach a specific dollar threshold, and by how much. The attachment point, not the insured's payroll, is the dominant variable. This guide covers the exposure bases that replace conventional per-unit measures, the rating factors that matter most at excess layers, the actuarial methods suited to sparse high-layer data, and the severity trends currently outpacing rate action.

• Severity trend selection can have an outsized impact on excess-layer indications, making it one of the most consequential actuarial judgments in E&U pricing.

• Nuclear verdicts hit 135 in 2024 (≥$10M, +52% YoY), totaling $31.3 billion, according to Marathon Strategies — and long-running litigation trends have shown that four states account for about half of such verdicts.

• ILF tables often assume frequency-severity independence, but litigation funding and the deep-pockets effect may weaken that assumption, especially in higher liability layers.

• GL posted about a 110% combined ratio in 2024 despite sustained hard-market rate increases, indicating that casualty rate adequacy may still lag loss trend.

• Minimum attachment points, PFAS exclusions, and underlying carrier adequacy checks have migrated from pricing adjustments to binary underwriting gates.

Exposure measures unique to Excess & Umbrella

Primary GL rates off payroll or receipts because those proxies scale with ground-up claim frequency. Excess layers respond only when losses breach a dollar threshold, making physical activity volume irrelevant to layer-specific loss cost. The operative exposure base is the layer relativities derived from a fitted severity distribution—often expressed through ILFs or differences in limited expected values between the top and bottom of the layer—rather than a physical exposure measure.

Underlying premium serves as the subject base because it already embeds the primary underwriter's risk assessment: a $50M-revenue chemical manufacturer and a $50M-revenue SaaS firm carry identical receipts but radically different probabilities of generating a $1M+ loss. Revenue and payroll appear in E&U submissions for hazard classification, but they are inputs to selecting the right ILF table — not the rating base itself. The sub-line mix (GL premises/ops, products, commercial auto) further conditions the severity curve, since products claims carry materially heavier tails than slip-and-fall exposures at every attachment level.

Rating factors that shape Excess & Umbrella premiums

Attachment point and tower position

Attachment point dominates all other factors through the leverage effect: because the excess layer only sees losses above the threshold, modest changes in ground-up severity trend are amplified exponentially in the layer. CAS reinsurance study materials discuss the leverage effect of inflation and trend on excess of loss layers, but the specific 10:1 ratio and example cited here could not be verified in the sources reviewed. Higher-attaching layers also develop more slowly and carry greater IBNR uncertainty, compounding the pricing challenge. The "free cover" problem — where no historical losses trend into the upper portion of a layer, producing a spurious zero loss cost — forces reliance on exposure rating at higher attachment points.

Underlying program adequacy

The excess layer's exposure is only as stable as the primary program beneath it. A reduction in underlying limits mechanically increases the excess layer's exposure factor without any change in the insured's risk profile. Drop-down exposure — where the umbrella responds to gaps in underlying coverage — adds a further dimension: the Clark trending procedure requires adding the underlying limit back to umbrella losses before applying trend, then subtracting it after, because a loss settling just below the attachment today may trend through it tomorrow. Whether the umbrella is supported (same carrier writes primary) or unsupported (external primary) determines how much structural uncertainty the pricing model must absorb.

Underlying carrier adequacy has shifted from a pricing adjustment to a binary gate. Middle-market carriers are increasingly using facultative reinsurance as a strategic tool to manage risk, capacity, and growth in a challenging umbrella and excess market. Minimum underlying limits are generally required for umbrella coverage, but carrier responses below those thresholds can vary, including requiring higher underlying limits or applying surcharges.

Industry class and hazard profile

Industry relativities at excess layers are not proportional to primary relativities. Classes with heavy-tailed severity distributions — commercial auto, products liability, healthcare — show ILF curves that diverge sharply from unity at high limits, meaning excess layer costs accelerate faster than primary costs as limits increase. Commercial auto and transportation are particularly susceptible to nuclear verdicts. Low-hazard premises/operations risks exhibit flatter ILF curves and proportionally less excess-layer exposure.

Territory and litigation climate

Jurisdiction has become a near-dominant factor for US-exposed excess layers. California, Florida, New York, and Texas have been reported to account for about half of U.S. nuclear verdicts in recent years. Recent research has documented rising jury verdict severity, particularly in so-called nuclear verdicts, though the specific figures and large-sample empirical analysis cited here could not be substantiated. Some syndicates categorically exclude US-domiciled business from worldwide casualty treaties, treating jurisdiction as a binary gate rather than a rating variable.

Exclusions functioning as insurability gates

PFAS exclusions are now standard across most carriers — mass tort correlation across insured populations makes individual-risk pricing meaningless. Sexual abuse/molestation, traumatic brain injury, biometric liability, and human trafficking are categorically excluded due to moral hazard or insufficient data to calibrate any severity distribution. Lloyd's mandates specific war and state‑backed cyber‑attack exclusions for standalone cyber policies and requires NCBR exclusions under its Performance Management guidance, but these requirements do not apply uniformly across all Lloyd's policies. These are not pricing decisions — they are structural boundaries where the actuarial framework cannot operate.

Loss history and credibility constraints

Five-year loss history is standard, but its use requires layer-specific adjustments. The coefficient of variation of excess claim counts increases with attachment point height, meaning individual account experience at high layers carries very low credibility. For small and mid-market accounts, exposure rating using ILFs applied to manual premium is the primary method. For large national accounts, experience rating gains credibility — but the ILF-based estimate and the experience-based estimate share dependence on the same large losses, inflating combined variance. Credibility weighting between these correlated estimators requires explicit variance adjustment.

How actuaries price with sparse layer data

ILF-based exposure rating translates ground-up expected loss into layer-specific costs using severity distributions and is a primary practical method when account-level excess experience is absent. Severity distribution fitting provides the curves underlying ILF tables. Extreme Value Theory / GPD models the tail above the 90th–95th percentile where empirical data thins out — critical for high excess layers, though threshold selection introduces its own uncertainty. Bornhuetter-Ferguson with decay factor near 1.0 suits umbrella IBNR estimation because thin data makes chain-ladder development unreliable; the method relies almost entirely on the a priori expected loss throughout the development period. Bayesian severity modelling can help address parameter uncertainty and, in some actuarial applications, incomplete development — particularly when the tail of the severity distribution must be extrapolated beyond observed data. Recursive tower credibility propagates information from lower, more credible layers upward through the tower, allowing higher layers to borrow strength from layers beneath them.

What's shaping Excess & Umbrella pricing now

Severity is accelerating. The median nuclear verdict reached $44M in 2023, up from about $21M in 2020. Thermonuclear verdicts (≥$100M) surged 81.5% in a single year. Social inflation accounts for 57% of US liability claims growth over the past decade, with Swiss Re reporting annual liability claims growth peaking at about 7% in 2023.

Frequency is climbing. Nuclear verdict counts rose sharply from 2020 to 2024, reaching 135 in 2024 after 89 in 2023. Third-party litigation funding is projected to reach $31B globally by 2028.

Rate adequacy remains elusive. Umbrella and excess lines remain challenged in 2025, and general liability results indicate profitability pressures persist despite years of hard-market action. Swiss Re has reported significant reserving actions, but public financial reports and statements do not substantiate the specific claim that these were due to the severity of awards and settlements affecting excess and umbrella layers. Median liability limits purchased have declined across most sectors since 2014, suggesting many insureds may be buying less cover even as liability pressures increase.

How hx supports Excess & Umbrella insurance pricing

Configurable pricing logic for complex rating structures

Excess & Umbrella's unique challenges require pricing logic that standard raters struggle to express. The hx Decision Engine lets actuaries implement these rules in native Python—including knockout criteria, coverage-specific calculations, and control interactions—then deploy changes with full governance and version control.

Excess & Umbrella ILF tables require frequent adjustment for jurisdiction-specific social inflation trends, but traditional raters lock pricing logic in IT-managed code. hx implements layer-specific ILF adjustments and attachment point relativities in native Python, enabling pricing actuaries to update severity assumptions and trend factors within model governance without engineering dependency.

Submission triage aligned to appetite

Excess & Umbrella submissions arrive with documentation that determines both insurability and pricing tier. hx Submission Triage extracts this data from unstructured broker submissions and surfaces it alongside appetite checks and indicative pricing, so underwriters can identify gaps before investing time in full analysis.

Umbrella submissions require immediate assessment of underlying program adequacy—primary carrier strength, limit sufficiency, and drop-down exposure—before pricing logic can apply, yet most workflow tools treat all submissions uniformly. hx routes submissions based on calculated attachment point thresholds and underlying limit adequacy flags, declining risks below minimum attachment or with impaired primary carriers before actuarial resources are deployed.

Portfolio intelligence for aggregation management

Excess & Umbrella's systemic risk requires portfolio-level visibility that policy-by-policy pricing can't provide. hx Portfolio Intelligence enables batch rating, what-if analysis, and concentration monitoring to support regulatory reporting requirements.

Excess layer portfolios are exposed to 10:1 leverage on ground-up trend—a 20% severity increase produces 200% layer cost increase—but aggregation by attachment point and jurisdiction is rarely real-time. hx aggregates tower exposure by attachment band and nuclear verdict jurisdiction, enabling actuaries to stress-test rate adequacy under social inflation scenarios and quantify concentration in high-severity venues (CA, FL, NY, TX produce 50% of nuclear verdicts).

Audit trails for evolving regulatory requirements

With increasing regulatory scrutiny, actuaries need documented lineage from model assumptions to individual policy pricing decisions. hx captures every action automatically, creating the governance trail Excess & Umbrella's regulatory environment demands.

Lloyd's has emphasized inflation considerations in market communications, but Excel-based workflows can lack audit trails linking rate changes to supporting analysis. hx maintains version-controlled lineage from ILF table updates and severity trend assumptions through to layer-specific rate changes, providing Lloyd's-ready documentation of inflation methodology and actuarial judgment at the individual layer level.

Explore hx for Excess & Umbrella insurance →

This guide is part of Hyperexponential's insurance pricing resource library. For more information on how hx supports Excess & Umbrella pricing, contact us.