Agencies, Brokerages, and Other Insurance Related Activities (U.S.) — NAICS 5242

A Histometrics rollup primer for public-market and private investors

1. Overview

Every insurance policy is created by a carrier that takes on the risk — but almost everything around that policy is done by someone else, for a fee. Someone sells and renews it. Someone prices, rates, and models the risk behind it. Someone runs the drug benefit or the claims back office. Someone investigates and settles the loss when it happens. NAICS 5242 — "Agencies, Brokerages, and Other Insurance Related Activities" is the U.S. federal statistical bucket for all of that non-risk-bearing work. (NAICS, the North American Industry Classification System, is the standard code set U.S. statistical agencies use to define industries.) [3]

The unifying idea across the whole group is simple and powerful: none of these firms underwrite risk. They take no premiums onto their own balance sheet, hold no policyholder reserves, and carry no catastrophe or investment "float." They earn commissions, fees, subscriptions, or spreads for a service, and they get paid regardless of which carrier ultimately wins the customer. That makes NAICS 5242 the "picks-and-shovels" layer of insurance — capital-light, labor- and data-intensive, recurring in its best corners, and structurally different from the carriers it serves (which live one level over in NAICS 5241) [3]. For an investor, the appeal is exposure to insurance activity — premiums written, claims processed, drugs dispensed — without exposure to hurricanes or claims spikes.

But the group is deceptively lopsided. It contains just two child industries, and they could hardly be more different: one is a swarm of 122,000-plus small brokerages earning commissions; the other is a handful of giant pharmacy-benefit and claims-administration firms whose reported revenue is swollen by drug dollars merely passing through. The distinctive job of this primer is the contrast — how the two children differ in size, growth, ownership, concentration, and how you actually reach them.

2. What's inside — the two child industries and how they differ

At the 5-digit "industry" level, code 5242 splits into two children:

• 52421 — Insurance Agencies and Brokerages: the distribution layer — retail agents and brokers, wholesale brokers and managing general agents (MGAs), and benefits/personal-lines/digital agencies that place and service coverage for a commission. [3]
• 52429 — Other Insurance Related Activities: everything else done around insurance for a fee — claims adjusting, pharmacy benefit management (PBM) and third-party administration (TPA), and advisory/rating/actuarial/modeling work. [3]

They share the "no underwriting risk" DNA, but almost nothing else. Here is how they compare.

(Receipts and firms are 2022 Economic Census; jobs and establishments are 2023 County Business Patterns — see §3. HHI is the Herfindahl-Hirschman Index, the sum of squared market shares on a 0–10,000 scale; U.S. antitrust agencies treat below ~1,500 as "unconcentrated." Tickers are defined and used in §4 and §10.) [1][2][3][6]

The four contrasts that matter:

1. Revenue says 52429 is bigger; everything else says 52421 is. By reported receipts the split is roughly 60/40 toward 52429. But that 60% is an illusion of accounting: the overwhelming majority of 52429's revenue is drug costs flowing through pharmacy-benefit managers on their way to pharmacies — money the PBM never keeps. Measured by the things that are harder to distort — jobs (62% vs. 38%), establishments (90% vs. 10%), and firms (~93% vs. ~7%) — the agencies-and-brokerages child is overwhelmingly the larger, more populous business. Read 52429's revenue share as "money flowing through," not "economic value captured." [1][2]

2. Concentration is opposite at the two poles. 52421 is one of the least concentrated industries in the entire economy — an HHI near 60 and a top-four share of just 11%, a base so deep that even record-breaking mergers barely move it. 52429 is topped by an oligopoly where it counts — three PBMs process about 80% of U.S. prescriptions (HHI ~1,449). The blended group HHI of 548.8 hides this barbell entirely. [2][6]

3. Growth points in different directions. Brokerage commissions track the insurance rate cycle, which has just turned down (commercial property-and-casualty prices slipped for the first time in nearly nine years). PBM/TPA rides structural drug-spend and employer-self-funding tailwinds — while simultaneously absorbing the fiercest regulatory assault in the group. Adjusting is catastrophe-driven and flat-to-cyclical; advisory/data grows steadily with premiums and hard-to-price risk. [6][8][12]

4. Ownership mix diverges — but private capital dominates both. 52421 is a barbell of a few listed brokers over a vast field of PE-backed platforms and independent shops. In 52429, the largest profits are buried inside diversified health insurers (the PBMs), the cleanest economics sit in one listed data franchise, and the biggest claims administrator is private-equity-owned. In neither child is the public market where most of the industry lives.

3. How big it is (the rollup figures)

Our ground-truth federal figures for NAICS 5242 blend the 2022 Economic Census (money and concentration) with 2023 County Business Patterns (CBP — counts, jobs, payroll). They are different vintages, not one year's financial statement, so receipts, firms, and employment are not interchangeable.

These are the figures in our ground-truth stats extract for this level [1][2]. Average pay works out to roughly $88,000 per employee (annual payroll ÷ employment) — a blend running from small-town personal-lines agents up through credentialed actuaries and PBM pharmacists. The group is "unconcentrated" on the standard test (HHI 549, well under 1,500), but that single number is deeply misleading — see the caveats below. [1]

The children reconcile to the level. Establishments (133,728 + 14,174) sum to exactly 147,902, and paid employees (818,561 + 491,943) to exactly 1,310,504 — line for line. Receipts ($194.3B + $301.1B ≈ $495.4B) and payroll ($74.35B + $40.8B ≈ $115.2B) reconcile within rounding. Firm counts sum to 131,167 versus the level's 131,043 — a tiny overlap because a firm operating in both children is counted once at the level but can appear in each child. So the two children genuinely partition the group. [1][2]

Undercount and interpretation caveats — the headline figures mislead in three directions:

• The group is unconcentrated overall, but the CR4 is a PBM artifact. The top-four share of 44.5% looks moderate, but it comes almost entirely from a handful of pharmacy-benefit giants sitting in 52429 — the top four firms of the whole group are essentially the top PBMs. Strip them out and what remains — 122,000 brokerages plus thousands of adjusters and advisers — is one of the most fragmented service economies in the country (52421's HHI is under 60). The blended HHI averages a swarm and an oligopoly into a number that describes neither. [2][6]

• Receipts are inflated at one end and understated at the other. 52429's receipts include pass-through drug costs the PBM never keeps, so the $495B overstates fee/service revenue; separately, 52421's receipts are broker compensation (commissions and fees kept), not the trillions of premium dollars that flow through brokers to carriers, so the same figure understates the industry's economic footprint. The group's dollar total is a distorted proxy, not a clean measure — the reliable signals here are the structural facts (two very different businesses, their headcounts, their concentration shapes). [1][2]

• A large small-operator and misclassified tail is missing. CBP counts only establishments with paid employees and excludes the self-employed [4]. Many single-agent brokerages, captive agents, solo public adjusters, and independent actuaries operate as sole proprietors, so the true count of people working in insurance distribution and services is meaningfully higher than 1.31 million — a small-operator undercount, not evidence that a few firms or government dominate. Meanwhile the marquee players are classified out of the code: the biggest PBMs are reported inside their parents' health-insurer codes, and the leading insurance-data franchise (Verisk) sits in the Information sector, not 52429. Our extract does not include Nonemployer Statistics for this level, so the solo-producer population is not stated here. [1][6]

• Our supplied figures do not include industry-wide profit, EBITDA (earnings before interest, taxes, depreciation, and amortization), premium volume, covered lives, or a public/private ownership split, so those are not stated as federal facts.

4. The investable universe (where value concentrates across the children)

The recurring theme across both children is the same: clean public pure-plays are scarce, and much of the quality is private. But value concentrates in very different places.

• 52421 (Agencies & Brokerages) offers the cleanest public exposure to the group's core idea — capital-light, recurring commission revenue with no underwriting risk. Value sits at two poles: a small cluster of large, high-quality listed brokers — Marsh McLennan (MMC), Aon (AON), Arthur J. Gallagher (AJG), Brown & Brown (BRO), and Willis Towers Watson (WTW) — plus specialty, personal-lines, and digital-agency names (Ryan Specialty (RYAN), Goosehead (GSHD), and the Baldwin Group). But most of the industry and most of the growth are private: the large PE-backed platforms (Acrisure, Hub International, USI, Alliant, Lockton, OneDigital) rolling up thousands of local agencies, reachable through private equity, private credit lent to the consolidators, or direct agency ownership. [5][15]

• 52429 (Other Insurance Related) holds the most reported revenue and profit — but almost none of it is a pure-play. The PBM economics live inside diversified health giants — UnitedHealth Group (UNH) (Optum Rx), The Cigna Group (CI) (Express Scripts/Evernorth), and CVS Health (CVS) (Caremark), with Elevance (ELV) and Humana (HUM) more peripheral — and the disruptive transparent-PBM challengers are private. The cleanest economics in the whole group belong to the advisory/data corner: subscription-based, data-moated, near-zero marginal cost, led by listed Verisk Analytics (VRSK) alongside insurance-software names (CCC Intelligent Solutions (CCCS), Guidewire (GWRE)), with Milliman (employee-owned), AM Best (private), and member-governed AAIS/NCCI not publicly investable. Claims adjusting has the most firms but the least revenue, and essentially one listed pure-play — Crawford & Company (CRD-A/CRD-B) — plus workers'-comp specialist CorVel (CRVL); the largest administrator, Sedgwick, is PE-owned. [6][7][10]

Cross-cutting reality. Two forces own the private tier across the whole group: private-equity roll-ups (of agencies in 52421; of claims/TPA platforms like Carlyle-controlled Sedgwick in 52429) and, uniquely in PBM, vertical integration into insurers. There is no dedicated exchange-traded fund (ETF) or index for the group or either child; broad insurance, financials, and health-care funds hold these names only in small weights. Tickers and valuations are developed in §10.

5. How the money works

The single throughline is fees or commissions for a service, not underwriting. Absent everywhere in NAICS 5242: premiums earned, the combined ratio, reserves, and investment float — those belong to the carriers in 5241. Present everywhere: labor, data, technology, and client-retention economics. But the fee mechanism differs sharply between the children.

• 52421 (Agencies & Brokerages) earns a slice of the premium that passes through it: base commission (a percentage of premium, paid by the carrier, renewing each year), fees on large or consulting accounts, volatile contingent commissions (carrier bonuses for volume, growth, and low loss ratios), wholesale/MGA overrides, and interest on premiums briefly held in trust. The insurer keeps the risk throughout. The metrics that matter are organic growth, client and producer retention (90%+ at the best books), EBITDA margin (~25–30%+), and the roll-up multiple arbitrage — buying small agencies at ~7–13× EBITDA into a platform valued far higher. [5][15]

• 52429 (Other Insurance Related) runs three different meters. Claims adjusting earns fee-per-claim or daily-rate revenue, with roughly half of every dollar going to payroll — labor-intensive and catastrophe-sensitive. TPAs earn recurring per-member-per-month (PMPM) administrative fees on self-funded plans (asset-light, sticky, software-like). PBMs layer on manufacturer rebates, spread pricing, and owned-pharmacy dispensing margin — thin per-script margins on colossal volume, with reported revenue mixing pass-through drug cost and true service revenue. Advisory/data runs on recurring subscriptions and data licensing (near-zero marginal cost), fee-for-service actuarial work, and issuer-pays ratings — the highest-quality economics in the group. [6][10]

Across both children the investor's discipline is identical: separate the true service/commission revenue from the money merely flowing through — the premiums a broker places, the drug costs a PBM passes on, the claim payments an adjuster disburses — then watch retention, recurring-revenue share, and revenue per producer.

6. Demand drivers

Demand rests on the size and complexity of insurance and health-benefit spending — but each child has its own swing factor.

• Premium volume = rate × exposure (drives 52421). Commissions rise when insurance prices harden or when insured activity grows (more payroll, property, revenue, vehicles). Structural tailwinds: new and worsening risks (cyber, artificial intelligence, liability), "social inflation" (rising jury verdicts) keeping casualty prices firm, and climate/catastrophe risk pushing business into specialty and excess-and-surplus (E&S) markets — all of which make an intermediary more valuable. [12]
• Drug spend and self-funding (drive 52429's PBM/TPA core). U.S. prescription-drug spending reached $467 billion in 2024 (up 7.9%) and keeps growing mid-single-digits; specialty and GLP-1 (glucagon-like peptide-1) obesity drugs raise the value of formulary management; and each employer that shifts to self-insurance adds PBM and TPA demand. [6][8]
• Catastrophe activity (drives claims adjusting). Hurricanes, wildfires, hail, and floods drive sudden surges in independent and catastrophe adjusting — the group's most volatile demand line. Federal data record 403 U.S. billion-dollar weather disasters from 1980–2024, and insured 2025 natural-catastrophe losses were estimated near $103 billion. [14]
• Hard-to-price risk and mandatory reporting (drive advisory/data). Climate, cyber, and litigation raise demand for modeling, reserving, and rate-adequacy work, while statutory filings create baseline demand across the cycle. [10]
• Artificial intelligence — a driver and a dampener. AI lifts demand for data and analytics vendors but trims the labor hours per routine claim (the federal labor forecast projects claims-adjuster employment down 5% from 2024 to 2034, citing AI photo-estimation) and threatens the simplest personal-lines distribution. [9]

7. Regulation

Insurance is regulated primarily at the state level, not federally — a structure rooted in the 1945 McCarran-Ferguson Act and coordinated loosely through the National Association of Insurance Commissioners (NAIC), which publishes model laws states may adopt. There is no single federal regulator for the group; the rulebook is activity-specific, and it is far heavier on one child than the other. [11][15]

• 52421 (Agencies & Brokerages): every agent and broker ("producer") must be licensed in each state where they operate, meet continuing-education rules, and follow sales-conduct standards; surplus-lines placements require a separate license and premium-tax filings; benefits brokers also fall under the federal Employee Retirement Income Security Act of 1974 (ERISA), the Affordable Care Act, and Department of Labor rules. This is a moderate barrier that favors scaled platforms but does not cap prices or returns the way utility or health-carrier regulation can. [11]

• 52429 (Other Insurance Related): the regulatory storm center. Adjusters face state licensing and (for public adjusters) fee caps. But because PBMs and TPAs hold the overwhelming majority of this child's revenue, they set the tone — and they are under the fiercest assault. The Federal Trade Commission (FTC) found the top three PBMs process ~80% of U.S. prescriptions, sued the big three over insulin rebates (reaching settlements by 2026), all 50 states have enacted PBM laws since 2017, and "delinking" (barring PBM pay from a drug's price) is spreading. Advisory/rating firms lean on McCarran-Ferguson's limited antitrust carve-out for data-pooling, and financial-strength raters are SEC-registered NRSROs (Nationally Recognized Statistical Rating Organizations). Data-security rules (the NAIC model law, HIPAA, state breach-notification laws) apply across the child. [6][7][10]

The net: regulation is a moderate, pro-scale barrier for brokerage, but intense, escalating, and politically charged for PBM/TPA — which is where essentially all of the group's regulatory risk sits.

8. Consolidation

One force dominates the entire group: private-equity (PE) roll-up. In 52421, PE-backed consolidators have been buying agencies at roughly 690–700 deals a year, with private equity involved in ~70% of transactions; 2025 brought the two largest strategic brokerage deals in history — Gallagher's $13.45 billion purchase of AssuredPartners and Brown & Brown's $9.8 billion purchase of Accession (Risk Strategies/One80). In 52429, PE-backed and hybrid buyers have likewise led insurance-services M&A, rolling small firms into platforms like Carlyle-controlled Sedgwick (adjusting + TPA) and Davies. [5][13]

Layered on top are two child-specific dynamics: vertical integration in PBM/TPA, where each big-three PBM is fused into an insurer that also owns pharmacies — capturing margin at the plan, the PBM, and the pharmacy at once (the group's most concentrated and most scrutinized structure); and technology and data as the new moat — AI damage-estimation, transparent-PBM software insurgents, and proprietary data franchises shifting competition from headcount toward software that scaled players can amortize and small firms cannot. [6][10]

Yet the striking fact is that all this deal-making barely dents the group's overall fragmentation. The base of 131,000 firms is so deep — and 52421's tail so long — that the group HHI stays under 550 even after record consolidation. The common outcome across both children is a barbell: a few scaled, technology-forward platforms above a fragmenting tail of small specialists that are prime roll-up targets. [2]

9. Risks

• Rate-cycle risk (52421's dominant one). Falling insurance prices directly cut premium-linked commissions; U.S. commercial premiums slipped ~1.2% in early 2026, the first broad decline in nearly nine years. Shrinking policy counts hurt even though the model is far less exposed to claim losses than a carrier's. [12]
• PBM regulatory/model risk (52429's dominant one). Because pharmacy-benefit management drives most of 52429's revenue, PBM reform — delinking, mandated rebate pass-through, spread-pricing bans, FTC enforcement — is the single biggest threat to that child's economics, and to the group's headline receipts. [6][7]
• Leverage and rich acquisition prices. The PE-owned platforms across both children carry acquisition debt sensitive to interest rates and integration risk, and public roll-ups have paid up. [5]
• Talent and books walking. Brokerage value walks out with departing producers; adjusting and actuarial work face experienced-labor scarcity and wage inflation. [5][10]
• Errors-and-omissions (E&O), conduct, and cyber litigation, plus custody of sensitive medical, financial, and claims data — a breach is a material-to-existential exposure across the group. [10]
• Disintermediation and automation. Insurtech and direct-to-consumer channels threaten the simplest personal-lines distribution; AI commoditizes routine claims and labor-heavy actuarial work even as it lifts the data-moated advisers. [9]
• Measurement/classification risk. As §3 shows, the federal dollar total for this group is distorted in both directions — analysts who anchor on $495 billion, or on the blended HHI, will mis-scale and mis-characterize the industry.

10. How to invest, and the outlook

Public-market routes (tickers reserved for here, per house style):

• The clean picks-and-shovels bet (52421): the listed brokers — Marsh McLennan (MMC), Aon (AON), Arthur J. Gallagher (AJG), Brown & Brown (BRO), Willis Towers Watson (WTW) — offer a high-quality, compounding financial-services business without underwriting risk, plus specialty/digital names (Ryan Specialty (RYAN), Goosehead (GSHD)). Compare them on organic growth, retention, producer economics, recurring-fee mix, EBITDA margin, free-cash conversion, leverage, and acquisition discipline; expect to pay premium multiples for the recurring revenue. [5][15]
• The revenue/profit heavyweight, indirectly (52429 PBM/TPA): only through diversified health giants — UnitedHealth Group (UNH), The Cigna Group (CI), CVS Health (CVS), and more peripherally Elevance (ELV) and Humana (HUM) — treated as integrated-managed-care bets whose PBM policy risk is now a first-order swing factor; Alight (ALIT) and CorVel (CRVL) are more targeted TPA/claims exposure. [6]
• The cleanest economics (52429 advisory/data): the nearest pure-play is Verisk Analytics (VRSK), with CCC Intelligent Solutions (CCCS) and Guidewire (GWRE) for insurance software, plus diluted exposure through RELX, S&P Global (SPGI), and Moody's (MCO). Claims adjusting is essentially one small-cap pure-play, Crawford & Company (CRD-A/CRD-B), with Carlyle (CG) and Onex (ONEX) offering indirect ownership of private Sedgwick. [10]
• There is no dedicated ETF or index for the group or either child.

Private-market routes. Much of the group's quality — and most of its growth — is unlisted: the PE-backed brokerage consolidators (Acrisure, Hub, USI, Alliant, Lockton), the claims/TPA platforms (Sedgwick, Davies), the transparent-PBM challengers, and employee-owned Milliman. The dominant institutional path is PE platform-and-add-on roll-up plus private credit lent to the consolidators. Diligence in every channel: recurring-contract share, renewal/retention, true service revenue versus pass-through, customer concentration, data security, licensing, technology ownership, and — for sponsor deals — acquisition accounting and debt. [5][6][10]

Near-term outlook (forward-looking judgments, not settled facts):

• The two children face opposite near-term swing factors. 52421's is the softening commercial insurance market, which pressures organic growth against an intact structural case (asset-light economics, recurring revenue, a still-fragmented base to consolidate, new risk categories). 52429's is PBM regulation — delinking and rebate reform could reshape incumbent economics, with the pace of employer defection to transparent models the other half of the story.
• AI is the group-wide variable — a tailwind for the data-moated advisers, a productivity boon but headcount/pricing pressure for adjusters and the simplest distribution.
• Catastrophe frequency and employer self-funding should keep feeding adjusting and TPA demand respectively.

Bottom line: NAICS 5242 is the defensive, fee-and-commission, capital-light services layer of insurance — the picks and shovels, not the risk. Its reported size is dominated and distorted by PBM drug dollars that make 52429 look the larger child, while by jobs, establishments, and firm count the story flips to a vast, fragmented field of brokerages. The cleanest public bet on the group's core idea is the listed brokers of 52421; the largest cash flows sit in 52429's PBMs, reachable only through diversified health insurers and squarely in the regulatory crosshairs; and the most durable economics belong to a short list of data-moated advisers. Across both children the pattern rhymes: scale, data, and technology increasingly favor a few consolidators — public where they exist, private where they don't.

---

Sources

Synthesized from the child primers (52421, 52429) and our ground-truth federal stats extract for NAICS 5242; citation numbers are this primer's own.

1. U.S. Census Bureau, County Business Patterns 2023 (NAICS 5242 and children 52421 / 52429: establishments, employment, annual and first-quarter payroll). 2025. https://www.census.gov/programs-surveys/cbp.html
2. U.S. Census Bureau, 2022 Economic Census — Concentration & selected statistics, NAICS 5242 / 52421 / 52429 (firms, receipts, CR4/CR8/CR20/CR50, HHI). 2025. https://www.census.gov/programs-surveys/economic-census.html
3. U.S. Census Bureau / NAICS, "5242 Agencies, Brokerages, and Other Insurance Related Activities; 52421 and 52429 definitions and adjacent codes." 2022. https://www.census.gov/naics/?input=5242&year=2022
4. U.S. Census Bureau, "County Business Patterns Methodology" (CBP covers employer establishments; excludes non-employers/self-employed). 2026. https://www.census.gov/programs-surveys/cbp/technical-documentation/methodology.html
5. MarshBerry / Risk & Insurance / Business Insurance, broker rankings and insurance-agency M&A ("Big Deals Drive Bigger Revenue for the Top 100"; ~690–700 deals/yr; ~70% PE participation; EBITDA multiples). 2025–2026. https://www.marshberry.com/resource/big-deals-drive-bigger-revenue-for-the-top-100-insurance-brokers/
6. U.S. Federal Trade Commission, Interim Staff Report on Prescription Drug Middlemen (top three PBMs process ~80% of U.S. prescriptions; top six >90%); NAICS 524292 stats. 2024. https://www.ftc.gov/news-events/news/press-releases/2024/07/ftc-releases-interim-staff-report-prescription-drug-middlemen
7. U.S. Federal Trade Commission / MultiState / Drug Channels, PBM enforcement (insulin settlements with Express Scripts and CVS/Caremark), state PBM legislation and "delinking." 2024–2026. https://www.ftc.gov/news-events/news/press-releases/2026/07/ftc-secures-major-settlement-caremark-resolving-antitrust-case-against-second-drug-middleman
8. Centers for Medicare & Medicaid Services, National Health Expenditure Fact Sheet (U.S. prescription-drug spending $467.0B in 2024, up 7.9%). 2026. https://www.cms.gov/data-research/statistics-trends-and-reports/national-health-expenditure-data/nhe-fact-sheet
9. U.S. Bureau of Labor Statistics, Occupational Outlook Handbook: Claims Adjusters, Appraisers, Examiners, and Investigators (projected −5%, 2024–34, citing AI photo-estimation). 2025. https://www.bls.gov/ooh/business-and-financial/claims-adjusters-appraisers-examiners-and-investigators.htm
10. Verisk Analytics / U.S. SEC; Milliman; AM Best; AAIS; NCCI, NAICS 524298 investable universe and advisory/data economics (Verisk FY2025 revenue ~$3.07B). 2026. https://www.globenewswire.com/news-release/2026/02/18/3240081/0/en/verisk-reports-fourth-quarter-and-full-year-2025-financial-results.html
11. NAIC (National Association of Insurance Commissioners), "Producer Licensing" and "Surplus Lines." 2025. https://content.naic.org/insurance-topics/producer-licensing
12. Risk & Insurance, "Commercial P&C Market Shifts Into Reverse as Soft Market Takes Hold" (U.S. commercial premiums ~−1.2%, first broad decline in ~9 years). 2026. https://riskandinsurance.com/commercial-pc-market-shifts-into-reverse-as-soft-market-takes-hold/
13. Insurance Journal, "Arthur J. Gallagher Completes $13.5B Acquisition of AssuredPartners" and "Brown & Brown to Acquire Risk Strategies/One80 Parent for $9.8B." 2025. https://www.insurancejournal.com/news/national/2025/08/18/836091.htm
14. National Oceanic and Atmospheric Administration (NCEI), Billion-Dollar Weather and Climate Disasters (403 events, 1980–2024); Insurance Information Institute, Facts + Statistics: U.S. Catastrophes (2025 insured cat losses ~$103B). 2025–2026. https://www.ncei.noaa.gov/access/billions/
15. U.S. SEC filings — Marsh & McLennan (MMC), Aon (AON), Arthur J. Gallagher (AJG), Brown & Brown (BRO), Willis Towers Watson (WTW). 2025–2026. https://www.sec.gov/cgi-bin/browse-edgar