Marsh McLennan and the Price of a Durable Moat
What the business is
A good insurance broker resembles good plumbing: scarcely noticed until something complicated begins to leak.
Marsh McLennan sits between corporate clients and the insurers and reinsurers willing to assume their risks. It designs coverage, negotiates terms, places business, and earns commissions linked to premiums alongside advisory fees. In 2025, the company generated about $27 billion of revenue across 130 countries with 95,000 employees.
Risk & Insurance Services contributes about 64% of revenue. It includes insurance broker Marsh and reinsurance broker Guy Carpenter. Consulting supplies about 36% through Mercer’s health, benefits, and pension work and Oliver Wyman’s strategy consulting.
This is a capital-light model. The thesis records 2025 free cash flow of $5 billion, approximately 18% of revenue. The real productive assets are accumulated risk data, regulatory licences, insurer relationships, and the professional judgment needed to structure unusual coverage. Large corporate programmes carry years of renewal history, while changing brokers introduces the possibility of coverage gaps. Switching can therefore be more troublesome than an orderly procurement spreadsheet implies.
Scale adds another layer. Large placement volumes strengthen negotiating power with insurers and produce market data that can attract more clients. Guy Carpenter operates in a reinsurance market described by the thesis as effectively a duopoly with Aon. Marsh McLennan also brings a 154-year brand and a licensing network spanning those 130 countries.
Artificial intelligence cuts both ways. Complex corporate, reinsurance, cyber, climate, and AI-liability risks still require advice and accountable judgment, while proprietary placement data may become more useful when paired with better analytical tools. Standardised small-business brokerage is more exposed to automation and direct distribution.
That distinction matters because the 2024 McGriff acquisition enlarged Marsh McLennan’s American middle-market presence. AI may reinforce the sophisticated centre of the franchise while eroding parts of its expanded periphery. I regard the net effect as uncertain rather than fatal. Moats are rarely removed by announcement; they tend to be worn down one renewal at a time.
What the numbers say
The operating record is strong. The thesis cites 17 consecutive years of adjusted operating-margin expansion, although softer commercial-insurance pricing slowed Marsh’s organic growth from 7% in 2024 to 4% in 2025. The screening evidence also reports a five-year average return on equity of 28.8%.
Annual owner earnings are $4.23 billion. Against the latest closing price, the valuation reads:
- Market cap (price-adjusted, 2026-09-16 close): $82.89 billion
- Owner-earnings yield (price-adjusted): 5.10%
- Share price (2026-09-16 close): $176.76
- 10-year Treasury: 5.01%
The 5.10% owner-earnings yield therefore sits only 0.09 percentage points above the 5.01% Treasury. That satisfies the rubric’s watch condition, but it leaves little compensation for ordinary business disappointments, let alone expensive ones.
The intrinsic-value evidence requires caution. Its parsed field lists a range of $4.2–4.2 billion, while the thesis describes a framework applying 12 to 15 times owner earnings. Those items do not connect cleanly enough for me to treat the range as a reliable independent valuation anchor. The deterministic evidence nevertheless records the current market cap at a 1860% premium to the parsed low end, so the rubric’s circle condition is plainly unmet.
Capital allocation deserves mixed marks. The thesis says the dividend rose from $0.90 in 2012 to $3.05 in 2024 and increased another 10% in 2025, while cumulative repurchases reached $11 billion. Yet acquisition activity in 2024 totalled about $27 billion, and debt-to-equity reached 1.28 following the McGriff financing. A business that produces abundant cash can carry debt; that does not make every acquisition price sensible.
The key question is whether future cash generation is used to reduce leverage and strengthen the franchise rather than to pursue further large transactions at rich valuations. Capital-light businesses occasionally tempt managers to forget that purchase prices are still made of capital.
What Wall Street believes vs what I see
There is no live consensus in the supplied material. The available Seeking Alpha headlines are favourable—Marsh & McLennan: Maintaining My ‘Buy’ Recommendation Following Q2 Earnings, Marsh & McLennan: 16 Years Of Dividend Growth And Counting, and Marsh & McLennan: The Valuation May Finally Be Here—but no publication dates or authors are provided. Three positive article titles are a mood sample, not a market census.
The favourable case is understandable. Marsh McLennan has recurring client relationships, high returns on equity, expanding margins, substantial free cash flow, and exposure to risks that are becoming more complicated. Its data, scale, licensing network, and embedded advisory relationships should remain difficult to replicate.
What I see is a fine business offered with a very small arithmetic cushion. A 5.10% owner-earnings yield against a 5.01% Treasury supplies a spread of only 0.09 percentage points. That asks the moat to endure, McGriff integration to proceed sensibly, leverage to decline, and AI disruption to remain concentrated in lower-complexity work.
I would watch Risk & Insurance Services organic growth, adjusted operating margins, large-client retention, low-complexity commission trends, free cash flow around the thesis reference level, and debt-to-equity. In particular, the thesis calls for reassessment if organic growth remains below 2% for two consecutive years or if debt-to-equity does not fall below 1.0 by 2027. A polished presentation cannot renew an insurance programme.
Verdict
Verdict: watch.
The 5.10% owner-earnings yield exceeds the 5.01% Treasury, satisfying the rubric’s watch condition, but the 0.09 percentage-point spread is slender beside the leverage and AI uncertainties. Circle is unavailable because the yield does not clear the required two-percentage-point hurdle and the market cap does not meet the requirement of being no more than 70% of the intrinsic range’s low end. A materially wider owner-earnings spread, satisfaction of the rubric’s market-value discount, or clear evidence of lower leverage and durable organic growth would change the verdict.
This analysis is AI-generated, educational, and not investment advice. Figures may contain errors or be delayed. Disclaimer