Aon Plc Class A AON

Thesis

Aon is the world's number-two insurance broker and risk advisor, an asset-light intermediary that sits between corporates and insurance carriers, collecting commissions and fees on the placement of property/casualty, reinsurance, health, and retirement risk. The economics are beautiful in steady state: ten-year average ROIC of 16.9%, FCF conversion of 1.29x (cash earnings exceed reported earnings), and a ten-year share count change of -2.64% — modest but real per-share compounding on top of mid-single-digit organic growth. The business is a textbook 'capital-light compounder': it owns no underwriting risk, no inventory, no factories — just relationships, data, and an actuarial brain. Buffett's 'four insurance disciplines' [1] describe carriers, not brokers, but the broker captures a slice of every premium dollar without bearing the loss exposure.

The problem is the price and the recent capital allocation choice. AON paid roughly $13B for NFP in 2024 to push down-market into the U.S. middle-market, and the after-tax marginal returns are showing in the math: 5-year ROIIC of 6.27%, far below the 10-year ROIC of 16.9%. The reverse-DCF embedded in $311.51 implies 7.56% growth in perpetuity — plausible but not conservative. Net-debt/EBITDA of 3.30x is the highest in Aon's modern history, draining the optionality that made this franchise great. IV-base is $259.72; IV-high is $329.88; price/IV is 1.20x. The only way to underwrite this at $311 is to assume NFP integrates cleanly and the high IV becomes the central case. That is a bet, not a margin of safety.

A Buffett-Munger investor pays up for quality only when the alternative is permanent capital impairment. Here the franchise is excellent, but $311 already prices the bull case. Buy below $230 (a roughly 11% discount to base IV); trim above $330 where even the bull case is exceeded.

Moat

1) Switching costs (the biggest moat). Aon's core product is the renewal of a multi-line insurance program for a Fortune 1000 client. The broker holds the loss-history database, the actuarial models, the claims relationships with the carriers, and the customized policy wording accumulated over years. Replacing a broker mid-cycle means re-marketing every line, re-loading every loss triangle, re-negotiating every wording, and accepting a quote-fatigue penalty from carriers. Industry retention rates routinely exceed 90% for Aon and Marsh, and the renewal cadence is annual, which means every year a client pays the switching cost rather than paying it once. This is the same dynamic Buffett identified for sound underwriting [1] — except Aon captures the relationship rents without bearing the underwriting losses.

2) Intangibles — brand and regulatory. Aon's brand carries weight in the boardroom; CFOs pay for the comfort that the world's largest reinsurance broker placed their D&O tower. Regulatory licensure (state-by-state in the U.S., FCA in the U.K., dozens of national regulators globally) is a non-trivial barrier to a new entrant trying to operate at Aon's scale. Buffett's 1981 letter [4] notes that the rare businesses with both pricing power and capital-light reinvestment are 'fierce to the point of being self-defeating' to acquire — the broker franchise is one of those.

3) Cost / scale advantages in data and reinsurance. Aon's Reinsurance Solutions and Health Solutions arms run on proprietary catastrophe models, mortality data, and benchmarking databases that smaller brokers cannot replicate. When a client wants to know what its peer group pays for cyber coverage, Aon has the answer; a regional broker does not. This is a flywheel: more clients → more data → better advice → more clients. Stress test: could a $10B competitor with five years of capital displace Aon? No — they could buy a top-20 broker (this is precisely what NFP was when Aon bought it), but they could not replicate the global reinsurance placement franchise or the Fortune 500 relationship book. The 'Big Three' (Marsh McLennan, Aon, WTW) plus Gallagher form an oligopoly that has been stable for 20+ years and is consolidating, not fragmenting.

4) Network effects — modest. Aon's Risk Capital data has weak two-sided network properties: more carrier participation makes Aon's placements more competitive, which attracts more clients, which gives carriers more reason to engage. But this is asymmetric — carriers need the broker more than the broker needs any single carrier. Not a primary moat.

5) Pricing power — limited at the line item, real at the bundle. On a single placement, brokerage commissions are competitive (often 10-15% of premium), and large clients fee-negotiate aggressively. But the bundle — risk advisory + claims + analytics + benefits + reinsurance — is sticky and re-prices upward with premium inflation. In a hard market (premiums rising), Aon's commission revenue grows automatically without any salesforce action. This is the 'inflation hedge' Buffett described in 1981 [5] — ability to grow dollar revenue with minimal additional capital.

Erosion risks. The structural threats are: (a) carriers disintermediating large clients via direct placement (slow-moving, has been threatened for decades, has not happened materially); (b) parametric and embedded insurance growing fast and not requiring traditional brokerage; (c) AI-driven underwriting compressing the analytical labor that justifies broker fees; (d) regulatory action on contingent commissions (Spitzer-era risk redux). None are imminent, but in a 10-year frame, (c) is the one to watch — if AI commoditizes actuarial analysis, the 'data moat' narrows. Berkshire's Buffett explicitly trimmed broker exposure (per the prompt) which is a signal worth weighing on (a) and (c).

Moat verdict: WIDE.

Management & Capital Allocation

Greg Case has been CEO since 2005 — twenty years, which already passes Buffett's 'long-tenured operator' filter. Christa Davies has been CFO since 2008. The duo has overseen the corporate inversion to the U.K. (later to Ireland), the divestiture of the Aon Hewitt outsourcing business in 2017 (a clean unwind of a low-return acquisition), the Willis Towers Watson merger attempt that DOJ killed in 2021, and now the $13B NFP acquisition closed 2024. Communication is structured (Aon United operating model, fee-vs-commission disclosure) but heavy on jargon; the 'Risk Capital and Human Capital' framing is corporate-glossy rather than Buffett-plain.

Capital allocation choices, ranked:

1. Buybacks — historically excellent, recently paused. Share count is down 2.64% over ten years, which understates the truth: prior to NFP, Aon was retiring 3-4% of shares annually, often well below intrinsic value. The buyback program has been suspended/throttled since the NFP announcement to deleverage. This is the right choice tactically but means the per-share compounding engine is off for 2-3 years.

2. Acquisitions — the NFP question dominates. NFP closed April 2024 for $13.4B ($10.6B cash + stock). The strategic logic is sound: NFP gives Aon a middle-market U.S. distribution channel where Aon historically under-indexed. The price logic is more debatable — at roughly 16-18x EBITDA on a peer-group basis, it is not a steal, and the 5-year ROIIC of 6.27% suggests marginal returns have already collapsed below the 16.9% historical ROIC. Aon also divested NFP Wealth in Q4 2025 (proceeds ~$1.4B sat on the balance sheet at YE2025; the Q1 2026 cash balance fell to $1.4B, implying redeployment to debt paydown and ongoing buybacks). The 'Aon Restructuring Program' charges of $147M in Q1 2026 indicate integration is still consuming real cash.

3. Debt — the biggest red flag. Net-debt/EBITDA of 3.30x is uncomfortable for an asset-light services firm; the historical Aon ran nearer 2.0-2.5x. Interest coverage was not provided by the scorer (null), which itself is a yellow flag — likely because TTM interest expense post-NFP is materially higher and rolling. A recession that compresses commission growth while interest stays high would squeeze FCF.

4. Reinvest organically — capex-light by nature. Aon's reinvestment is mostly people and technology; the FCF conversion of 1.29x confirms there is little capex drag. This is the cleanest part of capital allocation.

5. Dividends — modest, ~1% yield. Appropriate; Aon is not a dividend story.

The honest grade. Buffett's 1981 [4][5] warning is directly applicable: leaders with 'animal spirits' overpay in high-premium takeovers. NFP is exactly that — a transformational deal at a full price, financed with debt, ahead of an uncertain integration. On the other hand, Case's twenty-year track record includes Aon Hewitt divest (good), Willis-failed (sunk legal cost but no permanent damage), and consistent below-IV buybacks. He has earned benefit of the doubt but is using it up.

Capital allocator: B. Long-term track record is A-tier; the NFP bet drags it to B. Upgrade to A if NFP integration delivers >12% ROIIC by 2027; downgrade to C if leverage stays >3.0x in 2027 with organic growth below 5%.

Industry Structure

Buyer power — Moderate. Aon's clients range from Fortune 100 (high power, fee-based, aggressive negotiation) to middle-market (low power, commission-based, sticky). The post-NFP mix shifts toward middle-market, which strengthens Aon's pricing position because middle-market clients lack the procurement sophistication of large enterprises. Annual renewal cycles give buyers regular leverage but switching frictions blunt it. Net: moderate, trending favorable post-NFP.

Supplier power — Low to Moderate. The 'suppliers' are insurance carriers. In soft markets carriers compete for Aon's business and pay higher commissions; in hard markets carriers can dictate terms. The Big Three brokers control enough premium flow that no single carrier can dictate to them — Aon has channel power. The reinsurance market is more concentrated, but Aon's Reinsurance Solutions arm is itself one of three meaningful global players, balancing the dynamic.

Threat of new entrants — Low. Capital is not the barrier (broking is asset-light); the barriers are licenses, data, talent, and relationships. A startup cannot replicate a 100-year client book or a global reinsurance placement capability. Insurtech entrants (Embroker, Newfront, etc.) have nibbled the small-business segment for a decade with limited displacement of mid-market and enterprise. Private equity rollups (Acrisure, HUB, Alliant) are real entrants in middle-market — this is why Aon needed NFP to defend that segment.

Threat of substitutes — Low to Moderate (rising). Direct placement (carrier-to-corporate) has been a 30-year boogeyman that has materialized only at the small-commercial end. Captive insurance for very large corporates removes some risk from the brokered market but Aon often advises on captive formation. The medium-term substitute risk is parametric insurance (no claims-handling intermediation needed) and AI-driven self-service for SMBs. Risk grade: modest today, moderately higher in 10 years.

Rivalry — Moderate, oligopolistic. The Big Four (Marsh McLennan, Aon, WTW, Gallagher) plus the PE rollups define the competitive set. Pricing is professionally rational — large brokers do not cut commissions to grab share, they compete on capability and relationship. The 2021 Aon-WTW merger attempt and its DOJ block locked in the four-firm structure. In oligopolies with switching costs and slow share movement, returns on capital tend to be high and stable — exactly what Aon's 16.9% ten-year ROIC reflects.

Value pool location and trajectory. The global insurance broker fee pool is roughly $80-100B and grows mid-single-digits with global premium inflation. Aon captures roughly 8-10% share. Within the pool, value is shifting from pure placement (commoditizing) toward analytics, benefits administration, and reinsurance advisory — areas where Aon over-indexes. Health & benefits, in particular, is a secular grower as employer-sponsored healthcare costs balloon.

Industry Verdict: Good. Not 'Excellent' because (a) AI/parametric substitution risk is real on a 10-year frame, (b) PE-funded mid-market consolidation has compressed acquisition multiples for the public players (Aon paid up for NFP precisely because Acrisure et al. were bidding too), and (c) Berkshire's reduction in broker exposure is a tell from the smartest insurance allocator alive. 'Good' is the right grade for a stable oligopoly with one structural cloud on the horizon.

Inversion (Bear Case)

The single event that kills this thesis. A 2026-2027 recession (or a 'higher for longer' rate environment that crimps middle-market customer growth) collides with NFP integration shortfalls and elevated leverage. Aon enters such a downturn with net-debt/EBITDA of 3.30x — already 30-50% above its historical comfort range. If organic commission growth stalls from mid-single-digits to flat or down 2% (as in 2009), interest expense stays high (refi at 5%+), and integration synergies slip, the dividend and buyback get further restricted, the credit rating moves to BBB-, and the multiple compresses from 26.6x to 18-20x. That math — 18x on $10 of FCF/share that grew slower than promised — is a $180-200 stock. Catalyst: a single bad quarter where management is forced to widen its leverage timeline.

Why the moat is narrower than bulls think. The 'Big Three plus rollups' oligopoly story masks two erosion vectors. First, the middle-market segment Aon just paid $13B to enter is being attacked by PE-funded rollups (Acrisure at $80B revenue and growing, Hub, Alliant) using debt-financed roll-up math that public companies cannot match in the long run. Aon bought NFP partly because it had to, not because it wanted to — the moat in middle-market never existed for Aon, and now it is paying retail multiples for a segment where structural returns may be permanently 6-8%, not 16-17%. Second, the 'data moat' in placement and benchmarking is exactly the kind of analytical work LLMs can perform at 1/10th the cost; carriers and large self-insured employers can increasingly cut Aon out of analytical work that historically justified fees. The 5-year ROIIC of 6.27% is not a transient NFP-integration drag — it may be the new run-rate for marginal capital deployment.

Why management is worse than it appears. Twenty years of Greg Case is a feature in steady state and a bug at inflection points — long-tenured CEOs reliably overestimate their ability to integrate transformational acquisitions. Aon Hewitt was a 2010 mistake that took seven years to unwind. The 2021 WTW deal was killed by DOJ but management spent two years and ~$1B in deal costs pursuing it. NFP is the third large strategic bet in 15 years; pattern recognition says management's risk appetite for transformational M&A is higher than shareholders should want. The CFO's 'capital structure flexibility' communication is corporate-speak for 'we levered up and need 2-3 years to deleverage' — buyback throttling at exactly the moment when Aon stock is closest to fair value (or below it) is destruction of optionality. The 'Aon United' operating model jargon obscures real headcount and integration costs (the $147M restructuring program in Q1 2026 alone signals integration is harder than guided).

What bulls are extrapolating that won't hold. Bulls extrapolate (1) mid-single-digit organic growth perpetually, (2) NFP synergies hitting plan by 2027, (3) gradual deleveraging restoring buyback to 3% annually, and (4) the 26.6x multiple being justified by 'asset-light compounder' framing. Each is plausible individually; together they imply IV-base of $260-330 — but the reverse-DCF embedded in $311 is 7.56% growth in perpetuity. Insurance brokerage organic growth has averaged ~5%; 7.56% requires permanent above-trend, plus margin expansion, plus stable multiple. That is a stack of optimism. If any one fails, the stock pulls back to IV-base $260 (-17%); if two fail, it pulls back toward IV-low $137 (-56%).

Valuation trap (multiple compression / regime change). The TTM P/E of 26.64 sits below the 10-year average of 32.18, which bulls cite as 'cheap relative to history.' This is the wrong frame. The 10-year average reflects a zero-rate-environment compounder bid that no longer exists. In a 4-5% risk-free-rate world, 'asset-light compounder' multiples globally have re-rated 25-30% lower. EV/FCF of 30.7 is full pricing for a business growing at 5-7%. A reversion to a 20-22x P/E (peer median for slower-growing financials) on $11-12 of EPS yields $220-260 — at or below current IV-base. Berkshire's reduction in broker exposure is a thoughtful insurance allocator's vote that the regime has changed.

If I am right, the stock could be worth $190 within 24-36 months.

Lollapalooza Bias Check

Authority bias. The scorer hands me a 66 composite and a clean spreadsheet of 'great franchise' metrics. Berkshire owned Aon recently — that is authority on the bull side. Berkshire trimmed brokers — that is authority on the bear side. I notice I am cherry-picking the bear authority here because it suits my conclusion; I should treat both signals symmetrically. Net: the 'Buffett owns insurance brokers' frame is real but Buffett-Munger-Ajit have been quite specific that they like float-bearing carriers (Geico, GenRe), not commission-clipping intermediaries. The authority pull on this analysis is mild and roughly balanced.

Anchoring on price relative to IV. Px/IV of 1.20x feels like a clean 'overvalued' anchor and pulls me toward 'wait for $230.' But the IV range is wide — IV-low $137 to IV-high $330. The scorer's note that 'maintenance capex uncertain (>50% spread); widen IV range' is a flag I should not skip past — it means the IV-base anchor is itself uncertain. I have anchored on $260 base case as if it is precise; honestly, the range $200-$300 is roughly what the data supports. This argues for not being overly confident on either direction.

Recency bias around NFP. The NFP deal closed 18 months ago and integration restructuring charges are still being booked ($147M in Q1 2026 alone). I am extrapolating recent integration drag into permanent reinvestment-math weakness. It is possible that 2027-2028 produce structurally better marginal returns once integration laps. Symmetric possibility: it is also possible that 5-year ROIIC of 6.3% is the new baseline. I cannot tell from the data, which is itself a reason to demand a wider margin of safety.

Confirmation bias from the prompt. The user prompt explicitly mentioned 'Buffett MS reduction in brokers' — that is a thumb on the bear scale. I notice I have built the inversion section with more energy than the bull case. I should explicitly note the bull case has structural integrity: oligopoly, sticky clients, asset-light, mid-single-digit pricing power, durable ROIC. A 'Hold' or 'Trim' recommendation rather than 'Sell' reflects this asymmetric humility.

Commitment / consistency from the format itself. The 12-step pipeline rewards a definite answer. I notice a pull to manufacture conviction where the underlying data supports 'medium' at best. I am setting conviction explicitly to medium and recommendation to Hold to honor the actual epistemic state.

10-Year Outlook

Will Aon have the same fundamental business model in ten years? Probably yes at the core (placing complex risk for large enterprises), probably partially eroded at the edges (analytical work commoditized by AI, small-commercial cannibalized by direct/embedded insurance). The Risk Capital and Human Capital framing is durable; the value-pool composition will shift toward advisory/analytics and away from commodity placement.

Will the customer base be larger? Yes — global insurance premium volume grows with GDP plus inflation plus secular risk expansion (cyber, climate, supply chain). Aon's middle-market footprint via NFP also expands the addressable client count by an order of magnitude.

Will profit per customer be higher? Probably modestly. Inflation in premium translates into commission growth at near-zero marginal cost. Cross-sell of analytics and benefits raises wallet share. Offset: AI-driven price transparency may compress fees on commodity lines.

Will the moat be wider? Likely flat-to-narrower. The relationship and data moats remain strong for enterprise; the analytical-labor moat erodes for everyone in advisory services. Moat verdict: narrower at the edges, intact at the core, net narrower.

Single biggest threat: AI-driven analytical commoditization combined with PE-funded middle-market roll-up competition compressing structural returns from 17% historical ROIC toward 12-13% over a decade. The leverage from NFP is the secondary threat — it removes the optionality that previously let Aon compound through downturns by buying back stock cheaply.

Honest read: this is a durable, high-quality business that probably exists and pays a higher dividend in 2036, but the per-share compounding rate is more likely 7-9% than the 11-13% the bull case implies. At $311 that is not enough margin of safety. CONFIDENCE: medium.