← back to rankingMRSH · Marsh & McLennan Companies, Inc.
Insurance - Brokers · mkt cap $77.8B · calls: Q1 FY2026 vs Q4 FY2025
59.0 conviction · conf-adj 59
conf 6/10 partial
enthusiasm:27.0 · trend:8 · quantifies:0 · impact:0 · under_radar:5 · credibility:5 · business_impact:8 · disruption:0 · commitment:6 · confirmation:0
Enthusiasm latest 9 / prev 8 (rising)
Management's AI thesis expanded from Thrive-enabled productivity and digital infrastructure exposure in Q4 FY2025 to a three-pillar growth, productivity, and efficiency strategy in Q1 FY2026. Credibility improved because Q1 added concrete operating examples and quantified pilots, though management still declined to disclose total AI spend or direct revenue/margin dollars. The substance is broad-based: AI as client offering, internal automation, producer productivity, consulting demand, and potential broker consolidation advantage.
GROUNDED NEXT-FY IMPACT vs CONSENSUS
Grounded on actual base — revenue $27.0B · net income $4.2B · net margin 15.4% · diluted EPS 8.43
These are next-fiscal-year annual uplift estimates, not next-quarter numbers.
Aggregate next-FY est. rev uplift: % · next-FY EPS uplift: % · vs analysts: inline · priced in: medium (model's call-read: high; verdict above is the hard-data one used for ranking) · confidence: 6/10
| Claim | Figure | Arithmetic | Next-FY Rev % | Next-FY EPS % |
|---|
$50B capital advised on AI deployment (already)
revenue · soft | >$50B | $50B capital advised is disclosed, but it is CLIENT capital advised, not MMC revenue. No advisory take-rate/fee, recognized revenue, or next-FY conversion is disclosed. Illustration: a hypothetical 1% fee = $500M / $26.981B = 1.85%, but the fee rate is invented, so unanchored. Bookings/capital-advised ≠ revenue. | | |
Document ingestion +20% process efficiency
productivity · soft | 20% | 20% process efficiency disclosed, but the affected process cost/FTE/opex base is undisclosed -> cannot translate 20% to dollars or after-tax saving. | | |
Legacy tool -> broker workbench in days (saved months)
productivity · soft | days vs months | One-off effort saving, no team size/headcount/labor cost quantified. Not annualizable. | | |
Policy renewal center built in weeks vs many months
productivity · soft | weeks vs months | Implementation time compression disclosed, but no project cost, labor base, or recurring operating saving disclosed. | | |
Several thousand colleagues analyzing ~$200B loss info
engagement · soft | ~$200B; thousands of users | $200B is loss data analyzed, NOT revenue or fees. Adoption-scale/better-advice metric; no monetization rate, retained premium, fee, or cost saving disclosed. | | |
>2M AI-suite prompts/month (Marsh Risk)
engagement · soft | >2M prompts/mo | Usage volume disclosed, but no minutes saved per prompt, hourly cost, or revenue conversion -> no cost-displacement or revenue figure. | | |
Pilot: +50% sales velocity
revenue · soft | 50% | Pilot-only against an undisclosed pilot revenue book; rollout scope and next-FY conversion not disclosed. Cannot project to firm-level next-FY revenue. | | |
BCS: dozens of AI productivity tools
productivity · soft | dozens | Count of tools, no efficiency $ or FTE saving disclosed. Vague. | | |
~30% of consulting work draws on AI/analytics
engagement · soft | ~30% | Current mix/exposure metric, NOT incremental uplift. No consulting segment revenue base or AI-driven incremental growth delta disclosed. | | |
~$3T digital-infra investment over 5yrs (market)
other · soft | ~$3T (~$600B/yr) | External market TAM MMC advises into, NOT MMC revenue. No capture/share/fee disclosed. Opportunity framing only. | | |
2,000-3,000 data centers built worldwide / 5yrs
other · soft | 2,000-3,000 (~400-600/yr) | Market unit forecast, not MMC revenue. No revenue per data center, win rate, or service attach rate disclosed. | | |
Assumptions: Current revenue $26.981B, net income $4.160B, net margin 15.4%, diluted shares 490M; default tax 21% and default incremental margin = current net margin. EPS uplift would be sized on the consensus/adjusted base (~$4.717B adjusted NI implied by $9.60 EPS x 490M sh), NOT the depressed GAAP $4.16B, per the earnings-basis rule. Would take only the next-FY portion of any multi-year figure. None could be applied: every disclosed dollar figure ($50B, $200B, $3T) is a CLIENT or MARKET quantity, and every productivity figure (20%, days/weeks) lacks a disclosed MMC cost/FTE denominator. MMC is an ADOPTER (broker/consultant), not a supplier of AI compute/infra — the $3T/data-center figures are advisory-market TAM, so no supplier-side line exists. Bookings/capital-advised not treated as revenue absent disclosed conversion.
Top line: No quantifiable next-FY revenue uplift can be derived. The biggest numbers (>$50B AI-deployment capital advised, ~$3T 5-yr infra TAM, 2,000-3,000 data centers, $200B loss info) are client/market quantities, not MMC fees; none discloses Marsh's fee rate, share, or rollout base. The most revenue-adjacent claim — +50% sales velocity — is pilot-only against an undisclosed book, and '30% of consulting draws on AI' is current mix, not incremental growth. Hypothetical only: a 1% advisory take on $50B ≈ $500M / $26.981B ≈ 1.85%, but the fee rate is invented, so it stays soft.
Bottom line: All efficiency claims (20% document-ingestion efficiency, months-to-days/weeks build times, dozens of BCS tools, >2M prompts/mo) are genuine adopter-side margin levers but none carries a disclosed cost or headcount base, so no after-tax saving can be computed. Management itself frames these as fueling 'continued margin improvement' rather than a discrete savings number — diffuse, not a sizable standalone EPS event.
[impact n/m (all claims soft/unanchored)] Consensus already embeds ~10-11% growth: revenue ($26.942B-$24.358B)/$24.358B = +10.6%; EPS ($9.601-$8.692)/$8.692 = +10.5%; NI +~10.7%, driven by steady margin expansion. MMC explicitly says AI savings 'fuel additional growth investments... building confidence in continued margin improvement' — i.e. the AI benefit IS the margin-improvement trajectory consensus has already priced. With zero quantified incremental dollars above that trend, the calculable adopter-side uplift is 0% above consensus and there is no measurable gap; the AI story supports the embedded ~10% algorithm rather than pointing clearly above it.
MODEL CONSENSUS (impact)
partial
Both: MMC adopter, all 11 claims soft/null, priced_in high. Conflicts on aggregate pcts (used null) and verdict (used inline, better justified); confidence lowered to 6.
Conflicts reconciled
- est_rev_uplift_pct: X=null vs Y=0 -> null because all per-claim impacts are unanchored/null, so the aggregate is non-derivable, not a computed zero
- est_eps_uplift_pct: X=null vs Y=0 -> null for the same internal-consistency reason
- supplier_rev_uplift_pct: X=null vs Y=0 -> null (MMC is a pure adopter; no supplier line exists, so N/A not 0)
- vs_analyst_expectations: X=inline vs Y=unclear -> inline because X cites the ~10% consensus math and management's 'continued margin improvement' framing showing AI fuels the already-priced trend
| Field | Opus 4.8 | GPT-5.5 |
|---|
| Rev uplift % | – | 0 |
| EPS uplift % | – | 0 |
| Priced in | high | high |
| vs analysts | inline | unclear |
| Confidence | 3 | 7 |
| Top line | No quantifiable next-FY revenue uplift can be derived. The biggest numbers ($50B AI-deployment capital advised, $3T 5-yr infra TAM, 2,000-3,000 data centers, $200B loss info) are client/market quantities, not MMC fees. Even the most concrete revenue-adjacent claim — +50% sales velocity — is pilot-only against an undisclosed pilot book, and the '30% of consulting draws on AI' is current mix, not incremental growth. Real but unsizable here: at a purely hypothetical 1% advisory take on $50B that's ~$500M / $26.981B ≈ 1.85%, but the fee rate is invented, so it stays soft. | Management cites large AI-adjacent demand pools: >$50B capital advised, $3T infrastructure investment over ~5 years, 2,000-3,000 data centers, 30% AI/advanced-analytics consulting mix, and 50% pilot sales velocity. None discloses Marsh's fee rate, share, rollout base, or next-FY revenue conversion, so hard revenue uplift is 0.0%. |
| Bottom line | All efficiency claims (20% document-ingestion efficiency, months-to-days/weeks build times, dozens of BCS tools, 2M prompts/mo) are genuine adopter-side margin levers but none carries a disclosed cost or headcount base, so no after-tax saving can be computed (e.g. 20% of an undisclosed process opex = unknown $). Management itself frames these as fueling 'continued margin improvement' rather than a discrete savings number — diffuse, not a sizable standalone EPS event. | Management cites 20% document-ingestion efficiency, >2M prompts/month, dozens of productivity tools, and implementation time compressed from months to weeks/days. No affected labor or cost base is disclosed, so hard EPS uplift is 0.0%. |
| Reasoning | Consensus already embeds ~10-11% growth across the board (rev 24.358B->26.942B = +10.6%; EPS 8.692->9.601 = +10.5%; NI +10.7%) driven by steady margin expansion. MMC explicitly says AI savings 'fuel additional growth investments... building confidence in continued margin improvement' — i.e. the AI benefit IS the margin-improvement trajectory consensus has already priced. With zero quantified incremental dollars above that trend, there is no measurable gap to consensus; the AI story supports the embedded ~10% algorithm rather than pointing clearly above it. | Consensus already embeds growth from 2024 to 2025 of revenue: ($26.942B - $24.358B) / $24.358B * 100 = 10.61%, and EPS: ($9.60062 - $8.692) / $8.692 * 100 = 10.45%. The AI claims provide no hard incremental next-FY dollar uplift above the $26.981B revenue and $4.160B net-income base, so the calculable adopter-side uplift is 0.0% revenue and 0.0% EPS versus consensus growth already above 10%. |
Rows highlighted where the two models disagreed.
QUANTIFICATIONS
AI deployment capital advised: more than $50 billion (already, topline)
“We've already advised on more than $50 billion of capital investment in AI deployment.”
document ingestion volume and process efficiency: thousands of documents weekly; 20% (now, bottomline)
“For instance, our document ingestion capability is now handling thousands of documents weekly already improving efficiency in these processes by 20% and enhancing the quality of the data and its usability to further support clients with valuable insights.”
legacy tool modernization effort saved: in days; saving months of team effort (recently, bottomline)
“For example, we recently used AI to turn a legacy tool into a newly designed broker workbench in days saving months of team effort.”
policy renewal center implementation time: in weeks; otherwise many months (current, bottomline)
“And in our policy renewal center, AI has enabled us to transform a traditionally manual e-mail heavy process into a streamlined digital solution in weeks, a project that otherwise would have taken many months.”
AI-enabled claims analysis users and loss information: several thousand colleagues; almost $200 billion (now, both)
“we've got several thousand colleagues now drawing an AI-enabled analysis of almost $200 billion of loss information, which helps them support much better advice decision-making and advocacy.”
proprietary AI suite usage: more than 2 million prompts a month (current, bottomline)
“we've talked before about our general proprietary AI suite just within Marsh Risk and up to more than 2 million prompts a month.”
sales velocity pilot: 50% increase (pilot, topline)
“And in the areas where we pilot that, we see the amount of work that, that takes off our client teams drive a 50% increase in sales velocity in the pilot.”
AI-driven productivity tools: dozens (current, bottomline)
“BCS has introduced dozens of AI-driven productivity tools, and we're ramping up adoption to give our colleagues an edge.”
AI and advanced analytics consulting mix: maybe 30% of our work (current, topline)
“So within the business, maybe 30% of our work draws on advanced analytics and AI.”
digital infrastructure investment opportunity: roughly $3 trillion (over the course of the next five years or so, topline)
“We expect roughly $3 trillion of investment over the course of the next five years or so.”
data centers to be constructed worldwide: between 2,000 to 3,000 data centers (Over the next five years, topline)
“Over the next five years, it's estimated that between 2,000 to 3,000 data centers will be constructed worldwide.”
PAST (realized)
- We've already advised on more than $50 billion of capital investment in AI deployment.
- we moved early on AI, and we're excited about how it's already making us better and how it's going to make us better in the future.
- We have deployed agentic AI in our IT help desk, significantly reducing inquiries, improving colleague experience and creating downstream efficiencies in our support centers.
- BCS has introduced dozens of AI-driven productivity tools, and we're ramping up adoption to give our colleagues an edge.
- What we see in our business is that the use of AI tools and agents has had a significantly positive effect on productivity.
CURRENT (now)
- We are building AI-enabled applications and services that are generating new revenue streams as well as enhancing world-class capabilities and data-driven insights in insurance, health, human capital and investments.
- This is showing up in hundreds of different ways across a wide variety of roles.
- Across our business, we are starting to see the impact of AI automation.
- For instance, our document ingestion capability is now handling thousands of documents weekly already improving efficiency in these processes by 20% and enhancing the quality of the data and its usability to further support clients with valuable insights.
- AI is developing into a very large opportunity for us as a consulting business that works on strategy and transformation.
FORWARD (guidance)
- We also see significant AI growth opportunity in consulting.
- AI-enabled savings will fuel additional growth investments, including in producer talent and new capabilities while building our confidence in continued margin improvement.
- we're excited about how it's already making us better and how it's going to make us better in the future.
- We see strong growth potential in client-facing technology, virtual agents, and chatbots.
- The Thrive program will drive growth through investments in talent and AI, strengthen our brand, and generate greater efficiency.
TRACK RECORD — PROMISE vs DELIVERY
60/100 track record too-early 6 calls reviewed
Marsh talks extensively about AI (Len.ai, Centrisk, AIDA, BCS) but its only genuinely quantified, time-bound AI-linked target is the Thrive program's ~$400M of savings over three years (announced Q3 FY2025, partly AI-driven), consistently reiterated as on-track but whose window does not close until ~2028. With no AI-specific quantified promise yet judgeable, the track record is encouraging in tone but unproven.
Thrive program: ~$400M in savings over three years, driven substantially by AI-enabled automation/process reengineering in BCS (~$500M charges) — promised Q3 FY2025
too-early Reiterated as 'on track' in Q4 FY2025 and Q1 FY2026 with charges being incurred, but the 3-year window runs to ~2028, so savings are unproven
AI document-ingestion capability improving those processes by 20% (handling thousands of docs weekly) — promised Q1 FY2026
too-early Stated as an already-achieved result in the same call, not a forward target with a later checkpoint in this set
Len.ai handling ~1M colleague inquiries/week to drive efficiency/automation — promised Q3 FY2025
too-early Cited as current run-rate usage rather than a dated target; no quantified future milestone attached
PRICED-IN (REFINED)
MEDIUMEst. revisions flat · Fwd P/E 20.3 · EV/Sales 3.6x
AI claim maps to Marsh Insurance Group, Mercer Consulting Group, Oliver Wyman Group Consulting Group
Analyst ratings show only mild upward migration, but price targets are falling with last-month average below last-quarter and last-year averages, so revision momentum is best characterized as flat rather than rising. Forward EPS and revenue growth are solid but not dramatic, while the stock trades at a rich 20.3x forward P/E and 3.6x EV/Sales for a mature insurance broker. AI upside would most plausibly show up in Marsh Insurance Group and consulting segments through automation, analytics, and advisory productivity. Because estimates are not clearly rising but valuation is already elevated, the AI upside looks partially priced in, supporting a medium verdict.
COVERAGE — ENTHUSIASM TRAJECTORY + CATALYSTS
4Q4 FY20245Q1 FY20254Q2 FY20258Q3 FY20258Q4 FY202510Q1 FY2026
AI enthusiasm across 6 calls — trend ↗ rising
AI evolved from scattered digital-tool mentions to a detailed growth, productivity, and efficiency strategy with named products and quantified savings.
RECENT AI CATALYSTS & NEWS
BUSINESS IMPACT - QUALITATIVE MATERIALITY
7/10 qualitative impact material medium-term · mixed evidence
Where AI matters: broker productivity, risk analytics, consulting AI services
AI is already embedded in Marsh's broker workflows, claims analytics, document ingestion, renewal processes, and Oliver Wyman AI advisory work, with credible usage metrics such as thousands of colleagues, $200B of loss data, >2M prompts/month, and 20% process-efficiency gains. The upside is material but not transformational because disclosed metrics mostly show productivity/adoption rather than firm-level incremental revenue or EPS.
Caveats: No disclosed AI-specific revenue or EPS uplift; Consulting billable-hour deflation could offset some AI advisory growth; Productivity gains may be competed away or shared with clients; Execution risk on Thrive savings and broad AI rollout
AI DISRUPTION / CANNIBALIZATION RISK two-sided · 3/10
Core insurance brokerage is relatively durable because complex risk placement, carrier relationships, fiduciary trust, and fee-as-percent-of-premium economics are not easily automated away. The real pressure is in consulting and administrative brokerage workflows, where AI can compress billable hours and commoditize routine analysis, but this does not appear to overwhelm the core model.
OPTIONS / MARKET STRUCTURE
option liquidity: good
proxy inputs — dollar-ADV $484M · beta 0.637 · px $161.39
source: proxy (no options chain on FMP)
FMP /stable/ exposes no options-chain endpoint on this key, so ATM IV, bid-ask spread and open interest are unavailable. Liquidity below is a PROXY from dollar-ADV, beta and price level (a stand-in for option depth), not measured option-market data.
CONFIRMATION — INSIDERS · 13F · LANGUAGE
Mixed — insiders selling, institutions unknown, management language 8/10 committed.
INSIDERS selling 3 open-market sell(s) vs 0 buy(s) — net distribution
INSTITUTIONS (13F) unknown no 13F data returned
MGMT LANGUAGE 8/10 committed Strong ownership with deployed tools, quantified efficiency, and firm use cases, though some growth claims remain pipeline-oriented.
commit “We are building AI-enabled applications and services that are generating new revenue streams”
commit “our document ingestion capability is now handling thousands of documents weekly already improving efficiency in these processes by 20%”
commit “We have deployed agentic AI in our IT help desk, significantly reducing inquiries”
VERBATIM AI QUOTES
“I'd like to take a moment to discuss our AI strategy and why we believe Marsh will be an AI winner.”
— John Doyle, Q1 FY2026
“Our strategy leverages our scale and capacity to invest in AI to drive even greater value from our proprietary data assets and our role as our clients' trusted adviser.”
— John Doyle, Q1 FY2026
“We are focused on 3 main pillars. The first is growth. We are building AI-enabled applications and services that are generating new revenue streams as well as enhancing world-class capabilities and data-driven insights in insurance, health, human capital and investments.”
— John Doyle, Q1 FY2026
“Oliver Wyman's AI Quotient team created to help clients deploy their own AI strategies is its fastest-growing practice.”
— John Doyle, Q1 FY2026
“We've already advised on more than $50 billion of capital investment in AI deployment.”
— John Doyle, Q1 FY2026
“Our second pillar is productivity, which focuses on deploying AI capabilities to boost the performance of our colleagues.”
— John Doyle, Q1 FY2026
“The final pillar is efficiency. Across our business, we are starting to see the impact of AI automation.”
— John Doyle, Q1 FY2026
“For instance, our document ingestion capability is now handling thousands of documents weekly already improving efficiency in these processes by 20% and enhancing the quality of the data and its usability to further support clients with valuable insights.”
— John Doyle, Q1 FY2026
“AI-enabled savings will fuel additional growth investments, including in producer talent and new capabilities while building our confidence in continued margin improvement.”
— John Doyle, Q1 FY2026
“But when you do our advocacy is strong and if you take an example like Claims IQ, which is our AI-enabled toolkit, we've got several thousand colleagues now drawing an AI-enabled analysis of almost $200 billion of loss information, which helps them support much better advice decision-making and advocacy.”
— Nicholas Studer, Q1 FY2026
“But the AI investments are, I think, massively enabling of growth and of productivity and of efficiency.”
— Nicholas Studer, Q1 FY2026
“And in the areas where we pilot that, we see the amount of work that, that takes off our client teams drive a 50% increase in sales velocity in the pilot.”
— Nicholas Studer, Q1 FY2026
“Thanks, John. And Mike, you heard John in his prepared remarks, mentioned GC Quotebox, which is an AI-driven document ingestion tool. This is really a game changer for Guy Carpenter in our business.”
— Dean Klisura, Q1 FY2026
“So Mercer Fiber is one of the tools where we're leveraging the broader AI stack that we have at Marsh to kind of further enable our existing digital tools.”
— Patrick Tomlinson, Q1 FY2026
“AI is developing into a very large opportunity for us as a consulting business that works on strategy and transformation.”
— Ted Moynihan, Q1 FY2026
“Through BCS, we're building a data and technology ecosystem that harnesses AI and advanced analytics to improve client outcomes and drive operational excellence.”
— John Doyle, Q4 FY2025
“It accelerates expense savings and investment in AI and automation.”
— John Doyle, Q4 FY2025
“BCS has introduced dozens of AI-driven productivity tools, and we're ramping up adoption to give our colleagues an edge.”
— John Doyle, Q4 FY2025
“We see strong growth potential in client-facing technology, virtual agents, and chatbots.”
— John Doyle, Q4 FY2025
“The Thrive program will drive growth through investments in talent and AI, strengthen our brand, and generate greater efficiency.”
— John Doyle, Q4 FY2025
“What we see in our business is that the use of AI tools and agents has had a significantly positive effect on productivity.”
— Nick Studer, Q4 FY2025
“The second piece is our quotient platform which delivers AI work for clients, AI transformation for clients.”
— Nick Studer, Q4 FY2025
“But not one at the moment which is experiencing revenue headwinds because of this.”
— Nick Studer, Q4 FY2025
ANALYST QUESTIONS ON AI
Q (Q1 FY2026, Charles Peters): And maybe it's embedded in your AI comments. But with your margin results being so high, curious about the risks of AI disintermediation across the various businesses that you have?
A: we expect to be an AI winner, we moved early on AI, and we're excited about how it's already making us better and how it's going to make us better in the future.
Q (Q1 FY2026, Michael Zaremski): Just 1 question on maybe around the AI conversation, specifically on the value-add services that you offer your clients.
A: if you take an example like Claims IQ, which is our AI-enabled toolkit, we've got several thousand colleagues now drawing an AI-enabled analysis of almost $200 billion of loss information, which helps them support much better advice decision-making and advocacy.
Q (Q1 FY2026, Brian Meredith): So John, it's clear that AI is going to have productivity benefits,, it's going to benefit client experience and growth, et cetera. But 1 of the debates I'm having with investors is how much of the productivity gains is Marsh going to be able to keep and see a benefit from a margin perspective versus protects being competed away or giving back to clients.
A: Our fees have been for a long time, stable as a percentage of premium, and they're quite small compared to the cost of risk that we help our clients manage.
Q (Q1 FY2026, Robert Cox): I'm just curious how, if at all, AI is changing the M&A strategy. Are you staying away from certain businesses pivoting towards others and have your technology requirements or anything else changed?
A: I'm hopeful that actually the scale benefits that we bring to investing in AI and the data sets and client relationships and all the advisory work will create opportunities for us to consolidate smaller brokers over time who are going to struggle to compete and to invest in these technologies.
Q (Q1 FY2026, Meyer Shields): Are commissions still the right way to be compensated for that? Or do you expect compensation to become more transparently tied to the individual services?
A: We have fees, we have commissions. We have success fees, right? I'm sure there are other things I'm not even thinking about. We're very transparent with our clients about how we get paid.
Q (Q1 FY2026, David Motemaden): could you just remind us how much you guys are spending on AI just broadly within the tech budget. And I guess, who are you partnering with? What LLM providers are you partnering with? What tools are you using?
A: It wouldn't be a refresher because we've not shared that data in the past. We have a healthy tech CapEx budget.
Q (Q4 FY2025, Gregory Peters): I'd like to go back to your comments on AI and digital infrastructure. And I guess I'm curious how you think the trends of investment in these areas by your clients could affect the long-term revenue outlook for RIS, for the consulting business and the health business where, I guess, there could be some potential rising employment volatility.
A: We're excited about the investment in the digital infrastructure world. We expect roughly $3 trillion of investment over the course of the next five years or so.
Q (Q4 FY2025, Mike Zaremski): When we think about Thrive, would you say that that encompasses a lot of the new AI technologies that you all are deploying, or should we expect kind of a more to come?
A: We've introduced dozens of productivity tools to our colleagues. We're early movers on this.
Q (Q4 FY2025, Brian Meredith): I'm hearing some things in the marketplace that for the management consulting business, formerly Oliver Wyman, that there could be some project-related stuff that actually goes the way of AI and maybe a headwind.
A: What we see in our business is that the use of AI tools and agents has had a significantly positive effect on productivity.