← back to rankingCBRE · CBRE Group, Inc.
Real Estate - Services · mkt cap $37.4B · calls: Q1 FY2026 vs Q4 FY2025
31.0 conviction · conf-adj 31
conf 4/10 partial
enthusiasm:24.0 · trend:8 · quantifies:0 · impact:0 · under_radar:5 · credibility:0 · business_impact:4 · disruption:-6 · commitment:-4 · confirmation:0
Enthusiasm latest 8 / prev 7 (rising)
Management frames AI as a net positive: a large, quantified demand tailwind through data center/digital infrastructure and hyperscaler work, plus internal efficiency and broker-facing data tools—not a near-term threat to core brokerage. Credibility is stronger on the demand side (reported revenue/EBITDA scale, hiring constraints, Meta training partnership) than on cost takeout (25% research and ~25% call-center targets are directional, with Emma pushing headcount impact several years out). Enthusiasm rose from a structured Q4 risk/opportunity primer to a Q1 narrative that explicitly ranks AI-driven infrastructure growth alongside historical outsourcing and reiterates transactional disintermediation skepticism.
GROUNDED NEXT-FY IMPACT vs CONSENSUS
Grounded on actual base — revenue $40.5B · net income $1.2B · net margin 2.9% · diluted EPS 3.85
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: 4/10
| Claim | Figure | Arithmetic | Next-FY Rev % | Next-FY EPS % |
|---|
Research work cost −25% via AI
cost · soft | ~25% | 25% × undisclosed research-cost base — no $ base appears anywhere in inputs, so unsizable. Mgmt: corporate headcount cuts come 'a few years from now versus immediately' → ~0 FY2026 P&L. Cost claim → topline 0. | 0 | |
Call-center cost/headcount −~25%
cost · soft | as much as 25% | 25% × undisclosed call-center cost base — no $ base in inputs → unsizable; deferred a few years. Cost claim → topline 0. | 0 | |
Corporate function headcount cuts (research/HR)
cost · soft | not sized | Unquantified beyond the two ~25% items; Emma: reductions 'a few years from now.' No base. | | |
AI data/broker productivity 'concrete evidence by 2026'
productivity · soft | qualitative | No figure attached — vague 'real gains.' Cannot size. | | |
Data Center Solutions revenue $2B in 2026 (+20%/yr)
revenue | $2B in 2026, +20%/yr | $2B is TOTAL 2026, not incremental. At 20% growth, 2025≈$1.667B → incr = 2B−2B/1.2 = $333M; 100×0.333/40.55 = 0.82%. EPS left null (margin/segment overlap with infra lines). | 0.82 | |
Infrastructure services revenue +~50% in 2026
revenue | +~50% on disclosed $3B 2025 base | DISCLOSED-BASE $3B: 50%×$3B=$1.5B incr; 100×1.5/40.55=3.70%. Flow at consolidated 2.853% net margin → incr NI $42.8M; 100×42.8/1,906.4=2.25% of consensus adj NI. (GAAP-denominated would be ~3× larger — thin-margin artifact, not used.) | 3.7 | 2.25 |
Critical infrastructure services line +>60% in 2026
revenue | +>60% on disclosed $1.7B 2025 | DISCLOSED-BASE $1.7B: 60%×$1.7B=$1.02B incr (floor); 100×1.02/40.55=2.52%. incr NI $29.1M → 100×29.1/1,906.4=1.53% adj. SUBSET of the $3B infrastructure-services line — NOT additive (would double-count). | 2.52 | 1.53 |
Total infrastructure activities >$3B FY25; ~$950M Q1
other | >$3B FY25; ~$950M Q1 | Disclosed scale/denominator for the +50% line, not an incremental uplift on its own (captured via that line). 100×$3B/$40.55B=7.40% revenue mix; Q1 $950M=2.34% of annual rev. | | |
Data center work = ~14% of core EBITDA (2025)
other · soft | ~14% | Share-of-EXISTING EBITDA mix descriptor for 2025, not a forward uplift; no core EBITDA $ in base → no rev/EPS uplift computed. | | |
1,300+ data centers under building-ops management
other · soft | 1,300+ | Installed-base scale/count metric; no revenue/$ attached. | | |
US data center leasing revenue 'more than tripled' YoY Q1
revenue · soft | >3x YoY (Q1) | Growth multiple with no absolute $ base disclosed → cannot convert to a revenue %; not annualized to avoid inventing a base. | | |
Assumptions: Next fiscal year = 2026. Revenue uplift % = 100 × incremental $ / $40.55B FY25 revenue. EPS sized against the ADJUSTED basis consensus uses (FY2025 adj NI ≈ $1.906B / EPS $6.33; FY2026 ≈ $2.305B / $7.69), NOT GAAP (GAAP margin 2.85% is sub-3% / depressed; $1.157B base would manufacture a meaningless huge EPS%). Incremental revenue flows at consolidated 2.853% net margin unless noted; cost saves would use 21% tax but no $ bases exist for research/call-center. Phasing: only the NEXT-FY (2026) incremental slice counted; DCS $2B is a 2026 LEVEL, so only the +20% YoY step ($333M) is incremental. Overlap: critical-infrastructure (+$1.02B) is a SUBSET of infrastructure services (+$1.5B) — supplier headline uses $1.5B; DC Solutions (+$333M) possibly additive (up to ~4.5%). Adopter cost-saving claims left null/0 because no dollar cost base is disclosed and management defers headcount cuts 'a few years from now' → FY2026 P&L ≈ 0. Supplier = selling into the AI/data-center buildout; adopter = own cost/productivity.
Top line: Essentially ALL of CBRE's quantified AI-linked revenue is SUPPLIER-side — selling project, leasing and facilities services INTO the data-center/AI buildout, not AI improving CBRE's own product. Largest non-overlapping piece: infrastructure services $3B (2025) growing ~50% = +$1.5B = 3.70% of $40.55B revenue; DC Solutions adds ~$333M (0.82%, possibly additive → up to ~4.5%); critical infrastructure (+$1.02B/2.52%) sits inside the $1.5B. Adopter-side AI (own-cost savings) adds ~0 to the top line. Consensus 2026 revenue $43.20B vs $40.55B = +6.7% (+$2.65B); the $1.5B wedge is ~55–57% of that step — material but plausibly already embedded in infra-heavy models. Unanchored: DCS framing, leasing 'tripled' Q1.
Bottom line: Adopter EPS case is unsizable: the two concrete cuts (research ~25%, call-center ~25%) have NO disclosed dollar cost base anywhere in the inputs, and management twice flags the headcount cuts as 'a few years from now versus immediately,' so FY2026 adjusted EPS impact ≈ 0 (and any small saving off the depressed 2.85% GAAP base would throw a misleadingly large EPS%). Supplier infra revenue does drop incremental profit: +$1.5B at 2.853% net margin ≈ +$43M → ~2.25% of consensus adjusted NI (≈21% of the +21.4% 2025→2026 consensus EPS step), but it flows through segment economics consensus already models, not a discrete AI cost lever. GAAP-denominated EPS% would look ~3× larger — thin-margin artifact, not reported as headline.
[sized AI revenue is supplier-side (selling into the buildout), not adopter — uplift n/m] Consensus 2025→2026 builds revenue $40.55B→$43.20B (+6.7%, +$2.65B) and adjusted EPS $6.33→$7.69 (+21.4%, adj NI +~$396M). The headline AI-linked supplier revenue (infrastructure services +$1.5B = 3.7%) is ~55–57% of the modeled revenue step — clearly embedded, not surprise upside — so vs-expectations is inline and priced_in is high. EPS contribution (~2.25% of adj NI) runs through segment economics already in the model. Adopter-side cost AI (research/call-center −25%) is unquantifiable (no disclosed base) and explicitly deferred 'a few years,' so it adds nothing to FY2026.
MODEL CONSENSUS (impact)
partial
Merged 11-claim union; supplier topline 3.7% and adopter ~0 agreed. Disagreements resolved toward sounder arithmetic (Y on DCS, X on EPS flow) and the more conservative, better-justified priced-in verdict.
Conflicts reconciled
- DCS $2B rev_uplift_pct: X=null(soft) vs Y=0.82(hard) -> used 0.82, soft=false because Y correctly took only the +20% incremental step ($333M) per the multi-year-phasing rule
- infra/critical eps_uplift_pct: X=2.25/1.53 vs Y=null -> used X because method explicitly requires flowing incremental revenue to EPS at a stated margin
- cost claims rev_uplift_pct: X=0 vs Y=null -> used 0 because a cost saving's topline impact is definitionally ~0
- vs_analyst_expectations: X=unclear vs Y=inline -> used inline (better-justified by the ~55% share-of-growth math and more conservative)
- priced_in: X=medium vs Y=high -> used high (Y's embedding argument is stronger and less optimistic)
- est_rev_uplift_pct: X=null vs Y=0 -> used null (aggregate genuinely indeterminate given supplier/adopter split and overlaps; confidence lowered)
| Field | Opus 4.8 | GPT-5.5 |
|---|
| Rev uplift % | 0 | – |
| EPS uplift % | – | – |
| Priced in | high | – |
| vs analysts | inline | – |
| Confidence | 4 | – |
| Top line | Essentially ALL of CBRE's quantified AI-linked revenue is SUPPLIER-side — selling project, leasing and facilities services INTO the data-center/AI buildout, not AI improving CBRE's own product. The largest non-overlapping piece is infrastructure services: $3B (2025) growing ~50% = +$1.5B, i.e. 3.7% of $40.55B revenue; Data Center Solutions adds ~$333M (0.8%); critical infrastructure (+$1.02B) sits inside that. Adopter-side AI (own-cost savings) adds ~0 to the top line. Net adopter topline uplift ≈ 0%; supplier topline ≈ +3.7% (up to ~4.5% if DC Solutions is fully additive). | – |
| Bottom line | The adopter EPS case is unsizable on the disclosed numbers. The two concrete cuts — research ~25% and call-center ~25% — have NO disclosed dollar cost base anywhere in the inputs, and management twice flags the headcount reductions as 'a few years from now versus immediately,' so FY2026 adjusted EPS impact is ≈0. With GAAP net margin at 2.85%, any small saving would also throw a misleadingly large EPS% off the depressed $1.157B base, so I do not report one. est_eps_uplift_pct = null. Supplier infra revenue does drop incremental profit, but that flows through the segment economics consensus already models, not a discrete AI cost lever. | – |
| Reasoning | Consensus already builds FY2026 revenue to $43.20B, +$2.65B (+6.7%) over $40.55B. The headline AI-linked supplier revenue (infrastructure services +$1.5B) is ~55% of that modeled growth — clearly embedded, not surprise upside. Consensus adjusted EPS also steps $6.33→$7.69 (+21.4%) without needing the AI cost cuts, which management defers several years. So the math does NOT point above consensus: supplier revenue is priced in, and the adopter-side savings are too unanchored and too deferred to move the FY2026 number. As an ADOPTER thesis (which this product ranks), CBRE's quantified AI impact is ~0 incremental — the real AI exposure is supplier-side and already in the tape. | – |
Rows highlighted where the two models disagreed.
QUANTIFICATIONS
Research cost reduction from AI: ~25% (Over this year, maybe extending into next year (stated Q4 FY2025), bottomline)
“We think we're going to be able to save over this year maybe extending into next year, about 25% of the cost of our research work using AI and the data that we assemble and manipulate through AI to support the research work.”
Call center headcount/cost rationalization from AI initiatives: maybe as much as 25% (Anticipated (Q1 FY2026; no explicit completion date), bottomline)
“We think some of that -- we think we can rationalize that by maybe as much as 25%.”
Data Center Solutions revenue: $2 billion expected in 2026; growing 20% per year (2026 and ongoing growth rate (Q4 FY2025), topline)
“Revenue from this business is expected to reach $2 billion in 2026 and is growing at 20% per year.”
Data center / digital infrastructure share of core EBITDA: approximately 14% (2025 (Q4 FY2025), both)
“data center and digital infrastructure work across our 4 business segments accounted for approximately 14% of our core EBITDA in 2025.”
Total infrastructure activities revenue: more than $3 billion in 2025; nearly $950 million in Q1 (2025 full year and Q1 2026 (Q1 FY2026; AI secular tailwind framing), topline)
“We generated more than $3 billion of total revenue from infrastructure activities in 2025 and nearly $950 million in the first quarter.”
Infrastructure services revenue growth: growing almost 50% this year (2026 (Q1 FY2026; tied to AI secular tailwind), topline)
“$3 billion of revenue last year, almost $1 billion of revenue in the first quarter, growing almost 50% this year”
Critical infrastructure services line revenue growth: in excess of 60% this year (2026 (Q1 FY2026 prepared remarks), topline)
“Revenue in this business line totaled $1.7 billion in 2025 and $580 million in the first quarter and is expected to grow in excess of 60% this year.”
Data centers under building operations management: over 1,300 data centers around the world (Current scale (Q1 FY2026), topline)
“We do work on over 1,300 data centers around the world.”
Evidence of AI-driven data/broker productivity gains: concrete evidence by end of 2026 (By end of 2026 (Q4 FY2025), both)
“We think by the end of 2026, we'll -- there'll be concrete evidence that we've made some real gains in terms of extracting the data we have assimilating it and delivering it to our professionals in a way that we haven't done before.”
Corporate function headcount reductions from AI: not numerically sized beyond ~25% call centers and ~25% research (A few years from now vs immediately (Q1 FY2026), bottomline)
“even those head count reductions, we anticipate happening a few years from now versus immediately.”
U.S. data center leasing revenue growth: more than tripled (Q1 2026 vs prior-year Q1 (Q1 FY2026), topline)
“data center leasing revenue more than tripled from last year's first quarter.”
PAST (realized)
- Q4 FY2025 — Bob: "We are using AI today in 2 areas."
- Q4 FY2025 — Bob: "With the use of AI, we are moving toward gaining advantages that are more in proportion to the data advantage that comes with our market position."
- Q4 FY2025 — Bob: "What we're really working hard to do in thinking we're finally making some gains on using AI is to provide data to our brokers in a more efficient way and a more cost-effective way for us."
- Q4 FY2025 — Bob: "we've had an appraisal business there that was heavily, heavily automated" (Asia-Pacific valuations).
- Q4 FY2025 — Bob: "data center and digital infrastructure work across our 4 business segments accounted for approximately 14% of our core EBITDA in 2025."
- Q4 FY2025 — Emma: "data center leasing revenue more than doubled" (U.S. leasing context in Q4 remarks).
- Q1 FY2026 — Emma: "data center leasing revenue more than tripled from last year's first quarter."
- Q1 FY2026 — Bob: "we generated more than $3 billion of total revenue from infrastructure activities in 2025 and nearly $950 million in the first quarter."
CURRENT (now)
- Q4 FY2025 — Bob: "We're deploying AI where its economic value clearly exceeds the economic value of traditional efficiency levers like offshoring."
- Q4 FY2025 — Bob: "We are not reducing broker headcount. We're adding brokers."
- Q1 FY2026 — Bob: "we have AI initiatives underway to create efficiencies in the company."
- Q1 FY2026 — Bob: "We are developing AI-enabled tools in every 1 of those areas."
- Q1 FY2026 — Bob: "we are controlling it controlling who has access to it, controlling what we use it for."
- Q1 FY2026 — Bob: "We're building a capability there in multiple cities around the U.S. to recruit, train and place technical people to support Meta's data center initiative."
- Q1 FY2026 — Bob: "we can't hire enough people" in critical infrastructure / data center services.
- Q1 FY2026 — Emma: "we're constantly organically investing through our CapEx in technology and now AI to support our business"
FORWARD (guidance)
- Q4 FY2025 — Bob: "On balance... we are optimistic that the net impact will benefit CBRE in the long run."
- Q4 FY2025 — Bob: "Revenue from this business is expected to reach $2 billion in 2026 and is growing at 20% per year." (Data Center Solutions)
- Q4 FY2025 — Bob: "We think by the end of 2026, we'll -- there'll be concrete evidence that we've made some real gains" on data delivery enabled by AI.
- Q4 FY2025 — Bob: "We think we're going to be able to save over this year maybe extending into next year, about 25% of the cost of our research work using AI"
- Q4 FY2025 — Bob: "In the long run, will there be less office users because AI disintermediates some of the work people do, that's possible."
- Q1 FY2026 — Bob: "We think we can rationalize that by maybe as much as 25%" (call centers).
- Q1 FY2026 — Bob: "We're going to be able to cut back on research. We're going to be able to cut back on our human resources or people organization."
- Q1 FY2026 — Bob: "our move into critical infrastructure and data center services is going to be at least as profound as our move into outsourcing... and much faster."
- Q1 FY2026 — Bob: "growing almost 50% this year" (infrastructure services revenue, AI secular tailwind framing).
- Q1 FY2026 — Bob: "There will be some eliminations there" (efficiency areas after tool build/implement/reorg).
- Q1 FY2026 — Emma: "even those head count reductions, we anticipate happening a few years from now versus immediately."
- Q1 FY2026 — Emma: "I don't expect us to be investing in specifically AI companies"
TRACK RECORD — PROMISE vs DELIVERY
—/100 (no quantified promises) no-quantified-promises 6 calls reviewed
Across six calls (Q4 FY2024–Q1 FY2026), CBRE did not state a single quantified AI deployment or AI-outcome target with both a number and a deadline; AI commentary is qualitative until Q4 FY2025 and even then avoids metrics (e.g., ~25% call-center rationalization has no milestone). Forward numeric guides are data-center/infrastructure revenue and EPS, not AI delivery commitments, so AI promise-vs-delivery credibility is not scoreable on this transcript set.
PRICED-IN (REFINED)
MEDIUMEst. revisions flat · Fwd P/E 20.2 · EV/Sales 1.0x
AI claim maps to Advisory Services Segment, Project Management
Analyst rating counts improved only marginally (buy 6→7, hold 2→1, strongSell 1→0) while price targets are not in a raise sequence (lastYear 179.33 > lastMonth 178 > lastQuarter 177.33), so revision momentum is mixed-to-flat, not clearly rising. Forward valuation is moderate—not extreme for a mature services name (fwd P/E ~20.2, EV/Sales ~1.0, fwd PEG ~2.0)—but forward EPS already steps up materially (6.33→7.69→8.82), baking growth into the multiple. AI-driven efficiency/revenue would most plausibly flow through Advisory Services and Project Management, not the smaller Real Estate Investments line; with flat revisions and only moderate multiples, AI upside is partly but not fully reflected—hence medium rather than high (no rising estimates + rich valuation) or low (growth and bullish consensus already embedded).
COVERAGE — ENTHUSIASM TRAJECTORY + CATALYSTS
2Q4 FY20242Q1 FY20252Q2 FY20253Q3 FY20256Q4 FY20253Q1 FY2026
AI enthusiasm across 6 calls — trend → flat
No own-AI story for five quarters; Q4 FY2025 framed efficiency and data moat, then Q1 FY2026 dropped it.
RECENT AI CATALYSTS & NEWS
BUSINESS IMPACT - QUALITATIVE MATERIALITY
5/10 qualitative impact moderate medium-term · mixed evidence
Where AI matters: broker data tools and deferred corporate cost takeout
Adopter AI is real (research/call-center ~25% targets, broker-facing data tools) but dollars and FY26 EPS are immaterial and headcount cuts are years out; the hard uplift is mostly supplier-side infrastructure into the AI buildout, which guardrails exclude from adoption upside.
Caveats: Quantified revenue growth is largely selling into data-center/AI capex, not AI improving CBRE's own product economics; ~25% cost saves lack disclosed bases and are deferred, so near-term adopter P&L impact may be nil; Transactional brokerage and data layers face disintermediation pressure even if management is confident on large deals
AI DISRUPTION / CANNIBALIZATION RISK two-sided · 4/10
AI can compress valuation and research/data economics (Asia appraisal precedent: lower revenue per unit) and may erode office demand, but complex brokerage, relationships, and negotiation remain defensible and mgmt is adding brokers, not shrinking the core fee pool.
OPTIONS / MARKET STRUCTURE
option liquidity: good
proxy inputs — dollar-ADV $281M · beta 1.278 · px $127.86
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
Undercutting — insiders selling, institutions flat, management language 1/10 hedged.
INSIDERS selling 5 open-market sell(s) vs 0 buy(s) — net distribution
INSTITUTIONS (13F) flat as of 2026-03-31: 116 new / 151 closed positions; 512 increased / 433 reduced; institutional ownership -1.68pp; -28 net 13F holders
MGMT LANGUAGE 1/10 hedged Transcript has no AI/ML/automation for CBRE; only data-center real estate/infrastructure.
VERBATIM AI QUOTES
“Before I turn the call over to Emma, I want to address AI. We spend a lot of time thinking about this topic, and we know it's top of mind for investors.”
— Robert Sulentic, Q4 FY2025
“We are using AI today in 2 areas. The first is efficiency. We're deploying AI where its economic value clearly exceeds the economic value of traditional efficiency levers like offshoring. We're very disciplined about understanding the trade-offs before pursuing efficiency-related AI investments.”
— Robert Sulentic, Q4 FY2025
“The second area is developing a knowledge advantage to differentiate our product offerings. CBRE has more real estate data than any company in the world. Historically, we have not been able to turn this enormous base of knowledge into a comparably large competitive advantage. With the use of AI, we are moving toward gaining advantages that are more in proportion to the data advantage that comes with our market position.”
— Robert Sulentic, Q4 FY2025
“For instance, we've known for some time that our opportunity in the brokerage business is enabled by but not anchored to market data.”
— Robert Sulentic, Q4 FY2025
“AI can both enable and disintermediate the data and knowledge side of this.”
— Robert Sulentic, Q4 FY2025
“On the market-facing side, we believe there is and will be massive opportunity with owners and operators of data centers and digital infrastructure.”
— Robert Sulentic, Q4 FY2025
“On balance, when you add all of this up, there will be risks, risk mitigants and opportunities in our business associated with AI. We are optimistic that the net impact will benefit CBRE in the long run.”
— Robert Sulentic, Q4 FY2025
“What we're really working hard to do in thinking we're finally making some gains on using AI is to provide data to our brokers in a more efficient way and a more cost-effective way for us.”
— Robert Sulentic, Q4 FY2025
“In the long run, will there be less office users because AI disintermediates some of the work people do, that's possible. But what we're likely to see is a lot more AI-related workers backfill other types of workers that may go away because of AI.”
— Robert Sulentic, Q4 FY2025
“What's happening is that talent being used to support the growth of AI.”
— Robert Sulentic, Q4 FY2025
“We think by the end of 2026, we'll -- there'll be concrete evidence that we've made some real gains in terms of extracting the data we have assimilating it and delivering it to our professionals in a way that we haven't done before. And that is being enabled by AI, and that's one of the areas we're most encouraged to buy.”
— Robert Sulentic, Q4 FY2025
“We think we're going to be able to save over this year maybe extending into next year, about 25% of the cost of our research work using AI and the data that we assemble and manipulate through AI to support the research work.”
— Robert Sulentic, Q4 FY2025
“We are not reducing broker headcount. We're adding brokers.”
— Robert Sulentic, Q4 FY2025
“Big theme, obviously, is what's going on with artificial intelligence, big theme is what's going on with job creation and all the old jobs disappear.”
— Robert Sulentic, Q1 FY2026
“The whole AI job creation or job destruction thing that is unfolding and ping-ponging back and forth, lots of discussion around that, lots of headlines around all the jobs that are going to be eliminated by AI.”
— Robert Sulentic, Q1 FY2026
“I can tell you for our company when we look at what's going on, we anticipate some job loss in certain areas. So we have AI initiatives underway to create efficiencies in the company. And so for instance, we have lots of people in call centers around the world, thousands of them. We think some of that -- we think we can rationalize that by maybe as much as 25%. We're going to be able to cut back on research. We're going to be able to cut back on our human resources or people organization.”
— Robert Sulentic, Q1 FY2026
“AI is creating a considerable secular tailwind for our company right now, and it's fairly broad-based.”
— Robert Sulentic, Q1 FY2026
“To the point where I think our move into critical infrastructure and data center services is going to be at least as profound as our move into outsourcing was in the '90s and early 2000s and much faster.”
— Robert Sulentic, Q1 FY2026
“We are developing AI-enabled tools in every 1 of those areas.”
— Robert Sulentic, Q1 FY2026
“This is where there's going to be some potential loss of employees.”
— Robert Sulentic, Q1 FY2026
“We're going to see some efficiency in our offshore service centers. We're going to see some efficiency in the research area and financial planning and analysis and human resources.”
— Robert Sulentic, Q1 FY2026
“Where we think we're most protected, and I commented on this last quarter, is in our transactional businesses.”
— Robert Sulentic, Q1 FY2026
“That land development business is not going to be disintermediated by AI. It's going to be enabled by AI.”
— Robert Sulentic, Q1 FY2026
“One of the things we're watching very closely is it can get really expensive really fast if you don't control who has the access to use it and what they can use it for.”
— Robert Sulentic, Q1 FY2026
“we are controlling it controlling who has access to it, controlling what we use it for.”
— Robert Sulentic, Q1 FY2026
“In terms of investing in AI, as Bob said, it's similar to how we invest in technology. And we're constantly organically investing through our CapEx in technology and now AI to support our business, and that will remain unchanged. I don't expect us to be investing in specifically AI companies, like we didn't make large investments in technology companies historically.”
— Emma Giamartino, Q1 FY2026
“it will take a number of years, and it will start in our functions. I mean, Bob mentioned our HR teams, our shared service teams, but even those head count reductions, we anticipate happening a few years from now versus immediately.”
— Emma Giamartino, Q1 FY2026
ANALYST QUESTIONS ON AI
Q (Q4 FY2025, Stephen Sheldon (William Blair)): Appreciated the commentary around AI opportunities and risk; asked separately about capital markets pipeline and rate dependence (not AI).
A: No further AI answer beyond prepared remarks on this question.
Q (Q4 FY2025, Julien Blouin (Goldman Sachs)): Do you think there's a risk that AI maybe eats into some of these more market-making aspects of your brokerage business?
A: Bob: Clients want brokers' strategic input, negotiation, and relationships on big complex transactions—not leads from online; confident that business is driven by strategic creative thinking; using AI to provide data to brokers more efficiently and cost-effectively; same skills protect investment/development businesses; no evidence to the contrary.
Q (Q4 FY2025, Anthony Paolone (JPMorgan)): What do you think the impact might be on your end markets, particularly around office and whether you see any long-term diminution in space needs; and in areas like appraisal that can get streamlined and perhaps reduce fees?
A: Bob: Long-run fewer office workers from AI is possible but hard to predict; today companies use offices to attract talent; AI-related workers may backfill; office leasing is strong. On appraisals: Asia-Pacific automation lowered revenue per appraisal but raised volume sharply and profitability; valuations may face disintermediation pressure but scale could make CBRE a net winner; valves business expected to grow ~10% next year.
Q (Q4 FY2025, Steve Sakwa (Evercore ISI)): What visibility do you have on the data center business broadly, how far out can you see it, and are there longer-term bubble concerns?
A: Bob: Pipeline and duration support growth for a few years; talent shortage to support AI/data center demand conflicts with disintermediation fears; integrated data center solutions ~$2B revenue in 2026; not worried about bubble or running out of opportunity; limited ownership exposure; land sites are additive with little balance sheet risk; post-build maintenance/refit/management work will be substantial.
Q (Q4 FY2025, Ronald Kamdem (Morgan Stanley)): What are some of the moats to replicating CBRE's data advantage with an AI tool, and how long could it take?
A: Bob: By end of 2026 expect concrete evidence of gains extracting, assimilating, and delivering data to professionals—enabled by AI; will save money accumulating/buying data and make brokers more efficient; same tools to meaningfully cut research cost.
Q (Q4 FY2025, Alex Kramm (UBS)): On savings: how much spending on external vendors could be cut as AI improves; is research just reshaping internally; any external data monetization opportunities?
A: Bob: Empirical ~25% research cost savings this year maybe into next year via AI and assembled data; brokerage data spend savings from own collection plus purchased data but not specifying where; most excited about delivering data to brokers more usefully/self-serve; did not expand on external monetization beyond enabling the business.
Q (Q4 FY2025, Seth Bergey (Citi)): How do you think about head count needs balancing accelerating advisory and longer-term AI efficiency gains?
A: Bob: Not reducing broker headcount—adding brokers; savings in research, cost of data, and efficiency of delivering data to brokers.
Q (Q1 FY2026, Steve Sakwa (Evercore ISI)): C-suite macro views including AI job creation/destruction; what would challenge leasing and sales in 2H?
A: Bob: Headlines on AI job losses don't match client conversations; average office lease length hasn't decreased; CBRE has AI efficiency initiatives (call centers ~25%, research, HR) but biggest issue is inability to hire enough skilled people for critical infrastructure; net-net not a lot of fear in the foreseeable future.
Q (Q1 FY2026, Stephen Sheldon (William Blair)): Meta training partnership on data center capabilities—similar opportunities with other big tech and AI companies; one-time vs recurring revenue?
A: Bob: Definitively not one-time; building capability in multiple U.S. cities to recruit, train, and place technical people for Meta's data center initiative, including into CBRE teams, competitors, and the market; expects an enduring service; broad base of critical infrastructure/data center support opportunities with hyperscalers.
Q (Q1 FY2026, Julien Blouin (Goldman Sachs)): How have thoughts on AI risk evolved since last quarter—still BOE disintermediation risk vs capital markets; concern about AI proptech startups bypassing brokers in small deals and moving upmarket?
A: Bob: AI is a considerable secular tailwind ($3B infrastructure revenue 2025, ~$1B in Q1, ~50% growth, M&A optionality); developing AI-enabled tools across segments; efficiency will eliminate some roles in offshore centers, research, FP&A, HR over time; transactional/strategic brokerage, investing, and development are protected—broker expense is mostly brokers not data grinding; land development enabled not disintermediated; dismisses proptech disintermediation anecdotes without revenue proof.
Q (Q1 FY2026, Jade Rahmani (KBW)): AI rollout—what percentage of teams use it, any limits, and how to maintain a closed-loop system protecting proprietary data?
A: Bob: Still working through access and use cases; COO Vikram Kohli monitoring cost vs benefit; controlling who has access and what it is used for; measured expansion where benefit balances cost.
Q (Q1 FY2026, Seth Bergey (Citi)): Has AI changed capital allocation priorities (buybacks vs resilient bolt-ons vs investing in AI companies)?
A: Emma: Priorities unchanged—M&A first, especially data centers; buybacks fill gaps when stock undervalued; organic CapEx for technology and now AI like historical tech spend; do not expect investing in standalone AI companies.
Q (Q1 FY2026, Seth Bergey (Citi)): How could raising headcount in some areas while using AI for productivity change segment margins over time?
A: Emma: Hard to speculate; will take years; starts in corporate functions (HR, shared services); headcount reductions anticipated a few years out, not immediately.