Sunbelt Rentals Holdings, Inc. (NYSE: SUNB) — The #2 Rental Giant Comes Home to America, Cheaper Than the Champion It Chases, Into a Cooling Cycle

Independent fundamental research. The analysis that follows carries no recommendation and no price target — it discusses valuation only as embedded expectations and scenarios. The single deliberate exception is the Claude's Take block immediately below.

⚡ Claude’s Take

This block is the author’s own independent opinion. It is general information, not investment advice. The analysis that follows carries no position and no price target.

Verdict: HOLD / accumulate-on-weakness. A genuinely world-class rental franchise at a fair-to-slightly-attractive price, but not a fat pitch at ~$74 with the cycle cooling. Buy the discount in the low-to-high $60s; don’t chase the re-rating. Directional valuation zone: fair value ~$80–95 (base case — holds ~8.3x EV/EBITDA on FY27 guidance plus buyback accretion); accumulation zone ~$60–70 (where the stock traded in its March debut dip, ~7x); the bull re-rate-to-United-Rentals-plus-S&P-inclusion case gets you ~$100–115, the genuine-downturn case ~$55–60.

This is the former Ashtead Group — one of the great industrial compounders of the last two decades (Bloomberg once headlined a ~93,000% lifetime return; it was the #1 FTSE 100 stock of 2021) — that has just re-domiciled to the United States, where ~98% of its operating profit is earned, and re-listed on the NYSE as Sunbelt Rentals. The variant-perception hook is simple and quantifiable: Sunbelt trades at roughly a 13% discount to United Rentals (URI) on EV/EBITDA (~8.3x vs ~9.5x) despite comparable ~42% margins, lower leverage (~1.6x vs 1.9x), and a far higher free-cash-flow yield (~6.7% vs ~2%). The entire re-listing is a bet that this gap closes — through a U.S. investor base, eventual S&P 500 inclusion, and the disappearance of the “UK-conglomerate-on-the-LSE” discount. Closing it fully is worth ~+20%. The framing is therefore a special-situation re-rating, not a deep-value entry — the factor tape (beta ~1.6, negative alpha since listing, no usable factor model on a 4-month history) reads as a high-quality cyclical whipsawed by mechanical index/rotation flows, neither a falling knife nor a runaway momentum long.

The reason I won’t pound the table at $74 is the cycle and the returns trend. U.S. non-residential construction has been contracting for three straight years (the Architecture Billings Index has sat below 50 since January 2023; May-2026 = 44.5), GAAP earnings have fallen three years running (EPS $3.68 → $3.15), rental-revenue growth has decelerated to ~+3%, and — the Marathon tell — ROIC has compressed from ~15% to ~11% as the asset base scaled, meaning the heavy fleet-and-bolt-on growth now earns progressively less. Margins are near record, so ~8.3x sits on arguably peak-ish EBITDA. The offset, and the reason it’s a HOLD not an AVOID, is the counter-cyclical free cash flow: in a downturn Sunbelt slashes fleet capex and FCF balloons (FY26 FCF was $2.05bn, a ~6.7% yield, with the buyback running at $1.4bn), giving real downside support and a self-funding return engine while you wait. Conviction: medium. The single fact that flips me bullish: a durable re-acceleration in Specialty + data-center mega-project revenue (the funnel jumped from ~$10bn to ~$25bn in Q4) combined with confirmed S&P 500 inclusion. The single fact that flips me bearish: used-equipment recovery rates breaking below ~45% (a fleet-glut signal) or ABI/non-resi rolling into an outright construction recession.

Tag: “The crown jewel comes home — better-margined than the champion it chases, cheaper, and riding a cooling cycle.”

📈 Stock Price Action — Five-Year Event Map

Sunbelt’s NYSE tape is only ~4 months old (first trade 2026-03-02). The five-year arc is Ashtead Group plc (LSE: AHT) in GBP — the same operating company, the same Sunbelt Rentals business, ~98% of profit always North American. Ashtead shares were suspended on the LSE at £53.26 on 2026-02-27 ahead of the court-sanctioned scheme of arrangement, then re-debuted as SUNB on the NYSE at $73.79 on 2026-03-02 (one SUNB share per Ashtead share). The arc: a COVID trough near £17–18 (March 2020) compounding to an all-time high near £64 (November 2021), a 2022 rate-shock de-rating, a range-bound 2023–2025 (~£40–58) on soft U.S. construction, and then the U.S. re-listing into a choppy debut tape — $73.79 open → ~$63 dip (late March) → ~$86.60 intraday high (18 June) → ~$74 after the 23-June FY26 print. As of 26-June-2026 the stock sits ~11% below its post-listing high and ~17% above its debut-week low; the 52-week-equivalent range since listing is roughly $63–87. (Price levels: FACT, from companiesmarketcap AHT history and the AZI price CSV. Attributed drivers: INTERPRETATION.)

Cycle narrative. Events 1–2 are the canonical high-quality-cyclical V-recovery: Ashtead 5x’d off the COVID low on U.S. stimulus and the structural rental tailwind, peaking ~£64 in late 2021. Event 3 was a multiple event — the 2022 rate shock de-rated expensive cyclicals without breaking earnings. Events 4–6 are the long plateau: earnings held, the mega-project story provided a floor, but soft U.S. non-residential construction and the December-2024 profit warning capped the stock in a wide range. Events 7–11 are the re-listing special situation: a clean GBP-to-USD handover, a mechanical debut dip, a sharp spring rally on U.S.-buyer and index-inclusion demand, and a pullback when the FY26 results confirmed that fundamental momentum has cooled to mid-single digits even as the positioning story stays alive.

1. Executive Summary

Sunbelt Rentals Holdings (NYSE/LSE: SUNB) is the former Ashtead Group plc — the #2 equipment-rental company in North America behind United Rentals and #1 in the United Kingdom — which completed a re-domicile to the United States via a UK scheme of arrangement and began trading on the NYSE on 2 March 2026. The business is unchanged; what changed is the listing venue, the reporting currency (now USD), the accounting framework (now U.S. GAAP), and the brand on the ticker (Ashtead → Sunbelt Rentals). The stated rationale is alignment with reality: ~91% of revenue and ~98% of operating profit are North American, and management is chasing a U.S. investor base, deeper liquidity, and “potential inclusion in U.S. equity indices.”

The business itself is high quality. Sunbelt rents a standardized, ~$19.2bn fleet (original equipment cost) across 1,611 stores to ~800,000 North American customers (none more than 1% of revenue), through three segments: North America General Tool (58% of revenue, a 51% segment EBITDA margin), North America Specialty (33%, the faster-growing, stickier, higher-return leg), and the United Kingdom (9%, a structurally lower-return drag). FY26 (ended 30 April 2026) produced record revenue of $11,154m (+3.4%), adjusted EBITDA of $4,677m (41.9% margin), GAAP net income of $1,325m (EPS $3.15; adjusted $3.72), and free cash flow of $2,055m — of which $1,877m was returned to shareholders ($1,413m buybacks + $464m dividends). The moat is real and Greenwald-classic: economies of scale plus local branch density, which compound for the largest two players and earn ~42% EBITDA margins and (historically) high-teens ROIC that sub-scale independents cannot match.

The tension is twofold. First, cyclicality at a soft point in the cycle: U.S. non-residential construction has contracted for three years (ABI below 50 since January 2023), rental-revenue growth has decelerated from double digits to ~+3%, and GAAP EPS has fallen three years running. Second, returns compression: ROIC has fallen from ~15% (FY19) to ~11% (FY25) as the asset base scaled — a Marathon capital-cycle warning that incremental growth now earns less, even if it still clears the cost of capital. Against that, the counter-cyclical cash flow (capex falls in downturns, so FCF rises) and a fortress-adjacent balance sheet (1.6x net leverage, investment grade, 9x interest coverage) provide genuine downside protection. At ~$74 the stock trades at ~8.3x EV/EBITDA and ~19.9x adjusted EPS — a ~13% discount to United Rentals — with the re-rating and index-inclusion optionality not fully in the price, and a mild cyclical haircut, not a downturn, embedded. The verdict the body builds toward: a structurally good business and a disciplined allocator, fairly-to-attractively priced, whose near-term earnings trajectory and the U.S. construction cycle are the swing variables.

2. Business Overview

What it does. Sunbelt rents construction, industrial, and general equipment on a short-term basis, supplemented by sales of used fleet, new equipment, merchandise, and consumables. The fleet spans aerial work platforms, forklifts and material handling, earthmoving, power generation, HVAC and climate control, pumps, scaffolding, trench safety, temporary structures, flooring, traffic management, and a long tail of specialty categories. Customers rent because owning depreciating, increasingly complex iron is uneconomic for project-based or intermittent needs — the secular logic of rental penetration.

Revenue model and composition. This is a ~92.5% rental-revenue business. FY26 total revenue of $11,154m comprised: equipment rentals $10,320m (+3.4%); sales of rental (used) equipment $451m; and sales of new equipment, merchandise, and consumables $383m. Used-equipment sales exist to manage fleet age and mix (a ~$65m gain on disposal in FY26), not as a profit center. The model is recurring in aggregate — an enormous volume of short-duration contracts across ~800,000 U.S. customers averaging ~$11,200 of annual rental spend each, with the top 10 customers under 10% of revenue and no single customer above 1% — but each contract is short-term and demand is tied to construction and industrial activity. Revenue recurs the way a hotel’s does: high occupancy across many small transactions, not multi-year contractual lock-in.

Segments (FY26).

*Group adjusted EBITDA is after central costs; segment figures are pre-corporate. North America is ~91% of revenue and ~98% of operating profit. The two North American segments earn 46–51% segment EBITDA margins; the UK earns ~25% — roughly half — and is being restructured (see Changes and Headwinds).

Footprint and scale. 1,428 North American stores (814 General Tool, 614 Specialty) across all 50 U.S. states and 8 Canadian provinces, present in all top-100 North American markets; 183 UK stores (with a handful in Ireland, Germany, and the Netherlands). ~26,000 employees. The fleet is ~$19.2bn at original equipment cost, carried at ~$11.2bn net book value, with an average age of 53 months (up from 49 a year earlier — a modest aging as capex was trimmed). FY26 added 75 locations (51 greenfield + 24 via bolt-on acquisition).

Verdict. A simple, durable, high-margin rental model with extraordinary customer granularity and a genuine recurring-in-aggregate revenue base — but one whose top line is unambiguously tied to the construction and industrial cycle. The quality is in the margin structure and the asset-light-relative-to-peers branch network; the vulnerability is the demand cyclicality the model cannot escape.

3. Industry Dynamics

Market size and growth. The North American construction-and-industrial-equipment plus general-tool rental market is roughly $90bn (the ARA’s Q2-2026 forecast puts the U.S. at ~$83.5bn in 2026, +3.6%, with Canada ~$6.3bn), growing a forecast +3.8% (2027) and +4.4% (2028) — decelerating from the post-COVID boom (+13% in 2023, +8% in 2024) to a mature mid-single-digit pace. This is a GDP-plus industry where market growth alone will not carry the equity story; share gains and rising penetration must.

The penetration tailwind (the single most important structural fact). North American equipment-rental penetration — the share of equipment used that is rented rather than owned — rose for a fourth consecutive year to ~57% in 2024, above its pre-pandemic peak, and management sees a path “well over 60%” over the medium-to-long term. The drivers are structural and durable: contractors avoid owning capital-intensive, fast-obsolescing iron amid project uncertainty; renting preserves balance-sheet flexibility; and the rising cost and complexity of ownership (Tier-4 emissions compliance, telematics, maintenance-labor scarcity, safety/environmental regulation) makes renting cheaper, easier, and safer. Each point of penetration is roughly a 1–2% structural volume tailwind on top of construction-cycle volume — which is why the industry compounds through cycles even when core non-residential construction is flat.

Fragmentation and consolidation runway. The “big two” — United Rentals (~15% share) and Sunbelt (~11%) — together hold only ~26% of the North American market; the top three (adding Herc) reach ~35–40%. The remaining ~60% is a long tail of regional independents, ~40%+ of which have five or fewer locations. This is the consolidation flywheel: the leaders buy sub-scale independents at mid-single-digit EBITDA multiples and fold their fleet into a denser, higher-utilization branch network. The 2025 Herc/H&E bidding war ($5.3bn, won from United Rentals) confirmed both the scarcity value of scale assets and the leaders’ willingness to consolidate.

Capital-cycle location (Marathon lens): late-cycle, amber. Supply discipline among the majors is largely holding — rental rates are still positive (though decelerating), used-equipment recovery rates have normalized to ~50% from a 2022–23 peak above 60% (a healthy, not glutted, used market), and fleet OEC growth is modest (+3% at Sunbelt). But three warnings flash amber: (i) capex is re-accelerating (United Rentals guides 2026 gross capex to $4.3–4.7bn; Sunbelt FY27 gross rental capex to $2.45–2.85bn) into (ii) a three-year-contracting core non-residential backdrop (ABI 44.5 in May-2026), with (iii) the offsetting demand dangerously concentrated in data centers. Dodge projects mega-project starts rising from ~$765bn (2023–25) to >$1.5tn (2026–28), and FMI sees data-center construction +24.9% in 2026; meanwhile the earlier reshoring/manufacturing mega-project wave (semiconductor fabs, EV-battery plants) is flattening. The cycle is not yet at a fleet-glut top, but the demand bull case rests heavily on hyperscaler capex sustainability.

Verdict: a structurally good industry — for the scaled players (a B+). The secular penetration tailwind, the long consolidation runway, a healthy profit pool (40–46% EBITDA margins, ROIC above cost of capital through-cycle), and a genuine scale/density moat are durable positives. The permanent negatives are cyclicality and capital intensity: this is an asset-heavy business renting commodity iron tied to construction, where returns compress sharply in downturns. It is an excellent industry in which to own the leaders, and a poor one in which to be sub-scale — precisely the bifurcation that fuels consolidation.

4. Competitive Position

The moat: economies of scale plus local density. Sunbelt’s advantage is the Greenwald archetype of scale economies combined with local/regional customer captivity, operating at the market level rather than nationally. The mechanisms:

• Local density → utilization and logistics economics. Rental profitability is governed by fleet utilization and the cost of moving equipment. A dense cluster of branches in a metro (a top-25 U.S. market cluster can hold 15+ stores) lets Sunbelt share fleet, spare parts, drivers, and technicians across that market — spreading fixed costs over more revenue and enabling 24-hour delivery. A regional independent with one or two yards cannot match utilization, breadth, or service radius. Dollar utilization runs 55% group / 75% in Specialty.
• Scale purchasing and standardized fleet. ~$2bn+ of annual equipment purchasing, sourced from one or two suppliers per category, yields vendor discounts, supply priority, shared spares, and easy inter-store transfers that sub-scale players cannot replicate.
• National-account coverage → captivity. A national contractor — a data-center general contractor, an industrial-services firm — wants one vendor, one contract, one billing and telematics system, and consistent equipment specs across 30 states. Only United Rentals and Sunbelt can serve that. Re-bidding a national rental program is operationally painful and forfeits fleet-availability guarantees — a real, if convenience-based rather than contractual, switching cost.
• Data/telematics. >97% of the North American vehicle fleet carries telematics and ~95% has cameras, feeding utilization and predictive-maintenance advantages that widen with size and that small players cannot fund.

What financial outcome would deteriorate without it. Strip out density and scale and utilization falls toward independent-level economics, purchasing terms worsen, and the ~42% adjusted EBITDA margin and high-teens-historical ROIC compress toward the mid-single-digit returns that sub-scale yards earn. The moat is visible in the margin and the return — and in the ability to finance the multi-million-dollar fleet “load-in” on a mega-project that a small player simply could not.

Direct comparison vs United Rentals and Herc.

The reconciliation that matters: Sunbelt earns a higher EBITDA margin than United Rentals but a lower return on capital. Its margin edge is real (a high-margin General-Tool core at 51% segment EBITDA, plus high-margin Specialty and restoration), but United Rentals wins on scale, ROIC, U.S. national-account density, cold-start branch economics, and a far more mature capital-return engine — the attributes that actually determine #1 in a scale-economies business. Sunbelt’s lower group ROIC is dragged by (a) a younger, recently inflated fleet (newer iron at higher replacement cost depresses the return denominator) and (b) the structurally lower-return UK segment. Herc is a clear third tier — lower margins, ~5% ROIC, and a ~4x post-H&E balance sheet — not a quality comp.

Pressure-test. Can a regional independent compete? Yes, in a single local niche; no, for national accounts or broad-line breadth — and the economics prove it (independents earn lower returns and are the M&A feedstock). Are switching costs absolute? No — a determined large contractor can dual-source; this is captivity by convenience, not by contract. Does the moat survive a recession? Partly — cyclicality overrides the moat in a downturn (volume, rate, utilization, and residual values fall together), and the moat does nothing against residual-value risk. The clearest single proof the moat is local and not a transferable corporate capability is the UK: the same brand, playbook, and management earn ~2x the margin in North America as in the United Kingdom, because the density, penetration, and national-account structure that constitute the moat are far stronger in the U.S. market.

Verdict: a durable, scale-based advantage that is widening for the top two — but localized, convenience-based, and no shield against the cycle. It is a real moat (it surfaces in margins and ROIC-above-cost-of-capital through cycle), not a fortress.

5. Growth History and Forward Opportunities

History. Revenue compounded impressively off the COVID trough — $6,639m (FY21) → $7,962m (FY22, +20%) → $9,667m (FY23, +21%) → $10,859m (FY24, +12%) — before decelerating hard: $10,792m (FY25, −0.6%) → $11,154m (FY26, +3.4%). The deceleration is the central growth fact. It reflects the U.S. non-residential construction slowdown and softer rental rates, partly offset by the structural penetration tailwind, mega-projects, and Specialty. Growth has been ~70% organic (greenfields + same-store) and ~30% bolt-on M&A over the cycle; FY26 added 51 greenfields and 24 acquired locations and folded in 13 bolt-ons ($238m).

Quality of growth. Mixed and shifting. Specialty is the high-quality engine — rental revenue +5.8% for the year (and +15.1% in Q4), at a 46% segment EBITDA margin and 75% dollar utilization, in categories (power & HVAC, pumps, climate control, flooring, trench safety, scaffold, temporary structures) with lower rental penetration, higher service intensity, and less cyclicality than general tool. General Tool is the lower-quality, more cyclical leg — +2.1% rental growth, 47% utilization, an undifferentiated fleet (“similar to that of our peers,” per the 10-K). FY26 growth was volume- and utilization-led with roughly flat pricing, which is lower-quality than rate-led growth.

Forward opportunities.

• Specialty expansion — the strategic priority, broadening the highest-return product lines and cross-selling into the General-Tool customer base from co-located “clustered” stores. The Reliant/Aries acquisition ($650m, closed 1 May 2026) adds a new modular-space-solutions vertical (mobile offices, classrooms, storage), ~1% of FY27 revenue, EPS-accretive in year one but a year-one margin drag (sales-heavy mix shifting toward rental over time), with significant greenfield expansion planned.
• Mega-projects / data centers — the swing factor. Management defines mega-projects as >$400m construction value; the valuation of those in Sunbelt’s own funnel jumped from ~$10bn (Q1–Q3 FY26) to ~$25bn in Q4, the basis for FY27 optimism. Across a mega-project’s life the margin profile is “essentially the same as the business as a whole,” though early “load-in” periods (heavy fleet/labor deployment before billing ramps) cause a couple of quarters of margin drag — a live factor in late-FY26 compression.
• Greenfields — the “Sunbelt 4.0” plan targets 300–400 new North American locations over five years (99 added so far), densifying clusters.
• Penetration — the rising rent-vs-own share grows the addressable pie independent of the construction cycle and disproportionately benefits scaled players.

FY27 guidance: total revenue +4.5–7.5%; rental revenue +5–8%; adjusted EBITDA $4.85–5.05bn (vs $4.68bn in FY26); gross rental capex $2.45–2.85bn — a re-acceleration to mid-single-digit-plus growth, led by Specialty and mega-projects, against a “benign/flat” assumed core non-residential backdrop and “higher-for-longer” rates. Note a ~$70m FY26 events revenue tied to the FIFA World Cup that is largely non-recurring.

Verdict: decelerated, mixed-quality growth with a higher-quality forward mix. The headline growth rate has fallen to mid-single digits and is volume- not rate-led — lower quality. But the composition is improving as Specialty and mega-projects (higher-margin, stickier) take share from cyclical general tool, and the penetration tailwind is real. This is a mid-single-digit grower trying to up-mix its way to higher-quality growth, not a high-growth story.

6. Financial Quality

Revenue, margins, and the earnings trend. FY26 revenue was a record $11,154m (+3.4%), but profitability declined: adjusted EBITDA margin fell ~210bps to 41.9%, adjusted operating margin ~180bps to 22.4%, and GAAP net income fell to $1,325m ($3.15 EPS) from $1,510m ($3.47) in FY25 — the third consecutive year of GAAP EPS decline ($3.68 FY23 → $3.66 FY24 → $3.47 FY25 → $3.15 FY26). Margin-compression drivers: a higher Specialty/ancillary and sales mix, internal repair/repositioning and growth investment, mega-project load-in costs, UK weakness, the lapping of a +$28m FY25 receivables-provision reversal, and — importantly — one-time items.

Quality-of-earnings flags. Two are worth isolating:

1. FY26 one-time costs depress GAAP. FY26 carried ~$134m of restructuring (redomicile/U.S.-listing + UK operational restructure), plus $113m amortization of acquired intangibles and $65m stock-based comp, against just $15m of restructuring in FY25. The redomicile and UK-restructuring costs are genuinely non-recurring; adjusted EPS of $3.72 (vs $3.15 GAAP) is the better run-rate measure for FY26 specifically, with the redomicile costs falling away in FY27. (Caveat: the “adjusted” EBITDA also adds back items beyond the one-timers, so it is not a clean substitute — use both.)
2. Counter-cyclical cash flow is real but can flatter a soft year. FCF swings inversely with fleet capex. In fleet-growth FY24, FCF was just $126m (gross capex peaked); in soft FY26, with capex trimmed, FCF ballooned to $2,055m on $3,784m of operating cash flow. This is a genuine feature — the model self-funds returns in a downturn — but a high FCF print in a soft year is not the same as high-return growth, and the guided FY27 capex re-acceleration ($2.45–2.85bn gross) will lower FCF again.

Returns on capital — the Marathon tell. ROIC has compressed steadily: 14.9% (FY19) → 13.9% (FY23) → 12.3% (FY24) → 10.8% (FY25); ROE fell from 25.3% (FY23) to 17.5% (FY25). The company’s self-reported “return on investment” (a different, likely pre-tax/rental-asset denominator) was 14% in FY26 vs 15% in FY25 — same direction. Whether the slide is cyclical (post-COVID demand normalizing, fleet aging, rental-cost inflation in the denominator) or structural (industry fleet over-supply) is the key open question; either way, the incremental dollar of growth capital now earns less. ROIC still appears to clear a reasonable cost of capital (~8–9%), so this is returns compression, not value destruction — but it is the single most important metric to monitor.

Balance sheet and liquidity. Net debt ~$7,554m against ~$4.7bn adjusted EBITDA = net leverage ~1.6x, inside the 1–2x target; total-debt/EBITDA ~2.1x; EBITDA/interest ~9x; ~$3.5bn of availability; ~5% weighted cost of debt with a ~5-year average maturity. Investment-grade (BBB-area). Capital structure is an ABL facility plus senior notes, with ~$2.85bn of capital leases. The balance sheet is a genuine strength — lower-levered than United Rentals (1.9x) and far below Herc (~4x) — giving downturn cushion and M&A/buyback dry powder. Goodwill and intangibles total ~$3.8bn (roll-up legacy); a deep-downturn impairment is a tail risk but the businesses generate cash.

Unit economics. Equipment is depreciated to a salvage value of 10–15% of cost (range 0–35%); the gain on used-equipment disposal (~$65m FY26, down YoY) and the recovery rate are the key residual-value gauges to watch as the fleet ages (49 → 53 months) and capex re-accelerates.

Verdict: high-margin, cash-generative, conservatively financed — but with economics that are compressing, not improving, with scale right now. Do economics improve with scale? Historically yes (margins held ~45% as revenue doubled), but the recent ROIC trend says the marginal returns are deteriorating at this point in the cycle. The cash generation and balance sheet are unambiguous strengths; the returns trajectory is the watch-item.

7. Capital Allocation

Framework. A disciplined, returns-aware waterfall, explicitly stated: organic fleet investment first (highest-return use), then bolt-on M&A, then a progressive dividend, then buybacks — all maintained inside a 1–2x net-debt/EBITDA corridor. The framework demonstrably works counter-cyclically: Sunbelt invested heavily into FY23–24 demand (FY24 FCF just $126m, $846m of M&A) and then harvested in soft FY26 (FCF $2,055m, $1,413m of buybacks).

Track record.

• Buybacks. Share count fell from ~503m (FY16) to ~420m (weighted FY26), a ~16–17% decade reduction — a genuine, accretive shrink, not SBC-offset cosmetics. A new $1.5bn program launched on the NYSE-debut day (2 March 2026).
• Dividend. A roughly decade-long progressive dividend, raised to $1.125/share in FY26 (+4%), with no COVID cut, at a conservative ~36% payout — transitioning to quarterly payment in FY27.
• M&A discipline. A programmatic bolt-on roll-up — 13 deals in FY26 for ~$224m, habitually bought at mid-single-digit EBITDA multiples and folded into Sunbelt’s ~8x-rated, denser network (a multiple-arbitrage-plus-density value engine). The largest single deal in recent memory, Reliant/Aries at $650m, is small against a ~$38bn enterprise. There is no value-destroying mega-deal in the record — the opposite of United Rentals’ periodic large acquisitions.

The Marathon caveat. The asset base has grown enormously and per-share value has been created (EPS roughly doubled FY21→FY24), but returns on that growing capital are mean-reverting downward (ROIC 14.9% → 10.8%). The engine still creates per-share value, but at a declining marginal rate — the classic “high returns attract capital, then compress” pattern, here driven by the company’s own fleet growth and United Rentals’ and the broader industry’s. Disciplined allocation cannot fully outrun a maturing capital cycle.

Insider behavior — neutral, do not over-read. The Form 3/4 cluster around the re-domicile is entirely mechanical: the February-2026 Form 3s are initial-ownership filings triggered by U.S. registration; the March-2026 Form 4s are code-A conversions of Ashtead awards into SUNB shares (no cash paid); the June-2026 Form 4s are code-F shares withheld for tax on vesting (officers retained the large majority — Horgan kept 702,834 shares after withholding 24,567). Zero open-market purchases and zero discretionary sales — no conviction signal either way, though the large retained stakes are a modest positive.

Verdict: high-quality, disciplined capital allocation — among the best in the rental/industrials universe — marked down a half-grade only because returns on the growing capital base are visibly compressing. Counter-cyclical, shareholder-friendly, no empire-building, a real buyback and progressive dividend. The watch-item is not whether they allocate well but whether a now-$38bn enterprise at ~11% ROIC can replicate the historical compounding.

8. Changes and Headwinds — Last Two Years

The re-domicile (the defining change). Announced 10 December 2024, approved by shareholders 10 June 2025, implemented via a UK court-sanctioned scheme of arrangement, and completed 27 February 2026: Sunbelt Rentals Holdings, Inc. (Delaware) became the parent, each Ashtead share converted 1:1 into SUNB, NYSE trading began 2 March 2026 with a secondary LSE listing, the auditor changed from PwC UK to PwC US, and reporting converted to U.S. GAAP and USD. The rationale: ~98% of FY24 operating profit is North American, so align the listing, governance, audit, and currency with the business — and capture a U.S. investor base, deeper liquidity, brand profile, simpler U.S. employee ownership, and “potential inclusion in U.S. equity indices” (the S&P 500 ambition). The risk factor is candid: the benefits “may not be realized fully, or at all,” may take longer, or cost more; FY26 absorbed one-time costs and the company now carries first-year U.S.-registrant internal-controls and comparability risk.

The December-2024 profit warning. Issued alongside the move announcement: U.S. rental-revenue growth guidance was cut to +2–4% from +4–7% on soft U.S. commercial construction and higher-for-longer rates. This crystallized the cyclical deceleration that has since defined the numbers.

Management/board. CFO Alex Pease (ex-WestRock) joined as designate in October 2024 and became CFO in March 2025 — a credentialed U.S.-capital-markets hire to professionalize finance/IR for the listing. CEO Brendan Horgan (since 2019, with Sunbelt since 1996) and Chairman Paul Walker (since 2018) provide continuity; no executive turnover around the move. The board was refreshed with U.S./UK industrial directors.

The dividend modernization and a new buyback (quarterly dividend from FY27; fresh $1.5bn program) accompany the listing. Reliant/Aries ($650m, modular space) closed just after year-end (1 May 2026), opening a new Specialty vertical.

Headwinds. U.S. non-residential construction contraction (ABI sub-50 for three years); decelerating rental rates; an aging fleet and re-accelerating capex pressuring near-term FCF; UK weakness (now in restructuring); residual-value normalization; and tariff/inflation exposure on imported equipment that “may not be able to pass through to customers.”

Verdict: the changes are net thesis-strengthening on positioning but neutral-to-negative on fundamentals. The re-domicile is a sensible structural alignment with genuine re-rating/index optionality and a modernized capital-return cadence; the offsetting reality is that it lands into a cooling cycle, with a profit warning, three years of GAAP EPS decline, and compressing returns as the backdrop.

9. Risk Analysis

Catastrophic-loss / total-loss assessment. The probability of a permanent total loss is very low: the balance sheet is investment-grade at ~1.6x leverage, the model generates counter-cyclical cash, and the fleet is real, fungible, saleable collateral. The realistic downside is a cyclical 30–40% drawdown in a construction recession (volume, rate, utilization, and residual values falling together, with operating leverage amplifying the EBITDA decline and the multiple de-rating), not impairment of the enterprise.

10. Valuation Discussion (Embedded Expectations)

Note on data. ROIC.ai’s enterprise-value field for SUNB is broken (a stale ~$10.3bn from the GBP/float mapping) and AZI’s own-history valuation percentiles are empty (the U.S. listing is too new) — so the multiples below are computed from first principles.

Current multiples at ~$74. ~414m shares × ~$74 = ~$30.6bn market cap; + ~$7.55bn net debt = EV ~$38.2bn. Against FY26 figures: EV/adjusted EBITDA ~8.3x, EV/Sales ~3.4x, GAAP P/E ~23.5x, adjusted P/E ~19.9x, FCF yield ~6.7% ($2,055m / $30.6bn), dividend yield ~1.5%.

Peer comparison.

The central valuation fact: Sunbelt trades at a ~13% discount to United Rentals on EV/EBITDA (~8.3x vs 9.5x), despite a comparable ~42% margin, lower leverage, and a much higher FCF yield. Herc is not a clean comp (its multiple and ROIC are H&E-distorted). The re-rating thesis is the closing of the SUNB-vs-URI gap, aided by potential S&P 500 inclusion now that SUNB is a U.S.-domiciled NYSE primary listing. The arithmetic of full convergence: at URI’s 9.5x on FY26 EBITDA, EV would be ~$44.4bn; less ~$7.55bn net debt = ~$36.9bn equity ÷ 414m ≈ ~$89/share, roughly +20% holding EBITDA flat. The current ~8.3x is the market applying a ~13% “transition / UK-drag / possible-peak-earnings / execution” haircut to URI’s multiple.

Embedded expectations. At ~8.3x trailing EV/EBITDA and ~19.9x adjusted EPS for a beta-~1.6 cyclical at a soft point in the U.S. construction cycle, the price is underwriting: (i) modest, durable growth broadly consistent with the FY27 guide (+4.5–7.5% revenue, $4.85–5.05bn EBITDA) — not acceleration, but not a downturn; (ii) a cyclical-mid, not cyclical-peak read, with margins near record implying some discount for “are these peak rental rates?”; and (iii) optionality not in the price — full URI-gap closure and S&P inclusion are upside the current multiple does not capture. The price embeds the guide plus a mild cyclical haircut.

Scenario analysis (embedded expectations, NOT a price target).

The spread (~$58 bear / ~$83 base / ~$110 bull) shows the price sits near the base/mid: the market is paying for the guide, discounting only a mild cyclical haircut, and not yet paying for the re-rating. No price target and no recommendation — these scenarios characterize the embedded expectations only.

11. Variant Perception

Consensus belief. Sunbelt is a high-quality #2 rental compounder that just re-listed in the U.S.; the sell side is constructive (e.g., Citigroup Buy, target raised to $95) on the re-rating-toward-URI plus S&P-inclusion story and the structural penetration/mega-project tailwinds. Consensus expects mid-single-digit-plus growth to resume on data centers and Specialty, with the discount to United Rentals closing.

The strongest bull case. A best-in-class operator with a genuine scale/density moat and the highest General-Tool margins in the industry, trading at a ~13% discount to its larger peer with lower leverage and a ~6.7% FCF yield, re-domiciled into the market where 98% of its profit is earned — a clean catalyst path (U.S. investor base → S&P 500 inclusion → multiple convergence ≈ +20%), on top of a structural penetration tailwind and a data-center mega-project funnel that more than doubled to ~$25bn in a single quarter. Counter-cyclical FCF funds an aggressive buyback while you wait. The 2010s proved this management lineage can compound.

The strongest bear case. This is a capital-intensive, asset-heavy cyclical at a soft point in its cycle, dressed up as a re-rating story. ABI has been sub-50 for three years; GAAP EPS has fallen three years running; rental growth has decelerated to ~+3% and is volume- not rate-led; and ROIC has compressed from ~15% to ~11% — the textbook asset-growth-anomaly signal that the growth is earning less. Margins are near record, so ~8.3x sits on peak-ish EBITDA, and the demand bull case is dangerously concentrated in hyperscaler data-center capex that could air-pocket. The discount to United Rentals may be deserved (lower ROIC, the UK drag, peak-margin risk), and S&P inclusion is a one-time flow, not a durable value creator. A genuine construction recession takes EBITDA, the multiple, and residual values down together (~$58, the bear scenario).

The 3–5 assumptions that matter most.

1. Is FY26 EBITDA near a cyclical peak or a mid-cycle base? (Margin near record; the whole multiple debate turns on this.)
2. Does the data-center / Specialty demand layer durably offset core non-residential weakness — or roll over with hyperscaler capex?
3. Does the SUNB-vs-URI discount close (re-rating + S&P inclusion) — or is it a deserved quality/ROIC discount?
4. Is the ROIC slide cyclical (reverses) or structural (industry over-fleeting)?
5. Does management fix or divest the lower-return UK, an ROIC-accretive lever?

Falsification evidence. The bull case breaks if used-equipment recovery rates fall below ~45% and/or fleet productivity turns negative (a fleet-glut / pricing-give-back signal), or if data-center construction growth stalls. The bear case breaks if Specialty + mega-projects drive durable high-single-digit rental growth with stable rates and S&P inclusion is confirmed with the discount to URI closing. Factor-positioning input: with no usable factor model on a 4-month history, the empirical reads are beta ~1.6 and negative alpha since listing — a high-quality cyclical whipsawed by mechanical index/rotation flows, consistent with a special-situation re-rating riding a cooling cycle rather than either a falling knife or a crowded momentum long. Consensus may be offside in underestimating the cyclical risk if it is extrapolating the re-rating while the construction cycle deteriorates.

12. Fact vs. Interpretation Table

13. Open Questions

1. Peak or mid-cycle EBITDA? Are FY26’s near-record margins a cyclical peak (making ~8.3x optically cheap on inflated earnings) or a sustainable base?
2. Cyclical or structural ROIC slide? Does ROIC recover as the cycle turns, or has the industry over-fleeted into a durably lower-return regime?
3. Data-center durability. How concentrated is Sunbelt’s mega-project exposure in data centers, and what is the air-pocket risk if hyperscaler capex slows? (No hard-dollar data-center revenue is disclosed — only the ~$25bn funnel.)
4. S&P 500 inclusion — timing, probability, and the size of the passive flow. Named as a potential benefit only.
5. UK — fix, shrink, or divest? A divestiture would be ROIC-accretive and a plausible value-unlock.
6. First U.S. proxy (Sep-2026) — does the new comp plan add a one-time U.S.-listing mega-grant? ISS/Glass Lewis already flagged the rising quantum as “excessive.”
7. Residual values — how far do used-equipment recovery rates fall as the fleet ages and capex re-accelerates into a soft construction backdrop?

14. What Must Be True

For the bull case to be right:

• The U.S. non-residential cycle troughs and Specialty + data-center mega-projects drive durable high-single-digit rental-revenue growth with stable-to-rising rates — so FY26 EBITDA proves a base, not a peak.
• ROIC stabilizes/recovers toward the mid-teens as fleet inflation washes through and utilization rises — confirming the slide was cyclical.
• The SUNB-vs-URI discount closes (U.S. investor base + S&P 500 inclusion + recognition of the margin/leverage/FCF edge), delivering the ~+20% re-rating.
• Falsification test: the bull case is wrong if used-equipment recovery rates break below ~45% or fleet productivity turns negative (fleet glut / rate give-back), or if data-center construction growth stalls and core non-resi keeps contracting — i.e., EBITDA fails to grow and the multiple de-rates.

For the bear case to be right:

• The U.S. construction cycle deteriorates into an outright downturn; EBITDA goes flat-to-down as operating leverage bites; residual values fall, hitting disposal gains and depreciation assumptions.
• The discount to United Rentals is deserved and persists (lower ROIC, UK drag, peak-margin risk); S&P inclusion proves a one-time flow, not a re-rating; the multiple de-rates to ~7x (~$58).
• Falsification test: the bear case is wrong if Specialty + mega-projects sustain high-single-digit rental growth with stable rates and S&P 500 inclusion is confirmed with the URI discount closing — i.e., both earnings and the multiple rise together.

15. Source Appendix

See the Source Appendix below for the full citation list. Primary sources: SUNB FY26 Form 10-K (filed 2026-06-23); SUNB FY26 Q4/full-year press release (8-K Ex-99.1, 2026-06-23); SUNB Form 3/4/S-8 filings (Feb–Jun 2026); SUNB Q4-FY26 earnings-call transcript (2026-06-23); third-party financial-data providers (multi-year financials for SUNB and United Rentals; SUNB daily price history and news); American Rental Association forecasts; AIA/Deltek Architecture Billings Index; Dodge Construction Network and FMI mega-project/data-center forecasts; Ashtead Group historical price (companiesmarketcap) and annual-report remuneration disclosures; Bloomberg/Reuters/Investing.com archival coverage. Third-party aggregated data is reconciled to primary filings; management commentary is treated as hypothesis, validated against filings and external data.

APPENDIX A — Standard Diligence Questionnaire

Sunbelt Rentals Holdings, Inc. (NYSE: SUNB) — supplemental to the research memo. Fact/Interpretation/Assumption labels applied where it matters.

General

What thoughtful questions have other investors asked about this company? (1) Is the re-domicile a genuine re-rating catalyst (US investor base + S&P 500 inclusion) or a cosmetic move? (2) Is FY26’s ~42% EBITDA margin a cyclical peak? (3) Why does the #2 player earn a higher EBITDA margin than #1 United Rentals but a lower ROIC? (4) How exposed is the mega-project funnel to a data-center capex air-pocket? (5) Will management fix or divest the low-return UK? (6) Is the ROIC slide (15%→11%) cyclical or structural?

Cyclicality & Earnings Nature

Cyclical high or low? Interpretation: mid-to-late cycle on the soft side — US non-residential construction has contracted for three years (ABI <50 since Jan-2023; 44.5 May-2026), rental growth has decelerated to +3.4%, and GAAP EPS has fallen three straight years. Margins, however, are near record — so earnings are not at a clean trough; the cycle is soft but margins are elevated. External environment or internal actions? Both: external (US construction/rate cycle) drives the deceleration; internal actions (Specialty up-mix, greenfields, bolt-ons, buybacks, the re-domicile) are offsetting/transformational. Revenue stability? Recurring in aggregate (~800,000 customers, none >1%, ~92% rental revenue) but each contract is short-term and demand is cyclical. More stable than equipment manufacturing; less stable than a subscription model. Outlook / market size & direction? Fact: North American CIE+tool rental market ~$90bn, forecast +3.6–4.4%/yr (2026–28); penetration ~57% and rising toward 60%+. Growing, mostly domestic (NA ~91% of revenue), with a UK tail. The market grows GDP-plus; share gains and penetration are the equity drivers.

Business Quality & Competitive Moat

Industry more or less competitive? Consolidating (top-2 ~26%, top-3 ~35–40%; ~60% still fragmented independents) — competition for the biggest accounts narrows to ~3 players, while the long tail is being rolled up. Net: less competitive for scaled players, intensely competitive for sub-scale ones. How profitable (ROIC/ROE)? Fact: ROIC ~10.8% (FY25, down from 14.9% FY19); ROE 17.5% (FY25, down from 25.3% FY23); adj EBITDA margin 41.9%. Above cost of capital, but compressing. Industry profitability / barriers? Healthy profit pool (40–46% EBITDA for the majors); barriers = scale economies + local branch density + national-account coverage + purchasing power. Real but localized barriers. Easily understood? Yes — rent equipment, sweat the fleet, manage utilization and residual values. Undermined by foreign low-cost labor? No — a physical, local-logistics, service-intensive business; not offshorable. Do brands matter? Modestly — “Sunbelt Rentals” is a valued trade name and national-account credential, but the binding moat is density/scale/availability, not brand. Nature of competition? Availability, breadth, delivery speed, fleet reliability, national-account integration, price. Local/national, not global. Switching costs? Moderate for national accounts (program re-bid, spec consistency, integrated billing/telematics, fleet-availability SLAs); low for a single-site contractor. Captivity by convenience, not contract.

Financial Condition & Balance Sheet

Assets not fully on the balance sheet? The branch-density/national-account network and brand are intangible competitive assets not on the books. Conversely, ~$3.5bn goodwill + intangibles from the roll-up are on the books. Off-balance-sheet liabilities? Operating leases are capitalized (~$2.66bn ROU assets); the main exposure is fleet purchase commitments and the cyclical residual-value risk embedded in the $19.2bn OEC fleet. Accounting conservatism? Reasonable: equipment depreciated to 10–15% salvage; one-time redomicile/restructuring costs are disclosed and excluded from “adjusted.” Watch: the “adjusted EBITDA” basis adds back more than just one-timers — use GAAP and adjusted together. CapEx-hungry? Very — gross fleet capex ~$2.2bn FY26 (guided $2.45–2.85bn FY27); this is a capital-intensive, asset-heavy model. Counter-cyclical: capex (and thus FCF) swings inversely with the cycle.

Capital Allocation & Management

FCF generation & use / philosophy? Fact: FY26 FCF $2,055m; framework = organic fleet → bolt-on M&A → progressive dividend → buybacks, within 1–2x net leverage. FY26 returned $1,877m ($1,413m buybacks + $464m dividends). Significant acquisitions? Programmatic bolt-ons (13 in FY26, ~$224m, at mid-single-digit EBITDA multiples) plus Reliant/Aries modular ($650m, closed 1 May 2026). No value-destroying mega-deal. Buying back shares? Yes — share count ~503m (FY16) → ~420m (FY26), a ~16–17% reduction; new $1.5bn program from 2 Mar 2026. Issuing shares to insiders? Modest — SBC ~$65m FY26 (~5% of net income), more than offset by buybacks. Net share count falling. Compensation policy? Fact: LTIP weighted ~30% adjusted EPS + 30% return-on-investment + 30% relative TSR + 10% sustainability — strong per-share/returns alignment, no size/revenue metric. Watch: quantum rising toward US norms (ISS/Glass Lewis flagged “excessive”); first US proxy due Sep-2026. Management motivations? CEO Brendan Horgan (since 2019, with Sunbelt since 1996) holds ~703k shares; CFO Alex Pease (ex-WestRock). Disciplined, returns-focused; flag: the CEO’s brother (Kyle Horgan) is EVP Specialty (related-party).

Valuation & Market Data

ADR / MLP / K-1? No — SUNB is a Delaware C-corp common stock (NYSE primary, LSE secondary); ordinary 1099 treatment, not a K-1/MLP/ADR. Dividend policy? Progressive; $1.125/share FY26 (+4%, ~36% payout, ~1.5% yield); moving to quarterly payment in FY27. Profitability? High-margin (41.9% adj EBITDA), ~12% net margin, ROIC ~11%. Net income vs cash from operations? CFO ($3,784m) far exceeds net income ($1,325m) — normal for a depreciation-heavy rental model; FCF ($2,055m) is the cleaner figure and swings counter-cyclically.

Risks & Downside

What would cause the stock to decline? A US construction downturn; a data-center capex air-pocket; rate-driven multiple compression; falling used-equipment residual values; failure of the re-rating/S&P-inclusion thesis; continued ROIC compression. Catastrophic-loss risk? Low — investment-grade, ~1.6x leverage, counter-cyclical cash, fungible/saleable fleet collateral. Chance of total loss? Very low. The realistic downside is a cyclical 30–40% drawdown, not enterprise impairment.

Recent News & Events

Has the business environment changed? Yes — the re-domicile to the US/NYSE completed 27 Feb 2026 (trading 2 Mar 2026); a Dec-2024 profit warning crystallized the cyclical deceleration; CFO changed (Pease, Mar 2025); dividend moving to quarterly (FY27); fresh $1.5bn buyback. Significant acquisitions? Reliant/Aries modular ($650m, 1 May 2026) — a new Specialty vertical. Accounting-policy change? Yes — converted from IFRS to US GAAP and from GBP to USD reporting; auditor PwC UK → PwC US. Other recent changes? Brand Ashtead → Sunbelt Rentals; board refreshed; “Sunbelt 4.0” greenfield plan (300–400 NA locations); ~$70m FY26 FIFA World Cup events revenue (non-recurring).

APPENDIX B — Source Appendix

Sunbelt Rentals Holdings, Inc. (NYSE: SUNB) — research initiation, as of 2026-06-27. Primary sources before secondary; third-party aggregated data reconciled to filings. Management commentary treated as hypothesis.

Primary — SEC filings (CIK 0002083785)

• Form 10-K, FY ended 2026-04-30 (filed 2026-06-23) — business, segments, risk factors, MD&A, financial statements, LTIP metrics, share-based payments, redomicile disclosure. https://www.sec.gov/Archives/edgar/data/2083785/000162828026044888/snblt-20260430.htm
• Form 8-K + Ex-99.1 FY26 Q4/full-year press release (filed 2026-06-23) — FY26/Q4 results, segment tables, operating statistics, balance sheet, cash flow, FY27 guidance, Reliant/Aries acquisition, dividend, buyback. https://www.sec.gov/Archives/edgar/data/2083785/000162828026044797/sunbeltpressrelease-fy26q4.htm
• Form 8-K (2026-06-23) — results furnished; first annual meeting set 2026-09-01. https://www.sec.gov/Archives/edgar/data/2083785/000162828026044797/snblt-20260623.htm
• Form 10-Q (Q ended 2026-01-31) (filed 2026-03-12). https://www.sec.gov/Archives/edgar/data/2083785/000162828026017215/snblt-20260131.htm
• Form S-8 (filed 2026-03-02) — Ashtead Group LTIP 2021 + omnibus amendments. https://www.sec.gov/Archives/edgar/data/2083785/000119312526085489/d923968ds8.htm
• Form 3 cluster (2026-02-26) — initial beneficial-ownership filings (US-registration artifact; holdings only, no transactions).
• Form 4 cluster (2026-03-03) — code-A conversions of Ashtead awards into SUNB shares (no cash; redomicile mechanic). e.g. https://www.sec.gov/Archives/edgar/data/2083785/000119312526088692/xslF345X05/ownership.xml
• Form 4 cluster (2026-06-23) — code-F shares withheld for tax on vesting (not discretionary sales). e.g. https://www.sec.gov/Archives/edgar/data/2083785/000119312526279722/xslF345X06/ownership.xml
• Schedule 13G filings (Apr 2026) — institutional ownership.

Primary — transcripts

• SUNB Q4-FY26 earnings call (2026-06-23), via ROIC.ai — mega-project funnel ~$10bn→~$25bn (Q4); ~$70m FY26 FIFA World Cup events revenue; Aries/Reliant detail (17 locations, ~1% of FY27 revenue, year-one accretive/margin drag); FY27 guidance commentary. Prior calls Q1–Q2 FY26 and FY25 also available.

Secondary — financial/market data

• Third-party financial-data providers — SUNB income statement, balance sheet, cash flow, profitability/credit/valuation ratios and per-share data (FY2016–FY2025); United Rentals and Herc enterprise value and ratios. Aggregated data reconciled to primary filings; enterprise value computed from first principles (shares outstanding × price + net debt).
• SUNB daily price history, from 2026-03-02 (NYSE).
• News coverage (SUNB) — incl. Citigroup maintains Buy, target raised $85→$95 (Jun-2026).
• Factor/risk-model data (SUNB) — beta ~1.6, negative alpha since listing, market cap ~$30.5bn; full factor loadings unavailable on a sub-1-year US trading history.

Secondary — industry & macro

• American Rental Association (ARA) — North American equipment/event-rental forecasts (US ~$83.5bn 2026, +3.6%; +3.8% 2027; +4.4% 2028; Canada ~$6.3bn). https://news.ararental.org / https://ope-plus.com/2026/05/29/ara-updates-forecast-for-equipment-and-event-rental-markets/
• AIA/Deltek Architecture Billings Index — sub-50 since Jan-2023; 44.5 (May-2026). https://www.aia.org/resource-center/abi-may-2026-architecture-firm-billings-weaken-further
• Dodge Construction Network — mega-project starts pipeline ~$765bn (2023–25) → >$1.5tn (2026–28).
• FMI / ENR — data-center construction +24.9% (2026); hyperscaler capex ~$745–775bn (2026). https://www.enr.com/articles/62031-2026-forecast-megaprojects-data-centers-spur-growth-amid-shifting-policies
• RER / rermag.com — North American rental penetration ~57% (2024, 4th consecutive annual rise); Herc/H&E acquisition ($5.3bn). https://www.rermag.com

Secondary — peers & history

• United Rentals (URI) — ROIC.ai EV/ratios; stockanalysis.com / gurufocus (P/E ~23.9x, EV/EBITDA ~9.5–10x, ROIC ~12.5%, net leverage 1.9x); 2026 capex guide $4.3–4.7bn (BusinessWire, 2026-01-28).
• Herc Holdings (HRI) — gurufocus/8-K (P/E ~16.7x, EV/EBITDA ~16x deal-distorted, ROIC ~5%, net leverage ~4x).
• Ashtead Group plc (LSE: AHT) historical price — companiesmarketcap.com (annual returns +43% 2020, +70% 2021, −22% 2022, +14% 2023, −7% 2024, +2.5% 2025; ATH ~£64 Nov-2021; final LSE print £53.26, 2026-02-27).
• Ashtead Group annual-report / KPI & remuneration disclosures — historical LTIP metrics (adjusted EPS, RoI, leverage, TSR). https://www.ashtead-group.com/investors/investor-centre/kpis/
• Bloomberg / Reuters / Investing.com — “A 93,000% Return for the Rent-Everything Store” (Dec-2018); #1 FTSE 100 stock 2021 (Dec-2021); “Ashtead shares plunge after profit warning” (Dec-2024). https://www.investing.com/news/stock-market-news/ashtead-shares-plunge-after-profit-warning-amid-us-market-challenges-3763014
• Sunbelt Rentals IR — leadership (Pease/WestRock; Kyle Horgan). https://ir.sunbeltrentals.com/company-information/leadership-team

Frameworks applied

• Greenwald & Kahn, Competition Demystified — moat typed as economies of scale + local density + national-account captivity; share-stability and ROIC-vs-cost-of-capital tests.
• Marathon / Chancellor, Capital Returns — capital-cycle location (late-cycle, amber); asset-growth-anomaly read on the ROIC compression.