United Rentals URI

Reference table

1. The data constraint, stated first because it shapes everything

All six documents in the United Rentals filing archive are prose-only. A check across the archive returns zero [TABLE] markers in each of the 10-K filed 2026-01-28, the 10-Q filed 2026-04-22, the 10-Q filed 2026-07-22, and the 8-Ks filed 2026-05-08, 2026-06-18 and 2026-07-22.

What that means in practice. The extracted text preserves the management's discussion and analysis, the notes narrative, the legal and covenant descriptions and the 8-K item text — all of which are used extensively below and are genuinely informative. What it does not preserve is any financial statement: no consolidated balance sheet, no statement of cash flows, no debt maturity schedule, no lease table, no segment table.

The consequence for this dive, stated as a rule and applied throughout. Every revenue, growth-rate, fleet-productivity and operational figure in this dive is filing-verified from the management discussion. Every balance-sheet, cash-flow, debt and capital-expenditure figure comes from the vendor payload and is explicitly labelled as unverified. Where the standing rule says the filing wins over the vendor, there is no filing to win with, and the correct response is to say so rather than to present vendor data as though it had been checked.

The checks that therefore could NOT be run on this name, all of which fired on other names in this batch: the lease-inflation test on totalDebt (which found $17.2bn on FedEx, $5.8bn on Digital Realty and $982M on Boston Scientific); the capital-expenditure verification (which found a $3.18 billion zero on Digital Realty and a $22 million error on Simon Property Group); and the enterprise-value net-debt reconciliation.

One thing the prose does confirm about the capital structure, from the 10-Q's guarantor disclosure: "As of June 30, 2026, the indebtedness, net of debt issuance costs, of our non-guarantors was comprised of (i) $1.414 billion of outstanding borrowings by the SPV in connection with the Company's accounts receivable securitization facility, (ii) $147 of outstanding borrowings under the ABL facility by non-guarantor subsidiaries and (iii) $15 of finance leases of our non-guarantor subsidiaries." That is a partial view — non-guarantor subsidiaries only — but it establishes that finance leases at those entities are $15 million, which is a very small number against the $1,124 million the payload records as capitalLeaseObligations for the consolidated group.

And the covenant position is disclosed in prose and is favourable: "Based on our current total indebtedness leverage ratio (as defined in the applicable debt agreements) and usage of the ABL facility as of June 30, 2026, we met the criteria under the applicable debt agreements for these exceptions, and as a result we were not restricted in our ability to make share repurchases and dividend payments."

2. The business, on the numbers the filings do provide

Filing-verified from the 10-Q filed 2026-07-22, management's discussion:

The company's own explanation, quoted because it is the whole argument: "Equipment rentals increased $434, or 12.7 percent, primarily due to a 7.1 percent increase in average OEC and a 3.4 percent increase in fleet productivity."

Fleet productivity is the metric to watch and the 10-Q explains why: "Fleet productivity is a comprehensive metric that provides greater insight into the decisions made by our managers in support of equipment rental growth and returns. Specifically, we seek to optimize the interplay of rental rates, time utilization and mix to drive rental revenue. Fleet productivity aggregates, in one metric, the impact of changes in rates, utilization and mix on owned equipment rental revenue."

Two readings.

First, the growth is accelerating and it is not just fleet. Revenue grew 7.1% in FY2024, 4.9% in FY2025, and then 9.6% in the first half of 2026 and 11.8% in the June quarter alone. And the acceleration is roughly two-thirds fleet size and one-third productivity — a 7.1% OEC increase plus a 3.4% productivity increase compounding to the 12.7% rental growth. A cyclical business reaccelerating from 4.9% to 11.8% is either early in a new upcycle or benefiting from a specific demand source, and the filings do not disclose which.

Second, fleet productivity improved sequentially within the half — +2.9% for six months and +3.4% for the quarter, implying roughly +2.4% in the March quarter. Rate and utilisation are getting better, not worse, while the fleet grows 7%. That is the Cournot claim in Section 5 tested and passed.

The base-period distortion, disclosed and quantified. From the 10-Q: "In January 2025, we announced that we had signed a merger agreement to acquire H&E Equipment Services, Inc. d/b/a H&E Rentals. In February 2025, following the termination of that merger agreement, we received a break-up fee of $64. Our results for the six months ended June 30, 2025 include a net $39 merger termination benefit... For the six months ended June 30, 2025, the impact of the merger termination was a $29 after-tax benefit, or $0.45 per diluted share, to net income and a $52 benefit to adjusted EBITDA, cash flow from operating activities and free cash flow."

So the first half of 2025 was flattered by $52 million of adjusted EBITDA and $0.45 per share — which means the 2026 comparisons in this file are struck against an inflated base, and the underlying improvement is slightly better than the reported one. It also records a corporate action the payload does not contain: United Rentals attempted to buy H&E Equipment Services in January 2025 and the deal was terminated a month later, with United Rentals collecting a $64 million break-up fee. This is an acquisitive company whose most recent large attempt failed, and that context matters for anyone modelling capital allocation.

One other corporate action, from the 8-K filed 2026-06-18: Amendment No. 18 to the Third Amended and Restated Receivables Purchase Agreement, extending the accounts-receivable securitisation facility to 2027-06-18, further extendable on a 364-day basis by mutual agreement. Advances under the facility are reflected as debt on the balance sheet and stood at $1,414 million at 2026-06-30.

3. Balance sheet and cash flow — vendor data, explicitly unverified

Every figure in this section is from the vendor payload. The URI filing archive contains no financial statements against which to check it. Read accordingly.

Three observations, each carrying the unverified caveat.

Capital intensity is the defining feature. Capital expenditure of $4,528 million is 28.1% of FY2025 revenue and capexToDepreciationTTM is 1.590 — this business reinvests roughly one and a half times its depreciation charge every year. That is not a defect signature; it is what an equipment rental company does, and a ratio below 1.0 would be the warning sign.

But the vendor's free-cash-flow figure is not the company's, and the difference is large. From the 10-Q: "We define 'free cash flow' as net cash provided by operating activities less payments for purchases of, and PLUS PROCEEDS FROM, equipment and intangible assets." United Rentals adds back the cash it receives from selling used fleet; the vendor's calculation does not. The payload's own seg_prod block shows "Rental Equipment" revenue of $1,413 million in FY2025 — that is the sale of used rental fleet. On the company's definition, free cash flow would therefore be materially higher than the vendor's $662 million — our derivation puts it in the order of $2.0–2.1 billion, a roughly 2.9% yield rather than 0.92%. We label that a derivation rather than a figure because the cash-flow statement that would confirm the proceeds line is not in the archive. Consequently freeCashFlowYieldTTM of 0.883%, priceToFreeCashFlowRatioTTM of 113.3x and evToFreeCashFlowTTM of 137.4x are rejected as understating by roughly a factor of three.

Capital return is substantial and confirmed by the prose covenant disclosure. Repurchases of $1,969 million in FY2025 — 2.75% of the current market capitalisation — plus $464 million of dividends. Weighted diluted shares have fallen from 72,929 thousand in FY2020 to 63,600 thousand in the June 2026 quarter, a 12.8% reduction. Total shareholder yield is approximately 3.4%.

4. Valuation — priced in or room?

At $1,150.13 (market cap $71.59B, 62.2M shares implied):

Estimate coverage is good through FY2027 and the FY2029 row is corrupt. 15 analysts on FY2026 EPS and 15 on FY2027 — the anchor — with an FY2027 range of $54.019 to $63.409, a 17.4% spread. 9 on FY2028. The FY2029 row reports epsAvg of ZERO, epsLow of ZERO, epsHigh of ZERO and netIncomeAvg of ZERO while carrying revenue estimates from 5 analysts and an ebitdaAvg of $9,488 million — an internally impossible row that is excluded outright.

est.ebitdaAvg and est.ebitAvg are fabricated and rejected — the twelfth consecutive file. In every one of the eight years present, ebitdaAvg is exactly 44.598% of revenueAvg and ebitAvg is exactly 26.298% — with the single exception of the FY2022 row, where the ratios break, which is itself evidence the series is generated rather than aggregated.

Enterprise value. enterpriseValueTTM of $86,860M less market capitalisation is $15,273M, which is close to but not equal to the payload's own FY2025 netDebt of $16,018M — consistent with the vendor using a more recent quarter. minorityInterest is 0 and the payload shows no noncontrolling interest, so the NCI-omission check does not apply. None of this can be checked against a filed balance sheet.

4a. What today's price assumes (the inversion)

At $1,150.13 — 23.4x FY2026 consensus and 20.4x FY2027 — the price embeds:

• Adjusted EPS of $49.12 in FY2026. (Consensus, 15 analysts.) The first half delivered $22.47, leaving $26.65 — a 18.6% sequential-half increase, which the September consensus of $13.86 is consistent with. Achievable, and it requires the second half to be seasonally stronger than the first, which it usually is.
• Adjusted EPS growth of 14.9% in FY2027 on revenue growth of 8.7%. (Consensus, 15 and 16 analysts.) That gap is operating leverage plus buyback, and the buyback at the FY2025 pace of $1,969 million retires about 2.75% of the equity a year.
• Fleet productivity stays positive while the fleet grows 7%. (Our derivation; the filing supplies both.) This is the most fragile assumption in the price. The entire re-rating from a historical low-teens multiple to 20.4x rests on the market believing that this industry's pricing discipline is structural rather than cyclical. Fleet productivity of +3.4% is one quarter's evidence for that proposition.
• The market keeps paying 19-21x forward earnings for a cyclical. (Our number.) At 15x FY2027E the stock is $847; at 24x it is $1,355. United Rentals has traded well below 15x for most of its history, and the current multiple is the highest sustained level it has held.
• No cycle turn inside the forecast horizon. (Our number.) Equipment rentals are 87% of revenue and capital expenditure is 28.1% of it. In a downturn, utilisation and rate fall together while the fleet — already bought — depreciates.

4b. The return bridge (why the multiple moves)

Expected return over the next twelve months decomposes as: adjusted EPS growth (+14.9%, from FY2026E $49.12 to FY2027E $56.44) + multiple drift (COMPRESSION, from 20.4x toward roughly 19.5x) + shareholder yield (+3.4% — a 0.65% dividend plus roughly 2.75% of buyback).

Our base assumes modest multiple compression and that is why the base sits below the price. The reason is position, not deterioration: the stock is at its fifty-two-week high having risen 47.6% in six months, on a beta of 1.794, at the highest sustained multiple this business has carried. Nothing in the operating data argues for compression — quite the opposite — but paying a peak multiple at a peak price for a cyclical is where the arithmetic stops supporting the story.

Note what makes this different from a bearish call: if the multiple simply HOLDS at 20.4x, the stock is $1,151 in a year plus a 3.4% yield, so the twelve-month total return is about 3.4%. The earnings growth is real and it is already in the price.

4c. Variant perception (where we differ, what would surprise)

• We are 10.8% below the street's mean and 15.3% below its median, and the difference is entirely the multiple. Consensus is $1,233 and the median is $1,300 — 21.8x and 23.0x FY2027E respectively. We use the same consensus rows and apply 19.5x. Note also that this is the only name in this batch with five sell ratings, and the low target of $903 is 21.5% below the price — so the coverage universe here is genuinely divided, which is more informative than the unanimity found on SLB, MSI and CI.
• We think fleet productivity is the number the market should be watching and almost nobody quotes. +3.4% in the quarter, +2.9% for the half, improving sequentially, alongside 7.1% fleet growth. That combination — more machines AND better economics per machine — is the single fact that would justify a permanently higher multiple for this industry, and it is one quarter old. Watchable number: fleet productivity in the 2026-10-28 release against +3.4%.
• We think the prose-only archive is a genuine and under-appreciated analytical risk on this specific name, not a housekeeping note. This is a company with $16.5 billion of reported debt, $4.5 billion of annual capital expenditure and a $1,124 million line the vendor calls capitalLeaseObligations against $15 million of finance leases disclosed at the non-guarantor subsidiaries. On five other names in this batch, exactly this kind of check found errors between $22 million and $17.2 billion. Here it could not be run.
• Positive surprise that would force a re-rate: fleet productivity above 4%; or United Rentals beginning to disclose end-market exposure, particularly to data-centre and infrastructure construction, which would convert an inference into a valuation input; or FY2027 consensus revised above $58.
• Negative surprise that would break the thesis: fleet productivity turning negative, which would say the industry is adding capacity into the rate cycle; or capital expenditure guidance rising materially above $4.5 billion; or a return to the pattern of the September and December 2025 quarters, both of which MISSED consensus (11.70 against 12.29 and 11.09 against 11.79).

Synthos fair values

All three anchors are multiples of the FY2027 consensus adjusted EPS distribution (mean $56.443, low $54.019, high $63.409, 15 analysts).

• Bear ~$720 — 13.3x the FY2027 consensus LOW of $54.019, and 14.7x FY2026E. Cross-checks: 1.4% above the 52-week low of $710.47; 20.3% below the street's low target of $903. The scenario: non-residential construction turns, fleet productivity goes negative as the industry works off the capacity it has just added, utilisation and rate fall together, and a cyclical at 20x reverts to the low-teens multiple it carried for a decade. MINUS 37.4%.
• Base ~$1,101 — 19.5x the FY2027 consensus MEAN of $56.443. Cross-checks: 22.4x FY2026E; 17.0x FY2028E; a 3.4% total shareholder yield at that price. Sensitivity, stated openly: 15x gives $847 — 26.4% below spot — and 24x gives $1,355. The estimate has a 17.4% spread and the multiple assumption a 60% one, so the multiple is doing almost all the work. The scenario: consensus is met, fleet productivity stays modestly positive, the buyback continues at 2.75% a year, and the multiple gives back a fraction of its re-rating. MINUS 4.3%.
• Bull ~$1,500 — 23.7x the FY2027 consensus HIGH of $63.409. Cross-checks: 30.5x FY2026E; 30.4% above the 52-week high; 21.7% above the street's $1,233 and 5.6% above the street high of $1,421. The scenario: fleet productivity accelerates above 4%, the current construction cycle proves to be a multi-year infrastructure and data-centre build rather than an ordinary upturn, the industry holds pricing discipline through it, and the market re-rates equipment rental as a structurally advantaged oligopoly rather than a cyclical — which is precisely the knowledge base's argument. +30.4%.

Base is 4.3% BELOW spot; asymmetry roughly 0.81:1 (37.4% down, 30.4% up), before a shareholder yield of approximately 3.4%. A base below the price with a payoff ratio under one, on a 1.79-beta cyclical closing at its fifty-two-week high, is a Hold — and the operating momentum is why it is a Hold rather than anything worse.

5. Knowledge base — no claim on the company, two strong claims on the industry

Raw hits: 2. Case-sensitive entity matches: ZERO. Name-level claims on United Rentals: ZERO. Industry-level claims: 2, both bullish, both from one channel on one day.

The sweep ran entity terms URI and United Rentals plus free text on united rentals, equipment rental and aerial work platform across all 52,021 distilled claims.

Both hits, verbatim:

> 2025-04-29 · BULLISH · conviction 82 · horizon: principle · skill 1.0 · no entities · channel: invest_like_the_best · categories: cruise lines, car rental, airlines, travel, equipment rental

> "Cournot oligopolies with fixed short-term supply (cruises, rental cars, air travel) sustain monopoly-like profits since players effectively compete on quantity."

> 2025-04-29 · BULLISH · conviction 80 · horizon: thesis · skill 1.0 · no entities · channel: invest_like_the_best · categories: construction equipment rental

> "Equipment rental is a Cournot oligopoly with highly inelastic demand and constrained supply—an early high-conviction winner for him."

Three disciplines applied. Neither names United Rentals, so neither enters the conviction score under the entity test. Both are from the same channel on the same date, so the concentration test fails and a breadth reading of 2 would misrepresent them as independent — they are one voice on one day. And both carry speaker: null, the weakest sourcing the four-lane policy admits.

But the second claim is unusually specific and directly testable, and the filings test it. It asserts that construction equipment rental earns above-normal returns because supply is constrained and demand is inelastic — in a Cournot structure, competitors respond to demand by choosing how much capacity to add rather than by cutting price, so incremental demand flows to rate rather than to margin compression.

The June quarter is a clean test and it passes. United Rentals increased average original equipment cost 7.1% — it added capacity — and simultaneously increased fleet productivity 3.4%, which aggregates rate, time utilisation and mix. In a competitive market, a 7% supply increase would show up as flat or negative fleet productivity. It did not. And the improvement is accelerating: +2.9% across six months against +3.4% in the second quarter alone.

We record that as a corroborated claim that enters no score, for the same reason the pharmacy-benefit lane on Cigna and the oil-services lane on SLB entered none: the entity discipline is what stops this programme from attributing sector opinions to individual companies, and it has to apply when the sector view is favourable as well as when it is not.

What this means. Breadth 0 on the entity test, claim count 0, net conviction empty at the name level. The store has no view on United Rentals. It has a single-source, two-claim, high-conviction view on the industry structure that United Rentals dominates, and the filings support the mechanism. That is worth exactly as much as one voice on one day about a category, and it is one of the reasons the exponential score is 4 rather than 3.

6. Data integrity — what could and could not be checked

Four findings, and the first one is the absence of the ability to make findings.

1. The entire filing archive is PROSE-ONLY — zero preserved statement tables across all six documents. The 10-K filed 2026-01-28, the 10-Qs filed 2026-04-22 and 2026-07-22, and the 8-Ks filed 2026-05-08, 2026-06-18 and 2026-07-22 each return a [TABLE] marker count of zero. No consolidated balance sheet, no statement of cash flows, no debt schedule, no lease table, no segment table is available. Three standing checks could not be run: the lease-inflation test on totalDebt, the capital-expenditure verification, and the enterprise-value net-debt reconciliation. On five other names in this batch those same checks found errors ranging from $22 million to $17.2 billion. Every balance-sheet and cash-flow figure in this dive is therefore vendor data carrying an explicit unverified label, and the standing rule that the filing wins cannot be applied because there is no filed statement to apply it with.

2. capitalLeaseObligations of $1,124M sits against $15M of finance leases disclosed in the prose — unresolvable. The payload records $1,124 million of capitalLeaseObligations inside totalDebt of $16,477 million at 2025-12-31. The only lease figure available in the archive is from the 10-Q's guarantor note: "$15 of finance leases of our non-guarantor subsidiaries" at 2026-06-30. That covers only the non-guarantor entities and is not comparable to a consolidated figure, so it neither confirms nor refutes the $1,124 million. What can be said is that the pattern on five other names in this batch — where a vendor capitalLeaseObligations line turned out to be the operating lease liability relabelled — is a live possibility here and cannot be tested. We flag it as unresolved rather than assuming either way.

3. freeCashFlow of $662M is not the company's measure and understates it by roughly a factor of three. The 10-Q states: "We define 'free cash flow' as net cash provided by operating activities less payments for purchases of, and plus proceeds from, equipment and intangible assets." United Rentals adds back the proceeds of used-fleet sales; the vendor's calculation subtracts gross capital expenditure and adds nothing back. The payload's own seg_prod block shows $1,413 million of "Rental Equipment" revenue in FY2025, which is the sale of used fleet. Our derivation puts free cash flow on the company's definition in the order of $2.0-2.1 billion — a roughly 2.9% yield rather than 0.92% — and it is labelled a derivation because the cash-flow statement is not in the archive. freeCashFlowYieldTTM of 0.883%, priceToFreeCashFlowRatioTTM of 113.3x and evToFreeCashFlowTTM of 137.4x are rejected.

4. The est FY2029 row is internally impossible and is excluded. It reports epsAvg, epsLow, epsHigh and netIncomeAvg all equal to ZERO, while simultaneously reporting revenue of $21,275 million from 5 analysts, an ebitdaAvg of $9,488 million, an ebitAvg of $5,595 million and 3 analysts on EPS. A row with three EPS analysts and an EPS of zero is not a forecast. Separately, ebitdaAvg is exactly 44.598% of revenueAvg and ebitAvg exactly 26.298% in every year except FY2022, where the fixed ratio breaks — which is itself evidence the series is generated rather than aggregated. All est EBITDA and EBIT rows are rejected and forward valuation uses epsAvg for FY2026 to FY2028 only.

VERIFIED CLEAN — recorded because a clean check is a finding, and there are only two available.

seg_prod reconciles exactly. FY2025: owned equipment rentals $11,048M + ancillary and other rental revenue $2,483M + re-rent revenue $275M + rental equipment sales $1,413M + new equipment $348M + contractor supplies $163M + service and other $369M = $16,099 million, which is total revenue to the dollar, and the identity holds for FY2024, FY2023, FY2022 and FY2021. The block does not separate General Rentals from Specialty after FY2016, which is a real limitation — Specialty is historically the higher-growth, higher-margin half — but what is there is correct.

seg_geo reconciles exactly. FY2025: United States $14,671M + Non-US $1,428M = $16,099 million, again to the dollar, with the same identity in every year back to FY2012. The only defect is a label inconsistency: FY2024 and earlier call the international line "Canada and Europe" while FY2025 calls it "Non-US". In a batch where seg_geo failed on five of twelve names, United Rentals' is among the cleanest.

Non-equity tripwire — checked and passed. URI is common stock, NYSE-listed. Price of $1,150.13 is not par-like; beta is 1.794; volume was 0.59M shares (~$675M of turnover); the 52-week band of $710.47 to $1,150.13 is a 62% range. This is common equity.

Vendor composite rating — noted, not used. B- / 2 overall, with 1 of 5 on discounted cash flow, debt-to-equity and price-to-book. The discounted-cash-flow sub-score rests on the free-cash-flow figure rejected in finding 3 and is understated by roughly a factor of three.

7. Technicals and insiders

• Price $1,150.13, which IS the 52-week high. tech.hi52 is $1,150.13 and pct_from_hi is 0.00%. max_dd_from_peak is likewise 0.00%, and the two coinciding at zero is correct: the six-year peak is today's close. This is the only name in this batch to finish the session at its own high.
• +61.9% above the 52-week low of $710.47. Position within the annual range: 100th percentile.
• Well above both moving averages: +7.6% above a 50-day average of $1,069.13; +27.1% above a 200-day average of $904.61.
• RSI 63.1 — firm but, notably, not overbought despite the price being at its high, which happens when a stock grinds up rather than spikes. MACD +12.69, the strongest absolute reading in this batch.
• Relative performance is the best in this batch: 3-month +24.3% against SPY +7.6%; 6-month +47.6% against SPY +11.1%; 12-month +33.8% against SPY +24.3% and QQQ +30.8%.
• Beta 1.794 — the second-highest in this batch. A 10% market decline implies roughly an 18% move here, and the entry is at the top of a 62% annual range.
• Sentiment: 29 buy, 7 hold and 5 sell across 41 analysts — the most sell ratings of any name in this batch. Consensus target $1,233 (+7.2%), median $1,300 (+13.0%), high $1,421 (+23.6%), low $903 (MINUS 21.5%). The mean sits 5.2% below the median, which means the low targets are pulling the average down — and unlike the unanimity found on several other names here, this coverage universe genuinely disagrees.

Insiders — one sale that carries information, and seven grants that do not.

The chief financial officer sold 1,500 shares at $1,133.15 on 2026-07-24 — two days after the second-quarter release — retaining 6,061.653. That is a 19.8% reduction of his disclosed holding, executed within 1.5% of today's price, immediately after the largest earnings beat in the file. It is a small absolute amount (~$1.7 million) and it is the only discretionary transaction in the file. The remaining seven rows are A-Award director grants: 203 shares each to six directors at $937 on 2026-05-08, and 34 shares to one director at $1,132.89 on 2026-06-30 — the standard retainer, carrying no signal.

There is not one open-market purchase, and no transaction by the chief executive. Note that the directors' 2026-05-08 grants were priced at $937 and the stock is now $1,150.13 — 22.7% higher in under three months.

8. Verdict, kill-criteria and flip conditions

Hold.

The business is doing better than the verdict suggests and the price is why.

What is genuinely good. Revenue reaccelerating from 4.9% in FY2025 to 11.8% in the June quarter. Equipment rentals up 12.7% on a 7.1% increase in average original equipment cost and a 3.4% increase in fleet productivity — a company adding 7% more machines while getting 3.4% more economics out of each one. Fleet productivity improving sequentially, from roughly +2.4% in the March quarter to +3.4% in June. An adjusted EPS beat of 10.7%, the largest in the file. A buyback of $1,969 million in FY2025 — 2.75% of the market capitalisation — that has taken the share count from 72.9 million to 63.6 million since FY2020. And a knowledge-base industry claim, at conviction 80, whose central prediction the filings confirm.

What holds it at Hold, in three specifics.

First, the price and the position. The stock closed exactly at its fifty-two-week high of $1,150.13, up 47.6% in six months, at 20.4 times next year's consensus, on a beta of 1.794. Our base of $1,101 is 4.3% below spot with 0.81:1 asymmetry — the least favourable payoff ratio in this batch. If the multiple merely holds at 20.4x, the twelve-month total return is about 3.4%.

Second, this is a cyclical being valued as though it were not. Equipment rentals are 87% of revenue, capital expenditure is 28.1% of revenue, and the 20.4x forward multiple is the highest sustained level this business has carried. The entire re-rating rests on the proposition that the industry's pricing discipline is structural — the Cournot argument — and the evidence for that is one quarter of positive fleet productivity. Note also that the September and December 2025 quarters both MISSED consensus, at $11.70 against $12.29 and $11.09 against $11.79, which is what the beginning of a cyclical wobble looks like and which the last two quarters have reversed.

Third, and this is a constraint on us rather than on the company: we cannot verify the balance sheet. All six documents in the filing archive are prose-only. On a company reporting $16.5 billion of debt, $4.5 billion of annual capital expenditure and a $1,124 million lease line that cannot be reconciled to the $15 million of finance leases the prose discloses, that is a real gap — and in this same batch the equivalent checks found errors of $22 million, $982 million, $5.8 billion, $12.7 billion and $17.2 billion on other names.

Why not Watch. Because the base fair value is within the Hold band at MINUS 4.3%, and because the resolving event is a price rather than a disclosure — we would want this business meaningfully cheaper, not differently informed.

Why not Avoid. Because nothing is deteriorating. Revenue is accelerating, fleet productivity is improving, the beat was the largest in the file, and the balance sheet — as far as the prose permits us to say — is unrestricted for buybacks and dividends.

Pre-registered UPGRADE conditions — what would take this to Buy — Tactical:

• A price below roughly $900, which is 15.9x FY2027E, restores a payoff ratio above 2:1 on the base of $1,101, and is close to the street's low target of $903.
• Fleet productivity above 4% in the September or December quarter, from +3.4% — which would make the structural-pricing argument much harder to dismiss.
• End-market disclosure, particularly any breakout of data-centre, semiconductor or infrastructure construction exposure, which would convert an inference into a valuation input.
• FY2027 consensus adjusted EPS revised above $58, from $56.44.
• Insider buying — there is not one open-market purchase in the file, and the chief financial officer sold at $1,133.15.

Pre-registered KILL criteria — what would take this to Watch or Avoid:

• Fleet productivity turning negative, which would say the industry is adding capacity into the rate cycle and that the Cournot structure has broken.
• Two consecutive earnings misses, repeating the September and December 2025 pattern.
• Capital expenditure guidance rising materially above $4.5 billion without a corresponding step in rental revenue.
• Average original equipment cost growth exceeding 10% while fleet productivity falls below 1% — the specific combination that signals over-building.
• The buyback falling below $1.5 billion a year, from $1,969 million.
• A break below the 200-day moving average of $904.61 on company-specific news.

Where URI fits in the Synthos Framework Portfolio. The industrials sleeve, no position today, with a 1.5% starter contingent on a price below $900 or fleet productivity above 4%. Sizing note, and it is the operative discipline: beta is 1.794 and the entry would be at the 100th percentile of the annual range, so any position taken here should be sized on the drawdown rather than on the conviction. Logged as a tracked Synthos call (Hold) as of 2026-08-04 at $1,150.13, with the fair-value anchors, kill criteria and upgrade conditions all gradeable.

Single biggest risk: cyclicality bought at the top of the range. Equipment rentals are 87% of revenue and are sold into construction and industrial activity; capital expenditure is 28.1% of revenue and capexToDepreciationTTM is 1.590, so the fleet must be continuously replaced whatever the demand environment; beta is 1.794; and the stock closed the session exactly at its fifty-two-week high after a 47.6% six-month move. In a downturn, the two components of fleet productivity — rate and time utilisation — fall together, while the fleet that was bought at the top keeps depreciating. The offsetting consideration is the one the knowledge base makes and the filings currently support: if equipment rental really is a Cournot oligopoly with constrained supply, the industry responds to weaker demand by not adding fleet rather than by cutting price, and the downturn is shallower than it used to be. That proposition is the entire difference between a 13x multiple and a 20x one, it has never been tested in a genuine construction recession at this level of industry concentration, and at $1,150.13 the market is pricing it as settled. We are not willing to pay for it at the high, and $900 is the number at which we would be.

Provenance & disclosures

• Traceability: ZERO knowledge-base claims name United Rentals out of 52,021 distilled claims (raw hits 2, case-sensitive entity matches 0, name-level claims 0; breadth 0, claim count 0, net conviction empty at the name level). Both text-only hits come from one channel on one day — invest_like_the_best, 2025-04-29, both speaker: null, both skill 1.0 — and both are bullish on equipment rental as an industry structure: conviction 82, horizon principle, "Cournot oligopolies with fixed short-term supply (cruises, rental cars, air travel) sustain monopoly-like profits since players effectively compete on quantity"; and conviction 80, horizon thesis, categories construction equipment rental, "Equipment rental is a Cournot oligopoly with highly inelastic demand and constrained supply—an early high-conviction winner for him." Neither names United Rentals, so neither enters the conviction score under the entity discipline — which is applied here to a FAVOURABLE sector lane exactly as it was applied to the unfavourable pharmacy-benefit lane on Cigna and the favourable oil-services lane on SLB in this same batch. The second claim is specific and testable, and the filings support it: United Rentals grew equipment rental revenue 12.7% in the June quarter on a 7.1% increase in average original equipment cost and a 3.4% increase in fleet productivity — a company that adds 7% of capacity while improving the economics of each unit is not having its pricing competed away, which is what the Cournot claim predicts. Recorded as corroborated, scored at zero, and discounted because it is one voice on one day about a category rather than a company. All quotes are verbatim from the stored claim text.
• THE ARCHIVE CONSTRAINT — stated here because it governs the whole document: all six filings in the URI archive are PROSE-ONLY, returning a [TABLE] marker count of ZERO — the 10-K filed 2026-01-28, the 10-Qs filed 2026-04-22 and 2026-07-22, and the 8-Ks filed 2026-05-08, 2026-06-18 and 2026-07-22. No consolidated balance sheet, statement of cash flows, debt schedule, lease table or segment table is available to this dive. Every revenue, growth-rate, fleet-productivity, covenant and corporate-action figure in this dive is filing-verified from the preserved management discussion and note narrative. Every balance-sheet, debt, cash-flow and capital-expenditure figure comes from the vendor payload and is explicitly labelled UNVERIFIED. The standing rule that the filing wins over the vendor cannot be applied here because there is no filed statement to apply it with, and three standing checks — the lease-inflation test on totalDebt, the capital-expenditure verification and the enterprise-value net-debt reconciliation — could NOT be run. On five other names in this batch those same checks found errors between $22 million and $17.2 billion.
• Data as-of: operational and narrative data through 2026-06-30, filing-verified from the management discussion in the 10-Q filed 2026-07-22 · corporate actions from the 8-K filed 2026-06-18 · vendor balance-sheet and cash-flow data as at 2025-12-31, unverified · estimates 2026-08-04 · prices 2026-08-04, quote timestamp 1785873602 = 2026-08-04T20:00:02Z ($1,150.13, +3.07%; 50-DMA $1,069.13; 200-DMA $904.61; RSI 63.1; MACD +12.69; beta 1.794; pct_from_hi 0.00% — the close IS the 52-week high) · knowledge-base claims 2026-08-04. United Rentals' fiscal year is the calendar year. All figures come from the Synthos vendor data file for URI or from the SEC filings in the URI archive; no figure comes from memory, recall or external retrieval.
• What the prose DID establish, and it is substantial: second-quarter total revenue of $4,410 million, +11.8%, with equipment rentals of $3,849 million, +12.7%, "primarily due to a 7.1 percent increase in average OEC and a 3.4 percent increase in fleet productivity"; six-month revenue of $8,395 million, +9.6%, with rentals +10.8% on +6.4% OEC and +2.9% fleet productivity; equipment rentals at 87% of revenue and rentals plus sales of rental equipment at 95%; the definition of fleet productivity as the aggregate of "changes in rates, utilization and mix on owned equipment rental revenue"; the terminated H&E Equipment Services merger — signed January 2025, terminated February 2025, $64 million break-up fee, a net $39 million benefit in the first half of 2025 comprising $12M of professional fees in SG&A, $13M of bridge financing fees in interest expense and the $64M fee in other income, worth $29 million after tax, $0.45 per diluted share, and $52 million to adjusted EBITDA, operating cash flow and free cash flow — a distortion in the prior-year comparison base that the payload does not contain; Amendment No. 18 to the Third Amended and Restated Receivables Purchase Agreement, entered 2026-06-18, extending the accounts-receivable securitisation facility to 2027-06-18; the composition of non-guarantor indebtedness at 2026-06-30 ($1,414 million of securitisation borrowings, $147 million of asset-based facility borrowings and $15 million of finance leases); the company's own definition of free cash flow as "net cash provided by operating activities less payments for purchases of, and plus proceeds from, equipment and intangible assets"; and the covenant statement that the company "met the criteria under the applicable debt agreements for these exceptions, and as a result we were not restricted in our ability to make share repurchases and dividend payments."
• Where the payload was rejected or qualified (detailed in Section 6): freeCashFlow of $662M, which uses gross capital expenditure and omits the used-fleet sale proceeds that United Rentals' own definition adds back — the payload's seg_prod block shows $1,413 million of "Rental Equipment" sales revenue in FY2025, and our derivation puts free cash flow on the company's basis in the order of $2.0-2.1 billion, roughly three times the reported figure, so freeCashFlowYieldTTM (0.883%), priceToFreeCashFlowRatioTTM (113.3x) and evToFreeCashFlowTTM (137.4x) are all rejected; capitalLeaseObligations of $1,124M, which sits inside totalDebt of $16,477M and cannot be reconciled to anything in the archive — the only lease figure available is $15 million of finance leases at the non-guarantor subsidiaries, which is not comparable, so the possibility that this line is an operating lease liability relabelled (as it proved to be on five other names in this batch) is flagged as unresolved rather than assumed; and the est FY2029 row, which reports epsAvg, epsLow, epsHigh and netIncomeAvg all equal to ZERO while carrying revenue estimates from 5 analysts, 3 analysts on EPS, an ebitdaAvg of $9,488M and an ebitAvg of $5,595M — internally impossible and excluded, alongside the whole est EBITDA/EBIT series, fabricated at exactly 44.598% and 26.298% of revenue in every year except FY2022 where the fixed ratio breaks. Where the payload verified against itself or the prose — recorded, because a clean check is a finding, and only two were available: seg_prod reconciles exactly in every year from FY2021 to FY2025, with the seven FY2025 lines summing to $16,099 million, which is total revenue to the dollar — though it stops separating General Rentals from Specialty after FY2016, which is a real limitation given Specialty is historically the higher-margin half; and seg_geo reconciles exactly in every year back to FY2012, with FY2025 United States $14,671M plus Non-US $1,428M equalling $16,099 million, the only defect being a label change from "Canada and Europe" to "Non-US" between FY2024 and FY2025.
• Basis note: analyst estimates are on adjusted diluted EPS, which the 8-K of 2026-07-22 defines as "diluted EPS plus the impact of the defined special items." earn_cal's epsActual series is the adjusted figure — $12.76 for Q2 2026 against GAAP diluted EPS of $11.84 in inc_q, a $0.92 adjustment. Trailing adjusted EPS of $45.26 (25.4x) is built from the four quarterly releases; trailing GAAP EPS of $41.61 (27.7x) is reported alongside. Note the recent pattern: the September and December 2025 quarters both MISSED consensus ($11.70 against $12.29 and $11.09 against $11.79) before the March and June 2026 quarters beat by 8.5% and 10.7%.
• Estimate coverage: 15 analysts on FY2026 adjusted EPS and 15 on FY2027 — the anchor for all three fair values, with an FY2027 range of $54.019 to $63.409, a 17.4% spread — and 9 on FY2028, used as a cross-check. The FY2029 row is excluded as internally impossible.
• Peer note: the vendor peer set — Axon Enterprise, Cummins, Canadian National Railway, FedEx, Ferguson, W.W. Grainger, L3Harris, Norfolk Southern, PACCAR, Roper Technologies — contains NOT ONE equipment rental company. Ashtead Group (Sunbelt Rentals) and Herc Holdings, the only two meaningful comparables, are both absent, as is H&E Equipment Services — the company United Rentals attempted to acquire in January 2025. The set is assembled from industrials and distributors with no shared business model. No peer-multiple comparison is drawn.
• Fair-value caveat: the $720 / $1,101 / $1,500 anchors are multiples of the FY2027 consensus adjusted EPS distribution — 13.3x the low of $54.019, 19.5x the mean of $56.443, and 23.7x the high of $63.409. Stated arithmetic, not a discounted cash flow, and not cross-checked against enterprise value or free cash flow because neither can be verified from a filed statement. The base is sensitivity-disclosed: 15x gives $847, which is 26.4% below spot, and 24x gives $1,355. The base assumes modest multiple compression from 20.4x toward 19.5x — a judgement about position rather than about deterioration, since the operating data argues the other way — and if the multiple simply holds, the twelve-month total return is about 3.4%.
• Timing and entry: the stock closed 2026-08-04 exactly at its fifty-two-week high of $1,150.13, with pct_from_hi and max_dd_from_peak both at 0.00% — the only name in this batch to do so, and the correct behaviour of those fields when the six-year peak is the current close. Second-quarter results were released 2026-07-22 and beat by 10.7%; the chief financial officer sold 1,500 shares at $1,133.15 two days later. The next print is 2026-10-28, 85 days away.
• Accessibility note: no information in this dive is conveyed by colour. All emphasis is carried by bold text, table structure and explicit labelling.
• Not investment advice. Independent research, educational and informational only, never personalised. No recommendation to buy, sell or hold any security is made to any person.
• Version: 2026-08-04-full.