Fleet Mix, Damage, and the Winter Gap

Key takeaways

• United Rentals ended 2025 with 1,095,000 units and $22.48 billion of original equipment cost, about $20,500 of purchase cost per unit, and earned 61 cents of rental revenue per dollar of that fleet cost ($13.806bn of equipment rental revenue against $22.48bn of OEC, both from its FY2025 10-K).
• Depreciation is a forecast you write: United Rentals depreciates to a weighted average salvage value of 12% of cost, and says a one-year change in assumed useful life would move annual depreciation by roughly $311m to $410m, or 12% to 15% of its $2.670bn rental-equipment depreciation charge.
• Damage is a budget line. United Rentals booked $50m of insurance proceeds from damaged equipment in 2025 (0.36% of equipment rental revenue) but carried $265m of self-insurance accruals. The deductible layer, not the insurer, is where an operator’s losses land.
• Avis Budget flexed its average quarterly fleet from about 631,000 vehicles in Q1 2025 to about 746,000 in Q3 (+18%) while average quarterly utilization stayed in a 68%–72% band: the fleet moves so the utilization does not, and that flex is a financing capability before it is an operating one.

A fleet is a portfolio, not a pile

Three choices decide what a rental fleet earns, and none of them is the daily rate. Which categories you stock, how many identical units you hold inside each category, and how many distinct models you are willing to maintain. The first sets your revenue ceiling. The second decides whether a breakdown costs you a day or a customer. The third writes your repair bill.

United Rentals is a useful reference because it publishes the mix. Its FY2025 10-K reports a fleet of 1,095,000 units carrying $22.48 billion of original equipment cost, roughly $20,500 of purchase cost per unit, and splits equipment rental revenue by fleet type: general construction and industrial equipment 39%, aerial work platforms 22%, power and HVAC 11%, general tools and light equipment 9%, fluid solutions 7%, trench safety 5%, surface protection mats 4%, and mobile storage and modular office space 3%. No category carries more than four-tenths of the revenue. The four specialty categories together carry 27%, and they carry it at a better margin: the specialty segment produced a 43.6% equipment rentals gross margin in 2025 against 35.2% for general rentals, on a 38.0% blended average.

Illustrative only. Take two yards, each holding $1,000,000 of original fleet cost, each producing the 61 cents of rental revenue per OEC dollar that United Rentals managed in 2025, so call it $600,000 a year. Yard A holds 50 units across 30 models: 1.7 units per model. Yard B holds 50 units across 6 models: 8.3 units per model. Now a machine goes down. At Yard B the customer collects an identical unit off the row and never learns there was a problem. At Yard A the down unit is the entire model line, so the booking cancels. If 4% of units are out of service on an average day, and Yard A converts nearly all of that into lost bookings while Yard B substitutes away three-quarters of it, the gap is about three points of revenue: $18,000 a year, every year, on the same million dollars of steel.

The levers are depth per model, deliberate spread across categories that do not fail together, and a cross-hire relationship with a neighbouring yard for the days you are short. The risk is the mirror image of the benefit: six models chosen badly can all serve the same customer type, and then depth is just concentration with better parts availability. Test the mix by asking which single customer segment, if it stopped calling for a quarter, would idle the largest share of your OEC.

Depreciation is a forecast; the resale desk is the scoreboard

Depreciation is not a cost you pay. It is a guess you record about how fast an asset loses value, and it runs on two inputs, useful life and salvage value, both of which you choose.

United Rentals states its choices. The weighted average salvage value of its rental equipment is 12 percent of cost. It also publishes the sensitivity: if the useful lives of all its rental equipment increased or decreased by one year, annual depreciation expense would fall by roughly $311 million or rise by roughly $410 million. Set against the $2.670 billion of rental-equipment depreciation it actually recorded in 2025, a single year of assumption moves the charge by 12% to 15%. Nothing physical happened to the machines.

The market audits the guess at disposal. In 2025 United Rentals took $1.413 billion of proceeds on used rental equipment against $778 million of cost, the net book value of what it sold, for a $635 million gain, about 1.8 times book. Its average fleet age was 49.5 months at year-end, down from 51.3 a year earlier.

Illustrative only. A $40,000 machine on a seven-year life to a 12% salvage floor depreciates $5,029 a year. At four years, near that 49.5-month average, accumulated depreciation is $20,114 and book value is $19,886. Sold at 1.8 times book it returns about $35,800: just under 90% of what it cost, after four years of earning. That is what a conservative schedule looks like when the used market cooperates.

It does not always cooperate, and United Rentals says so in its own risk factors: the market value of any given piece of rental equipment could be less than its depreciated value at the time it is sold. A 1.8x ratio is a portfolio average across a mixed fleet in one particular used-equipment market. One machine in a soft market sells below book and the gain line becomes a loss line. The practical lever is to set life and salvage to match how you actually sell, then check the assumption every year against realised prices. If disposals keep printing gains, you are depreciating faster than the market and understating current profit; if they print losses, you have been distributing money you had not earned.

Where the rental dollar actually goes

Operators talk about utilization because it is the number that moves. The cost structure underneath is what decides whether moving it is worth anything, and it is remarkably stable across the industry.

United Rentals discloses the full stack against its $13.806 billion of 2025 equipment rental revenue: depreciation of rental equipment $2.670bn (19.3%), labor and benefits $2.161bn (15.7%), repairs and maintenance $1.087bn (7.9%), delivery $987m (7.2%), and all other rental expenses $1.653bn (12.0%), leaving equipment rentals gross profit of $5.248bn, a 38.0% gross margin, before any selling, general and administrative cost, interest, or tax.

Two of those lines deserve more attention than they get. Repairs and maintenance at 7.9% is the true cost of keeping steel working, and it is separate from the depreciation that supposedly accounts for wear. Delivery at 7.2% is the line that quietly scales with how far your customers are, and it grows fastest where the equipment is heaviest. In United Rentals’ specialty segment, delivery cost rose from $350 million in 2024 to $473 million in 2025, up about 35%, well ahead of that segment’s revenue growth.

Illustrative only. A yard with $600,000 of rental revenue, applying those same 2025 ratios, would carry roughly $116,000 of depreciation, $94,000 of labor, $47,000 of repairs, $43,000 of delivery, and $72,000 of everything else, leaving about $228,000 of gross profit. Out of that $228,000 come the office, the software, the yard lease if it is not already in the other line, the sales cost of winning the bookings, and the interest on whatever financed the fleet. That is the honest gap between gross margin and money in the bank.

The levers are pricing delivery separately rather than absorbing it, routing so a truck drops two units per trip, and preventive maintenance scheduled in the off-season when the unit was not going to earn anyway. The risk is treating any of these ratios as a target rather than a description: a yard that cuts repairs to 5% of revenue for two years has not found efficiency, it has moved the cost into next year’s disposal price.

Damage, deposits, and who is actually carrying the risk

Damage in a rental business is not an accident. It is a rate: a predictable percentage of transactions that must be priced, reserved, and financed like any other cost.

United Rentals reports insurance proceeds from damaged equipment as a distinct line: $50 million in 2025, $51 million in 2024, $38 million in 2023. Against 2025 equipment rental revenue that is 0.36%. But that line records only what an insurer paid. Underneath it sits the retention, the layer the operator eats before any policy responds. United Rentals carried self-insurance accruals of $137 million in current liabilities and $128 million in long-term liabilities at 31 December 2025, $265 million in total, up from $247 million a year earlier. Avis Budget separately carried $284 million of current public liability and property damage insurance liabilities at the same date, against $245 million a year earlier. Those reserves are the real shape of the exposure; the recovery line is the tip.

Illustrative only. An operator does 800 rentals a year at an average $400 ticket, or $320,000 of revenue. If 1.5% of rentals produce damage averaging $900 to repair, that is 12 incidents and $10,800 a year, about 3.4% of revenue. A $500 deposit covers 56% of the average incident and under 1% of a $52,000 machine. And a single total loss of that machine is roughly five years of the ordinary damage line, gone in one afternoon.

So the three instruments do three different jobs and only one of them is risk transfer. The deposit is behavioural: it makes the renter careful and gives you leverage in a conversation, not coverage. The damage waiver is a margin product. Avis Budget sells collision and loss damage waivers as an ancillary revenue line, and notes in its 10-K that certain US states mandate disclosure to the customer and that some have statutes establishing or capping the daily rate chargeable for a loss damage waiver, so it is a regulated product with a price ceiling in some places. The insurance policy is the only actual transfer, and it starts above your retention.

The failure mode is designing for the average and being killed by the tail. Set the deposit high enough to change behaviour and low enough to close bookings; price the waiver against your own observed loss rate rather than a competitor’s posted price; and size your retention against the single worst unit in the fleet, not the average claim. Then check annually whether your loss rate is drifting, because a rate that moves from 1.5% to 2.5% is a 67% increase in a cost line nobody was watching.

Winter is a financing problem, not a demand problem

Every rental operator knows the season. Fewer plan for the fact that the assets keep costing money in the months nobody is calling, and that the correct response is a balance-sheet move rather than a marketing one.

Avis Budget shows the mature version. Its average quarterly vehicle rental fleet in 2025 ran from about 631,000 vehicles in the first quarter to about 746,000 in the third, an 18% swing, while average quarterly fleet utilization stayed inside a 68% to 72% band. That is the whole discipline in one sentence: the fleet flexes so the utilization does not. Shrink the denominator in the slow quarter and your idle-asset problem largely disappears.

That flex is bought with financing, not willpower. Avis Budget’s 2025 vehicle programs cash flows show $25.973 billion of proceeds from borrowings against $24.714 billion of payments, a fleet debt book that turns over many times its own balance in a year, funding $15.056 billion of vehicle purchases against $10.406 billion of disposal proceeds. The seasonal fleet is a revolving credit machine with wheels attached.

Equipment rental cannot flex as cleanly, because there is no manufacturer repurchase programme for a scissor lift. United Rentals simply states that its business is seasonal, with demand tending to be lower in the winter months, and that borrowings under its asset-backed revolver are used to fund seasonal expenditures. Its release valve is the resale channel: $1.413 billion of used rental equipment sold in 2025 against $4.149 billion of purchases, for $2.736 billion of net rental capital expenditure. Buy in spring, sell in autumn, and let the net number carry the season.

Illustrative only. A party-rental operator books $480,000 a year with 62% of it arriving May through August, so $297,600 in four months, leaving $182,400 across the other eight, or $22,800 a month. Fixed costs (storage, insurance, base payroll, debt service) run $26,000 a month, so those eight months cost $208,000 and burn $25,600 net. Add the spring restocking spend and the cash reserve needed before the first tent goes up is well past $50,000. The operator who financed the fleet at 100% and kept no reserve does not fail in January because demand fell; they fail because a payment was due in a month with no revenue behind it.

The defences are ordinary. Size the reserve from the actual off-season gap, not a rule of thumb. Push fleet purchases as late into pre-season as suppliers allow. Time disposals into the strong used market rather than the panic one. And negotiate the debt schedule against the revenue shape: seasonal amortisation exists, and a lender who understands the industry will structure to it if asked before signing rather than after missing.

Four numbers to run the fleet on

Most rental dashboards report bookings. Bookings tell you about the last month. These four tell you about the asset base, and every one of them is computable from a spreadsheet you already have.

One: rental revenue per dollar of original fleet cost. United Rentals produced 61 cents in 2025 ($13.806bn of equipment rental revenue against $22.48bn of fleet OEC). This is the single number that says whether your capital is working, and unlike time-based utilization it cannot be flattered by cheap add-on units sitting on the yard.

Two: fleet age. United Rentals ran 49.5 months at the end of 2025, down from 51.3. Age is a leading indicator for both the repair line and the disposal price, and it moves slowly enough that a drift takes two years to hurt and two years to fix.

Three: recovery on disposal against book value. United Rentals realised about 1.8 times net book value in 2025. If yours is consistently above 1.0 your depreciation schedule is conservative and your reported profit is understated; if it is below, your schedule is optimistic and you have been paying yourself out of a valuation that was never there.

Four: damage recoveries and retained losses as shares of revenue. United Rentals’ recovery line was 0.36% of equipment rental revenue in 2025, against $265 million of self-insurance accruals. Track both, because the ratio between them tells you how much of your loss exposure you are actually keeping.

One caution on all four: these are figures from a company with $16.099 billion of total 2025 revenue, national purchasing power, and its own resale infrastructure. They are a reference frame for the shape of the cost structure, not a benchmark a two-truck yard should expect to hit. Use them to ask why your number differs, since the answer is usually mix, geography, or delivery distance, rather than as a target to hit.

Put it to work

Rebuild your fleet ledger this week. List every unit with purchase cost, book value, assumed life, and salvage. Compute rental revenue per dollar of fleet cost, then repairs, delivery, and labor as shares of that revenue. Price a total loss of your most expensive unit against a full year of damage claims. Finally, size the off-season cash gap before you buy anything in spring.

Sources & references

Linked entries open the named source directly. Entries without a link say exactly what kind of reference they are — and how to check them yourself.

• United Rentals, Inc. — Form 10-K for fiscal year 2025 (fleet OEC, fleet mix, rental cost breakdown, salvage and useful-life disclosures, used equipment sales, self-insurance accruals)
• Avis Budget Group, Inc. — Form 10-K for fiscal year 2025 (quarterly fleet size and utilization range, loss damage waiver disclosures, vehicle programs financing)
• American Rental Association — The trade association for equipment and event rental operators in North America; a starting point for industry-level education and operator resources. Figures in this briefing come from the SEC filings cited above, not from ARA.