Carol Tomé, chief executive since 2020, spent eighteen months making United Parcel Service smaller on purpose, and the stock has been punished as though the shrinkage were an accident.
UPS runs a single integrated air-and-ground parcel network, reported in two segments — U.S. Domestic Package and International Package — with everything else, from ocean and air freight forwarding to contract logistics, healthcare distribution and the Roadie and Happy Returns platforms, disclosed as Supply Chain Solutions, a grouping of operating segments that individually fall below the reportable threshold .
In the June quarter, the company finished the hardest thing it has attempted in decades: cutting Amazon's volume by more than half from 2024 levels , roughly two million pieces a day, alongside some 78,000 operational positions, close to 150 buildings and about $4.5 billion of expense across 2025 and 2026 .
The second quarter, reported July 28, was the first in which the rebuilt network's economics were visible rather than promised. Consolidated revenue of $22.8 billion rose 7.6% and non-GAAP adjusted operating profit of $2.1 billion rose 12.0%, lifting the adjusted consolidated margin to 9.2% . The line that matters sits inside U.S. Domestic: adjusted operating margin of 8.0%, up 100 basis points year over year and 400 basis points from the first quarter , earned on volume that fell 3.3%. Reported GAAP diluted EPS was $0.71 against $1.76 adjusted, the $891 million after-tax gap being separation costs from the driver buyout program .
Shares last closed at $93.02, 24% below the $122.41 intraday high set in February . The argument of this piece is that the drawdown was delivered by fuel and by a volume decline the company announced eighteen months in advance, and that the price now embeds essentially none of the cost structure that came out alongside the volume.
Diesel, not demand, took the multiple apart
The de-rating happened in two distinct moves, and neither was an earnings event. The first came in March, when the Gulf conflict pushed crude sharply higher; UPS went from a $115.96 weekly close in late February to $94.80 four weeks later, an 18% slide . The stock then recovered almost all of it by mid-July, reaching $117.72 . The second move is the strange one. UPS beat and raised on July 28 — full-year revenue guidance to approximately $91.2 billion, adjusted operating profit to approximately $8.65 billion and adjusted diluted EPS to approximately $7.22 — and the shares fell from $114.79 to $104.22 in that week , then ground lower through September to the current level.
The mechanism is real but it is optical as much as economic. UPS's fuel expense jumped to $1,697 million in the quarter from $1,058 million in Q2 2025, and the surcharges that recover it reset weekly against Department of Energy kerosene and on-highway diesel benchmarks, by design offsetting cost rather than adding profit . The segment data supports management's claim that the mechanisms worked. But because surcharge revenue carries the cost it recovers, a fuel spike mechanically dilutes a reported percentage margin even when profit is unharmed — in International it cost 120 basis points in the quarter . Consensus earnings have moved a few percent; the price has moved twenty-four. What the fuel shock supplied was selling pressure; what the Amazon exit supplied was a story that made it easy not to buy the dip. Neither changed the model.
Strip the property sales and the spread is still there
The whole thesis reduces to one metric management has been explicit about: the gap between U.S. revenue per piece and cost per piece. Base pricing of 250 to 350 basis points plus a revenue-per-piece-over-cost-per-piece spread of 50 to 100 basis points is the stated algorithm for domestic margin accretion . In the second quarter revenue per piece rose 9.3%, driven by an average 5.9% net increase in base and accessorial rates implemented through 2025, higher fuel surcharges and favorable customer mix , while adjusted cost per piece rose 8.0% — a 130-basis-point spread, above the top of the frame.
The sharpest bear objection is that the quarter was flattered. Gains on sales of properties and aircraft parts, booked inside operating expenses and primarily within U.S. Domestic, were $101 million against $20 million a year earlier . On an adjusted U.S. cost base of about $13.7 billion ($14,930 million of revenue less $1,188 million of adjusted profit ), that $81 million swing is worth roughly 59 basis points of cost growth. Take it out and the spread is around 70 basis points. The objection is correct on the facts and wrong on the conclusion: 70 basis points still sits inside the range management committed to, and the range is what the earnings model needs.
There is a second, better objection, which is that the Amazon packages just removed were disproportionately expensive to serve, so part of the cost-per-piece improvement is a one-time mix artifact rather than a repeatable structural gain. That is almost certainly true of some of it. What argues against it carrying the whole explanation is that the cost reduction and the volume reduction were engineered together — positions, buildings and hours came out in step with the pieces — and that the automation share of U.S. volume kept climbing independently, to 68.5% at quarter-end from 64% a year earlier, at roughly 28% lower cost per piece .
The sensitivity is worth doing precisely, because it is the hinge. On 2027 U.S. revenue near $61.7 billion at roughly 91.6% costs, the domestic cost base is about $56.5 billion; each 100 basis points of spread is therefore some $565 million of operating profit, $435 million after tax at the guided 23% rate , or roughly $0.51 per share on the 851 million shares outstanding at June 30 . If the ~70-basis-point rate proves durable — the central dependency of the entire case, and one quarter is not proof of it — that is about $305 million of incremental after-tax profit a year against perhaps $2 billion of reset outlay, this year's transformation cash plus next year's growth capital, a return near 15% against a weighted cost of capital around 9%. The spread does not need to be heroic. It needs to exist.
One more piece of evidence is easy to miss. Of the 6.0% U.S. revenue increase, rate and product mix contributed 5.2 points and fuel surcharges 4.1, against a 3.3-point drag from volume . Ex-fuel, domestic revenue grew roughly 1.9% while the company was deliberately removing packages.
International is the drag; healthcare is the engine
International Package delivered the quarter's best top line and its worst margin. Revenue of $5,044 million rose 12.5% on an 18.9% increase in revenue per piece, but the adjusted margin fell 280 basis points to 12.4%, of which management attributes only 120 basis points to fuel . The unexplained remainder is geographic mix, and it is not trivial: the China-to-U.S. lane is highly profitable, and the growth has migrated elsewhere, with Asia-to-Asia export volume up 13.6% . Guidance is 13% to 14% for the third quarter and mid-teens for the year . This is the segment where the forward model is least proven.
Supply Chain Solutions is the opposite case — revenue of $2,860 million up 7.8% at a 10.2% adjusted margin, 220 basis points better year over year — and it contains the portfolio's clearest growth engine. Healthcare generated more than $3 billion of revenue for a second consecutive quarter, with 27 temperature-controlled cross-dock facilities added to the network, and UPS's own claim is that it is the only carrier operating end-to-end complex healthcare logistics on its own assets . The growth is not yet primarily organic: healthcare logistics added $232 million of revenue in the quarter, largely from the AHG cold-chain acquisition closed in the fourth quarter of 2025 . Against it, Mail Innovations shrank by $186 million in the quarter and $563 million year to date , and the goodwill carried by the freight-forwarding and healthcare-distribution reporting units — $877 million and $738 million of a $4.8 billion balance — showed only a limited excess of fair value over carrying value at the last annual test . A non-cash write-down there would not touch the cash thesis, but it would land in the segment the growth story depends on.
What breaks it: USPS fees, an unsettled pilot contract, and July 2028
The most underappreciated line in the quarter is purchased transportation, which rose 26% to $3,187 million from $2,522 million . Ground transportation expense alone increased $362 million, primarily fees paid to the Postal Service for outsourcing the final mile of Ground Saver . The history matters: UPS insourced the product from USPS at the start of 2025, found self-delivery expensive, and re-outsourced the final mile from 2026. The economics may well be better, but the company has made itself a price-taker to a counterparty that sets its own holiday rates, on a product aimed squarely at the volume Amazon can carry itself. Air carrier expense added a further $175 million from leased aircraft covering capacity lost to the fourth-quarter 2025 aircraft retirements and Middle East charter disruption , and two aircraft sat temporarily idled at quarter-end .
Labor is the structural cap. Nearly 80% of U.S. employees are union-represented , and the Teamsters national master agreement runs through July 31, 2028 . The pilots' agreement with the Independent Pilots Association became amendable on September 1, 2025 and remains in negotiation — a live, unpriced cost line. One overhang did clear: airline mechanics at Teamsters Local 2727 ratified an extension on June 26, 2026, pushing that contract's amendable date to November 1, 2029 . But any valuation that capitalizes margin gains past mid-2028 is paying today for the outcome of a negotiation in which the productivity UPS has just extracted from automation becomes the other side's opening argument.
Then there is the cash. Free cash flow is guided to approximately $5.5 billion this year, after about $3 billion of capital expenditure and a $1.3 billion pension contribution, against roughly $5.4 billion of planned dividends . No shares were repurchased in the first half, and $1.3 billion remains of the January 2023 authorization . The declared rate of $3.28 per share for the first half annualizes to $6.56, or about $5.6 billion on the current share count — the gap with the $5.4 billion of planned payments is timing, and coverage either way is essentially one times. Management's guidance record argues for discounting the optimistic reading: 2025 incentive targets were set at $89.0 billion of revenue and $9.6 billion of adjusted operating profit, and the company delivered $88.7 billion and $8.7 billion, for a 75% payout . Nor is the competitive backdrop benign. Amazon, 10.6% of consolidated revenue as recently as last year , delivered 6.7 billion parcels in 2025 against UPS's 4.4 billion — a customer that is now the larger domestic carrier and can bid for exactly the small-business and residential volume UPS is repricing.
12.9 times free cash flow, and a 22% discount to FedEx on identical guidance
EV/EBITDA is the wrong lens on a business that spends roughly $3 billion a year on aircraft, buildings and sort automation ; the relevant number is cash after that spend. Against roughly $6.15 billion of free cash flow over the next twelve months — a quarter of this year's guide blended with three quarters of my 2027 estimate of $6.4 billion, both on the company's own definition of operating cash flow less capital expenditure and both after pension and transformation cash — the $79.0 billion market capitalization works out to about 12.9 times. The declared dividend rate is a 7.1% yield on the current price.
The peer check is unflattering to the bears. UPS trades at 12.9 times its own approximately $7.22 calendar-2026 adjusted EPS guide ; FedEx, at $290.36 , trades at about 16.6 times the $17.50 midpoint of the $16.90-to-$18.10 calendar-2026 adjusted diluted EPS range from continuing operations it issued in June . Same calendar year, same adjusted construct, both company-set — UPS at a 22% discount, on a franchise with the better domestic margin trajectory of the two.
At a cost of equity near 10.3%, 12.9 times free cash flow implies a cash yield of about 7.8% against essentially no sustainable growth. That is the flat-spread outcome, priced as the base case. Putting 14.5 times on $6.15 billion — a 6.9% reported yield, which corresponds to roughly 7.3% once the non-recurring transformation and pension cash is grossed up, against 3% growth — gives about $105 a share, roughly 13% above the current price and near 20% with the dividend. The twelve-month sell-side consensus target sits near $119 . The downside is honest: if the spread closes entirely and International stalls at 12% to 13%, free cash flow of roughly $5.5 billion fails to cover the run-rate dividend, 11.5 times is the right multiple, and the stock is worth about $74, some 20% lower.
The spread is one quarter old, and the price says it is zero
The evidence base here is thinner than the bulls admit. One quarter of demonstrated spread, part of it flattered by property sales and part by lapping a first quarter that carried excess staffing from the Ground Saver transition, is not a track record. The next two prints resolve it: a third-quarter U.S. adjusted margin near the guided 7% — against a prior-year quarter that carried a $0.30 per-share sale-leaseback benefit — followed by the step to an approximately 8.8% second-half rate would convert a claim into a fact.
The asymmetry is what makes the position. A multiple of 12.9 times free cash flow does not price a 70-basis-point spread with error bars around it; it prices the spread at zero, on a network that has already given up roughly 78,000 positions and close to 150 buildings . What a buyer is genuinely not compensated for is two things: July 2028, when the Teamsters contract reopens and automation becomes a bargaining chip rather than a margin lever, and Amazon's ability to keep bidding for the premium small-business volume on which the repricing depends. Both are real ceilings on terminal value, and together they are why this is a cash-yield-plus-repair story rather than a compounder. Within that ceiling, at roughly 12.9 times free cash flow with a 7% dividend and two quarters that should settle the central question, United Parcel Service is mispriced to the upside.
Disclaimer: Tenzing Analytics, Inc. is not a registered investment adviser, broker-dealer, or financial institution, and all content provided by Tenzing Memo is AI-generated for informational purposes only. It does not constitute investment advice, a recommendation, or an offer or solicitation to buy, sell, or hold any security or financial instrument. Content is general in nature, not tailored to any individual's financial situation or objectives, and may contain errors, omissions, or inaccuracies; it should not be relied upon as a primary basis for any investment decision. Any forward-looking statements are subject to risks and uncertainties and may not reflect actual future results. Users should conduct their own research and consult qualified financial, legal, or tax advisors before making investment decisions, and by accessing this content, you acknowledge and accept these limitations.