Strip the $330M Comparison and Q3 Grew 14%: Domestic Held Its Guide, International Beat It, and Supply Chain Became the New Miss

Results vs. Consensus

Q3 2026 Scorecard

The two consensus compilations we checked agree closely. The widely cited pair was $1.63 on adjusted EPS and $22.29B on revenue; a second compilation carried $1.62 and $22.27B, on which the beat reads +4.3% and +2.6%. The table uses the first pair, and the characterization does not change either way. One note on the revenue line: the comparison above uses the filed total of $22,853M rather than the $22.9B rounded figure the release headline and the wires carried, because the rounded version would overstate the beat by roughly 20bp.

The more important fact about the consensus this quarter is what did not happen. Three months ago the GAAP line was the row that moved the stock: a $0.71 print against a published GAAP consensus near $1.64, because the Street had not modelled the $891M after-tax Driver Choice charge. This quarter the GAAP-to-adjusted wedge narrowed to $0.32, composed of $0.15 of transformation cost and $0.21 of goodwill impairment less $0.04 of asset-sale gains. There was no collision, and the GAAP print of $1.37 sat close enough to the adjusted figure that the headline readers and the model readers were looking at the same quarter, as they had in March, when the wedge was $0.05.

Performance Against the Company's Own Q3 Guidance

UPS gave no consolidated third-quarter revenue or EPS guide in July, only three segment ranges plus a half-year margin commitment. That makes the guide-versus-actual scorecard the more informative of the two, because it is the standard management set for itself ninety days ago.

Three quarters into the year the pattern is becoming the story. In each of the last two quarters exactly one segment has missed its own guided margin range, and it has been a different segment each time: International in June, Supply Chain Solutions in September. Both times every revenue line came in above its guide. A company that beats its revenue guides and misses one margin range per quarter is telling you something about where its forecasting confidence actually sits, and it is not in the top line.

Year-Over-Year Comparison

UPS reports cash flow only on a year-to-date basis in the quarterly schedules, so the two cash lines above are nine-month figures against nine-month comparatives while every other line is the quarter.

Quarter-Over-Quarter Comparison

The sequential table is where the shape of the guided step-down is visible, and where the one genuinely uncomfortable number in the quarter sits. The June quarter was the inflection; September was the first period guided to give part of it back.

Revenue. Consolidated revenue of $22,853M grew 6.7%, with all three segments contributing for a second consecutive quarter. The composition continues the pattern June established: price running well ahead of a volume base that is still shrinking, but shrinking more slowly in the U.S. and faster abroad. Consolidated average daily volume fell 4.2% and consolidated revenue per piece rose 7.7% to $15.96. The sequential line is the one worth pausing on. Total revenue was flat quarter over quarter at $22,853M against $22,834M, on one extra operating day, with U.S. Domestic and International both down slightly and the whole of the sequential increase coming from Supply Chain Solutions. For a company whose core narrative is premium growth, a flat sequential top line in the quarter before peak is a reminder that the volume base has not turned.

Margins. The consolidated margin went backwards and the business did not. Adjusted operating margin of 9.0% was 100bp lower than a year ago, and the bridge is three lines: U.S. Domestic up $129M, International up $55M, Supply Chain Solutions down $262M, summing to the reported $78M decline exactly. Supply Chain Solutions is a tenth of consolidated profit in a normal quarter and it accounted for more than the whole of the decline, because the prior-year quarter carried a $330M gain that was never excluded from adjusted results. On a comparable basis the consolidated margin expanded, from 8.41% to 8.99%, which is 58bp of expansion rather than 100bp of contraction.

EPS. The headline and the substance point in opposite directions and both are arithmetically correct. Adjusted diluted EPS of $1.69 fell 2.9%. Excluding the $0.30 the company itself footnotes in the prior-year quarter, it grew 17.4%. In July the CFO told the market to expect "mid-teens EPS growth in both the third quarter and the fourth quarter" once the sale-leaseback was normalized out. He got 17.4%, above the range he set, and the headline reported a decline. That gap between the reported number and the committed number is the single most important thing to understand about this print, and it is why a stock quoted 2.3% lower before the open closed 5.8% higher, a swing of more than eight percentage points.

Segment Performance

Revenue by Product Line

Next Day Air revenue growing 6.3% while its average daily volume fell 7.0% is the premium-mix claim in a single line, and the mix is now being delivered entirely by price. Revenue per piece on the company's most expensive domestic product rose 12.7% to $29.47. The qualifier matters: management disclosed that air average daily volume from small-business and enterprise customers was flat year over year. So the premium air story in the September quarter was repricing a shrinking book, not winning a growing one, and the only domestic product line whose volume held up was Ground at a 2.7% decline.

Volume and Pricing KPIs

The bottom rows carry the thesis, as they did in June, and this time they carry it in two directions at once. Year over year the spread is positive for a second consecutive quarter at 60bp, inside the 50 to 100 basis point algorithm management committed to, and contribution per piece rose 15.5% from $0.84 to $0.97. Applying the new rate to the quarter's 1,015.4 million pieces against the old rate on 1,033.6 million produces $116.7M of additional contribution dollars, against the $129M of actual segment adjusted profit growth, a residual of $12.3M that is cargo and other activity plus rounding in the one-cent cost-per-piece disclosure. The arithmetic ties for the second quarter running.

Sequentially the same metric is the least comfortable number in the print. Contribution per piece fell from $1.15 to $0.97, a 15.7% decline. There is genuine seasonality in that: the same metric fell 7.7% from the June to the September quarter of 2025. But this year's give-back was twice as large, and the spread narrowed 70bp from June's 130bp to land at the bottom of the committed band rather than the middle of it. Management's framing is that 130bp was above the algorithm and 60bp is the algorithm. That is internally consistent and it is also the more conservative of the two readings of the June quarter, which is the one we carried into this print and the one the evidence now supports.

U.S. Domestic Package

Revenue of $14,865M grew 4.5% and adjusted operating profit of $1,034M grew 14.3%, on 3.3% fewer pieces and 4.5% more revenue. The 7.0% adjusted margin landed exactly on the figure management guided in July, up 60bp year over year and down 100bp sequentially. Average daily volume fell 3.3%, the same rate as June and a long way from the mid-single-digit decline management had guided, which means the non-Amazon base held better than the company's own plan assumed.

The customer-mix disclosure improved again and, importantly, in absolute terms. Small-business average daily volume grew 3.7%, so the record 35.1% small-business share of U.S. volume, up approximately 240bp year over year and 60bp from June's 34.5%, is being driven by a growing numerator. Combined small-business and enterprise average daily volume grew 1.8%, which management sized as more than 200,000 additional pieces per day, with Ground up 2.1% and air flat. Global Digital Access Program revenue reached $1.4B, with $1.3B of that in the U.S., and air services plus ground commercial crossed nearly 30% of total program volume for the first time. Healthcare revenue exceeded $3B for a third consecutive quarter with growth in every segment.

"In the third quarter, the growth in revenue per piece was 60 basis points higher than the growth in cost per piece, continuing the positive spread between the two and reflecting our focus on revenue quality and productivity delivered by our network reconfiguration efforts." — Brian Dykes, CFO

Total segment expense rose 3.9% on an adjusted basis, with the increase coming primarily from fuel and purchased transportation. The Driver Choice Program finished inside the quarter, with approximately 1,500 participants exiting, mostly at the end of September, so the full benefit of those departures lands in the fourth quarter rather than this one.

Assessment: The segment did exactly what it said it would do, which is a different and in some ways more valuable thing than beating. A business whose domestic margin fell 300bp year over year in March printed the guided number to the basis point, and did it while beating the revenue and volume halves of the same guide. The caution is the one we flagged in July and it has not resolved: the segment has now proved twice that it can hit a number, and it still has not proved that the number compounds. The full-year arithmetic, laid out under Guidance below, has domestic operating profit dollars finishing 2026 flat against 2025.

International Package

International was the quarter's clean win and the reversal of June's blemish. Revenue of $5,007M grew 7.1% on an 8.7% volume decline, a 1,580bp price-and-mix offset driven by a 15.0% increase in revenue per piece with around half of that from fuel. Adjusted operating profit of $746M grew 8.0% and the 14.9% adjusted margin came in above the 13% to 14% range management guided, expanding 10bp year over year. That is the segment's first year-over-year operating profit growth in six quarters.

The operational detail underneath is better than the margin and it is where the forward case sits. The year-over-year rate of volume decline improved in each month of the quarter, with September finishing well above July, and U.S. inbound volume, which carries the most profitable trade lanes, returned to year-over-year growth in September. Asia-to-Asia export volume grew 8.4%, a third consecutive quarter of growth.

"When the de minimis exemption was eliminated in Europe, the volume had a sharp decline. But every month following, it has gotten better. We actually are seeing growth in the most profitable trade lanes, which is China-U.S., Asia to Asia." — Carol Tomé, CEO

On the forward path, management upgraded its own language. The full-year framing in July was "mid-teens." On this call the CEO offered "mid-teens, mid to high teens" and the CFO confirmed the wider phrase, which takes the segment's forward margin range above mid-teens.

Assessment: This was the pillar moving the wrong way in July and it moved back, decisively enough that the 90bp beat against its own range is not a rounding item. The honest caveat is that 8.7% fewer pieces produced 8.0% more profit, which is price and geographic mix doing the work rather than a recovering franchise, and a 15.0% revenue-per-piece increase of which half is fuel will not repeat when crude normalizes. The fourth-quarter guide of approximately 20.0% is where the credibility of the recovery gets tested, because it asks for 200bp of year-over-year expansion in a segment whose volume is still falling at a high single-digit rate.

Supply Chain Solutions

The smallest segment produced the quarter's only missed range and its only write-down. Revenue of $2,981M grew 18.2%, the fastest growth of any segment in any quarter this year, driven by forwarding up 17.5% on higher international air and ocean freight rates and logistics up 17.2% on healthcare. Adjusted operating profit of $274M fell 48.9% and the 9.2% adjusted margin landed below the guided 10.0% to 11.0%. Separately, a $182M goodwill impairment was recorded against an unnamed reporting unit inside the segment, which took the GAAP margin to 3.1%.

Two things need separating here, because the headline mixes them. The 48.9% profit decline and the 1,210bp margin compression are almost entirely the $330M prior-year gain: excluding it, the comparison is $274M against $206M, which is 33% growth. The genuine problem is narrower and more specific. Adjusted operating expense in the segment rose 36.3% against revenue up 18.2%, which is the opposite of operating leverage, and the reason is the composition of the growth. Forwarding revenue arrives with purchased transportation attached, and consolidated purchased transportation rose 33.0% to $3,275M.

"Logistics revenue increased 17.2% year-over-year, driven by strong growth in healthcare logistics, partially offset by a decline in our Mail Innovations business." — Brian Dykes, CFO

The impairment received one sentence in the prepared remarks, no sentence at all in the slide deck's segment page, and no question in Q&A. The reporting unit was not named. What is disclosable from the balance sheet is that goodwill fell $247M over the nine months against a $182M charge, so currency accounts for the remainder, and that the segment houses at least one business management described on the call as declining.

Assessment: This is the first goodwill impairment in any period the company's current reconciliation tables cover, and it deserves more weight than the market gave it. A goodwill impairment is management telling the auditor that a business it owns will not earn what the acquisition price assumed, and doing so in the same quarter that the segment missed its margin guide and grew revenue 18.2% makes the mix question concrete rather than theoretical. The fourth-quarter guide compounds the discomfort: approximately 12.2% would be the highest clean quarterly adjusted margin this segment has produced in the seven-quarter series the company publishes, whose best clean reading is 10.3%, and it is asked for one quarter after a 9.2% print and an impairment, with no bridge offered. We are modelling this segment below its guide and it is the single largest reason our full-year estimate sits under the company's.

Operating Expense Detail

Two of these eight lines cannot be read at face value, and the company does not give the disclosure that would fix either. Compensation and benefits fell 2.6%, which looks like the cost program arriving in the largest expense line. Both periods carry separation cost inside it, however, $171M this year against $328M last year, so if the whole of each charge sat in compensation then roughly half of the $311M decline would be the smaller charge rather than lower underlying labour cost. UPS does not disclose how transformation cost splits across expense lines, so the underlying figure is not recoverable from the filed table.

Other expenses up 27.1% is the second unreadable line and the more misleading one. The prior-year figure of $1,682M is the lowest of the last seven quarters, every other one of which falls between $1,958M and $2,509M, which is consistent with the $330M sale-leaseback credit sitting in that line. If it does, and if the $182M impairment sits there too, then adding back one and removing the other turns a 27.1% increase into a low-single-digit decline. We flag the inference rather than relying on it, but a reader taking the 27.1% at face value will conclude the cost base is deteriorating when the likelier reading is that it is flat.

The line that carries real information and no distortion is purchased transportation at $3,275M, up 33.0% and up $812M, which is more than half the total expense increase. It carries three things at once: the United States Postal Service tender for last-mile Ground Saver running at scale, leased aircraft capacity, and the freight bought to serve a forwarding business whose revenue grew 17.5%. The company does not break the line into those three parts, and until it does, the single largest growing cost at UPS is also its least explained. Depreciation grew 4.2% against 4.7% in June and 8.0% in March, which is the clearest sign yet that the automation capital wave is flattening on the income statement.

Key Topics & Management Commentary

Overall Management Tone: Confident and, for the first time in this cycle, forward-leaning rather than defensive. Where July reported a finished plan, this call spent most of its oxygen on 2027 and beyond, offering an addressable-market framework, a multi-year margin commitment and a new operating structure, none of which was requested by the print. The two soft spots were both omissions rather than evasions: the segment write-down was disclosed and then left alone, and the customer whose departure has defined three years of this story was not sized at all, for the first time in the three quarters we have covered. Analyst pushback was notably thin, with no question on either.

1. The 3.7% Decline Is a Comparison, Not a Quarter

Every negative number in the consolidated headline traces to one footnote. The September 2025 quarter included a $330M pre-tax gain inside Supply Chain Solutions from a sale-leaseback, worth $0.30 of earnings per share, and UPS did not exclude it from adjusted results. Confirm it from the reconciliation: prior-year GAAP operating profit of $1,804M, one adjustment of $328M for transformation cost, adjusted operating profit of $2,132M. The gain sits in both bases, which means it inflates the denominator of every growth rate in this quarter's adjusted comparison.

Strip it and the quarter reads: adjusted operating profit $2,054M against $1,802M, up 14.0%; adjusted earnings per share $1.69 against $1.44, up 17.4%; consolidated adjusted margin 8.99% against 8.41%, up 58bp. The company footnoted the gain in the release and the slide deck. It did not translate the footnote into a comparable growth rate anywhere in either document.

"Our third quarter results included gains from asset sales, partially offset by related expenses. On a net basis, these items were not a significant driver of the outperformance versus our expectations." — Brian Dykes, CFO

That statement is about this quarter's $43M of MD-11 parts gains, which are excluded from adjusted results, and it is accurate. It is not about last year's $330M, which is not excluded, and nobody on the call made the distinction out loud.

Assessment: This is the quarter's largest gap between what the tape read and what the business did, and it is the mirror image of June. In June the Street's estimate base had not absorbed a charge the company had disclosed, and the stock fell 6.6% on an operationally good quarter. In September the headline growth rates had not absorbed a prior-year gain the company had disclosed, and the stock was quoted 2.3% lower before the open and traded as low as 0.7% below the prior close before closing 5.8% higher. Both episodes were resolved by reading a footnote. The practical lesson for anyone modelling UPS through the transformation is that the adjusted series has been contaminated in both directions, and the only clean cross-period measures are segment contribution per piece and the expense detail.

2. Domestic Delivered the Guide Exactly, and the Spread Returned to the Algorithm

Ninety days ago the question was whether an 8.0% domestic margin could step down to approximately 7.0% as guided without the step-down becoming a slide. It printed 7.0%, with segment profit up 14.3% and revenue up 4.5% against a guide of approximately flat. The spread between revenue-per-piece growth and cost-per-piece growth came in at 60bp, inside the 50 to 100 basis point algorithm, with revenue per piece up 6.1% against adjusted cost per piece up 5.5%.

What the spread history now shows is a business whose unit economics turned and then settled rather than compounded: negative 300bp in March, positive 130bp in June, positive 60bp in September. Management's account of that path is that June was above the algorithm and September is the algorithm.

"This kind of range of 50-100 basis points spread between RPP and CPP will help us continue to drive improvement in the margin as we go forward." — Brian Dykes, CFO

The deceleration in the revenue-per-piece line is also now visible and management decomposed it. The 6.1% increase was about half base rates and customer mix and about half fuel, against a July commitment that the second half would decelerate to 4.0% to 4.5%. The line ran hotter than guided, and it ran hotter for the reason that does not help profit.

Assessment: A 60bp spread is a pass and a narrow one. It confirms the structural reset we underwrote in July and it also confirms the conservative reading of June that we chose at the same time, which is that 7.0% rather than 8.0% is the honest domestic run-rate at this point in the year. The forward question is no longer whether the reset is real. It is whether a 50 to 100 basis point algorithm applied to a volume base that is still shrinking produces enough margin accretion to matter, and on the full-year arithmetic the answer for 2026 is no. For 2027 it depends on volume turning, which management now says will not show in the reported number until the second half.

3. International Returned to Profit Growth and Got a Better Multi-Year Range

The segment that missed its range in June beat it in September, at 14.9% against 13% to 14%, and delivered its first year-over-year operating profit growth in six quarters. The mechanism was price and geographic mix rather than recovering volume: revenue per piece up 15.0% on volume down 8.7%, with around half the price gain from fuel surcharges that management says offset the fuel expense and preserved profit.

The more durable disclosure was the monthly volume shape. The company published monthly year-over-year average daily volume charts by region, which show the sharp July decline when the European de minimis exemption was eliminated and monthly improvement after it, with September finishing well above July. U.S. inbound volume returned to growth in September and Asia-to-Asia export volume grew 8.4% for a third consecutive quarter of growth.

"Mid to high teens is the right range for the international margin as the trade lanes start to settle down a bit, and we are hoping that that's actually going to happen." — Carol Tomé, CEO

That phrase is an upgrade. The July framing was "mid-teens" for the full year and nothing beyond it. "Mid to high teens" as a settled range takes the forward margin statement for this segment above mid-teens, and it was volunteered rather than extracted.

Assessment: Our July note said this was the guide we expected to be trimmed rather than met, and it has been met at its middle rather than its floor. On the fourth-quarter guide the full year computes to 15.0%, against the 14.0% we modelled and the 17.8% fourth quarter we calculated as the minimum requirement. We were wrong on this segment by about a point of full-year margin, which is roughly $200M of operating profit, and the correction accounts for most of the increase in our estimate. The remaining caution is that nothing in the recovery is yet volume: a segment earning more on 8.7% fewer pieces is earning it from surcharge and mix, and the approximately 20.0% fourth-quarter guide leans on peak demand surcharges that management described in exactly those terms.

4. Supply Chain Solutions Missed Its Range and Impaired Goodwill

A $182M goodwill impairment against an unnamed reporting unit, in the same quarter the segment missed its margin guide, is the one genuinely new negative fact in this print. Its disclosure footprint was one clause in the prepared remarks, one line in the reconciliation tables, and nothing else. No analyst asked about it.

The accounting mechanics matter less than what they imply. An impairment is a statement that the future cash flows of an acquired business no longer support the price paid for it. The segment contains at least one business management named as declining on this call, and the charge took segment GAAP margin to 3.1% against an adjusted 9.2%. Over the nine months goodwill on the balance sheet fell $247M, of which the $182M is the charge and the balance is currency.

The forward guide makes the silence harder to accept. Fourth-quarter Supply Chain Solutions margin is guided to approximately 12.2% on revenue up nearly 13%, which would be the highest clean quarterly adjusted margin the segment has recorded in the seven-quarter series the company publishes. The clean series runs 3.6%, 8.0%, 10.3%, 8.1%, 10.2% and 9.2%, with the one higher reading being the 21.3% that contained the sale-leaseback gain. No bridge from 9.2% to 12.2% was offered in the release, the deck or the call.

Assessment: Treat this as the segment that has replaced International as the risk half of the story, and treat the absence of questioning as a market inefficiency rather than as evidence the item is immaterial. The write-down is 6.1 percentage points of a quarter's segment margin and it signals a portfolio problem rather than a trading problem. Our estimate assumes approximately 11.0% in the fourth quarter rather than 12.2%, which would still be the best clean quarter in the series. The event that would change this read is a named reporting unit and a stated basis for the fourth-quarter step-up, neither of which has been given.

5. The Raise Is $1.4 Billion of Revenue for $100 Million of Profit, Again

UPS raised the full-year revenue outlook to approximately $92.6B from approximately $91.2B and adjusted operating profit to approximately $8.75B from approximately $8.65B. Adjusted earnings per share went to approximately $7.28 from approximately $7.22.

The incremental margin on the raise is $100M on $1,400M, or 7.1%, against the 2.6% that the July raise carried. Better, and still well below the 9.45% the business earns on the revenue it already had, which is the arithmetic signature of a top-line increase made of fuel surcharge and forwarding pass-through. The implied full-year adjusted margin fell about 3bp, so for the second consecutive quarter the company raised revenue and profit while the implied margin went slightly backwards.

Assessment: Two consecutive quarters of raising both lines is a real pattern and the commitment weight keeps improving, with free cash flow and the pension contribution moving from spoken remarks into the filed release this quarter. But the composition of the raises has not changed: revenue goes up by more than a billion dollars and profit goes up by tens of millions. A reader should treat the revenue guide as close to mechanical and the operating profit guide as the only one that carries information, and should note that at $8.75B the company is guiding full-year adjusted operating profit to grow 1.0% against 2025's $8.661B.

6. The Free Cash Flow Raise Is $450 Million of Pension Bonds

The headline change in the quarter's capital story is that free cash flow guidance went to approximately $6.1B from approximately $5.5B, which takes the dividend from 98% of guided free cash flow to 89%. Nine-month free cash flow of $4,220M was up 53.8% and covered 104% of the $4,064M of dividends paid, against 68% coverage a year ago. On the headline figures this is the first material loosening of the capital-allocation constraint since we initiated coverage.

The composition is where the care is required, and to its credit management volunteered it without being pressed.

"Remember, a part of that is the business performance, which the business is performing better than our initial expectations. Part of it's also the pension contribution in the form of bonds rather than cash, which is good for our pension plan." — Brian Dykes, CFO

UPS issued $1.3B of debt in the quarter to refinance a maturity and to contribute $450M of its own bonds to the pension asset portfolio. A pension contribution funded with paper rather than cash does not pass through the cash flow statement, so $450M of the $600M increase in the free cash flow guide is a change in funding method rather than in cash generation. Excluding it, the full-year figure is roughly $5.65B and the payout ratio is 96% rather than 89%. The CFO was also explicit that the item may reverse.

"Depending on how we fund the pension, we may then have to offset the $450 million in there." — Brian Dykes, CFO

Assessment: The direction is right and the magnitude is three-quarters smaller than the headline. Against the underlying $5.65B the payout improves from 98% to 96%, which is progress rather than release, and the 2027 starting point may be $450M worse than the 2026 exit depending on how the pension is funded. Set against that, the pension and postretirement obligation fell $816M over the nine months and the balance sheet absorbed the contribution without touching cash, which is a sensible piece of liability management. What the company has not done is convert any of it into shareholder optionality, which is the next section.

7. Capital Allocation: A Third Quarter of Zero Buyback and $1.95 Billion More Debt

UPS repurchased no shares in the first nine months of 2026, against $1,000M in the same period of 2025. The word was not used in the release, in the slide deck or on the call, for a third consecutive quarter. Nine-month dividends of $4,064M were up 0.5% on an unchanged $1.64 quarterly declared rate.

The balance sheet moved in the other direction. Total debt and finance leases rose $1,951M over the nine months to $26,078M, while total shareowners' equity fell $1,229M to $15,026M and retained earnings fell $1,553M. Debt to equity moved from 1.48x at the year end to 1.74x. Cash of $6,837M was up $950M with no commercial paper outstanding. Interest expense in the quarter was $289M against $291M, the first year-over-year decline of the three quarters of 2026, as the comparison laps the heavy 2025 issuance.

Asked directly about the dividend, the CEO gave the strongest statement she has made on it.

"So clearly … our dividend is safe and secure." — Carol Tomé, CEO

Assessment: We take the dividend statement at face value, because the nine-month coverage figure supports it and the fourth-quarter requirement is undemanding. What has not changed is the absence of optionality. The business has now gone three quarters without buying a share while a direct competitor repurchased roughly $2.0B of its own stock over June through September, and leverage increased in a year in which free cash flow improved 54%. That combination says the improvement is being used to repair the balance sheet rather than to return capital, which is defensible and which also means the yield remains the only return of capital on offer. A repurchase authorization remains the single cleanest signal management could give that the lower capital intensity is permanent.

8. The Cost Program Is at 60%, and Three Disclosures Got Quieter

Approximately $1.8B of program benefits were booked in the first nine months against a full-year target the release now describes as "nearly $3 billion." That is 60% at the three-quarter mark, against 40% at the halfway mark and 20% at the quarter mark, so the back-loading management described in January is tracking. The Driver Choice Program finished in the quarter with approximately 1,500 participants exiting, mostly in late September, and the CFO said all the buildings have now exited.

Three things in the filed documents moved, none of them flagged as a change. The full-year excluded transformation cost went from $1.3B to $1.5B in the July release to $1.4B to $1.6B in October, a $100M increase at both ends, with the Driver Choice component moving from $1.1B back to $1.2B. The program-benefit wording went from "approximately $3 billion" to "nearly $3 billion." And the completion language went from "these initiatives are expected to conclude by 2027" to the Efficiency Reimagined initiative "not expected to conclude until the fourth quarter of 2027."

A fourth item is a basis mismatch rather than a change. The filed release sizes the Driver Choice element of the excluded operating expense at $1.2B. On the call the CFO sized this year's Driver Choice spend at $1.4B. Those are a profit-and-loss exclusion and a cash figure respectively, and the company does not bridge them.

"That was always going to be weighted to the last quarter of the target, and we feel comfortable. Additionally, the driver buyout, the roughly 1,600 people finished exiting at the end of the third quarter, so you get the full wrap of that. Then obviously all the buildings now exited." — Brian Dykes, CFO

Assessment: The program is on track and the disclosure around it is thinning at exactly the point where verification matters most. Through June every quarter carried a building count, an operational-position count and an automation percentage. This quarter carried none of the three: no buildings closed, no positions removed, no hours eliminated, and no share of volume flowing through automated buildings, only a forward statement that nearly 70% of peak packages will. The $100M increase in the exclusion range is small in absolute terms and it is the second consecutive quarter in which a guidance line moved in the company's favour on revenue and against it on a cost item without the release saying so.

9. Amazon Vanished From the Disclosure and the Volume Tail Got Longer

For the first time in the three quarters we have covered, UPS did not size its largest customer. June gave 9% of revenue. This quarter gave neither a percentage nor a dollar, only references to "the deliberate actions we took earlier in the year to reduce Amazon volume" as the explanation for a 3.3% volume decline. No analyst asked for the figure.

What did get disclosed is a longer tail than the July framing implied. In June the glide-down was described as complete, with volume growth excluding Amazon arriving in the back half across all segments. On this call the reported-volume headwind was extended.

"We will have the wrap of the Amazon volume on the top-line volume number until we get through the second half of next year. But focus on the core volume growth, right? We had nearly 2% core volume growth in the third quarter." — Brian Dykes, CFO

The fourth quarter carries the same shape: total U.S. average daily volume is guided down approximately 1.3% year over year, with growth excluding Amazon. The business-to-business line also went undisclosed in percentage terms for the first time this year. The CEO said business-to-business volume was down year over year with 60% of the decline Amazon-related, where March and June both carried a penetration figure, 45.2% and 43.8% respectively.

On Amazon as a competitor rather than a customer, management was unusually direct and gave a structural rather than a reassuring answer.

"Our networks are just different. They have a linear network. We provide reverse logistics. Their ground offer is two to five days reliable delivery. We have ground guarantees. We're just a different network." — Carol Tomé, CEO

Assessment: This is the third consecutive quarter in which the single largest variable in the volume line is described qualitatively, and the first in which it is not quantified at all. Nine percent of approximately $92.6B is roughly $8.3B, so the omission is not a rounding item, and it now sits alongside a statement that the reported-volume drag runs for another three to four quarters. We continue to treat the retained Amazon book as a flat contributor. The competitive answer, by contrast, is the most substantive thing management has said on the subject, because it rests on service-level differences rather than on relationship assurances.

10. Peak Is Guided to a 25% Sequential Ramp on Lower Year-Over-Year Volume

The fourth quarter is the quarter that matters most in this guide and the operational preparation was described in more detail than the financial requirement. A little over 100 U.S. customers account for approximately 80% of the peak surge, and on their forecasts UPS expects average daily volume to rise approximately 25% from the third quarter to the fourth, which management called consistent with last year. In absolute terms that is about 4 million additional pieces per day. Total fourth-quarter U.S. average daily volume is still expected to be down approximately 1.3% year over year.

The cost shape of a smaller peak is the part worth modelling, and it is genuinely favourable.

"By bringing down the water level on some difficult volume, right? It means that we are renting less trailers, we're leasing less aircraft. While we're hiring, we're not having to hire and train as many people in such a short time period. So it's going to allow us to deliver a great peak for our customers, a great service peak, and a great cost peak." — Brian Dykes, CFO

Industry forecasts for holiday online retail sales were cited at approximately 9% growth. Internationally the peak step-up is "not quite 25%, but it's close to 20%," generating demand surcharges that management explicitly named as part of the approximately 20.0% fourth-quarter International margin guide. Eight consecutive years of industry-leading peak on-time service were cited, with a stated intention to make it nine.

Assessment: This is a peak that UPS enters with capacity ahead of volume rather than behind it, and the cost argument is the right one: a network that rents fewer trailers and leases fewer aircraft converts a given surge more profitably. The financial ask behind it is the largest single-quarter year-over-year requirement of the year, at 13.2% adjusted operating profit growth and 92bp of margin expansion. The reconciling point is that the September quarter just delivered 14.0% on a comparable basis, so the fourth quarter is being asked to repeat what the third produced rather than to accelerate from it.

11. Pricing: the 2027 Rate Increase Matches 2026, and the Mix Tailwind Fades

UPS published its 2027 general rate increase in the week of the call, at a 5.9% average net increase on Ground, air and international daily rates effective 21 December 2026. That is the same headline rate as the 2026 increase, which was also 5.9%, so the CFO's characterization checks out precisely. The principal competitor announced a matching 5.9% average increase effective early January. The headline understates the realized increase, because the most-used accessorial charges rise by an average close to 10%.

The more consequential pricing disclosure was about composition rather than level, and it is the first clear statement of what revenue per piece looks like once the Amazon effect stops helping.

"As we wrap the Amazon declines, that mix benefit will come down, and it'll normalize to something that's closer to historic levels. However, we continue to see a strong pricing environment... We expect to continue to get that kind of 200-300 basis point base rate pricing and mix benefit as we go forward." — Brian Dykes, CFO

Assessment: A 200 to 300 basis point combined base-rate and mix contribution is a materially lower forward number than the 6.1% revenue per piece this quarter produced or the 9.3% June produced, and management named it unprompted. That is the honest framing of what the premium strategy delivers in steady state, and it is the number to build 2027 on rather than either of this year's prints. Combined with the 50 to 100 basis point cost spread, it implies a business that grows revenue per piece in the low-to-mid single digits and expands margin by a few tens of basis points a year. The variable absent from both halves of that sketch, for a third consecutive quarter, is the contractual wage escalator.

12. A New Operating Model, a $922 Billion Market Frame, and the 2028 Labour Clock

Two structural items arrived with this print and neither was a response to the quarter. In September UPS announced a new global operating model splitting responsibility three ways: U.S. domestic businesses including the retail, returns and financing adjacencies; international small package together with global healthcare, forwarding and logistics; and a single global operations organization covering air, automation and surface transportation. The stated purpose is standardizing operational processes across geographies while keeping local flexibility.

The second item is a market frame the company has not previously put on the record at this granularity: an addressable market of approximately $922B from "manufacturing floor to the front door," of which approximately $535B is premium globally and approximately $150B is premium in the United States. Inside it, a $15B logistics opportunity attached to data centres, complex healthcare logistics growing around 6% a year, and U.S. high-tech small package revenue up 30% year over year on a small base.

The third structural item is a date. The Teamsters agreement expires on 1 August 2028, which now sits inside a normal forecast horizon, and the question about it was answered with the CEO's own experience rather than with a number.

"Except for our Founder, Jim Casey, I believe I will be the only CEO who has negotiated two Teamster contracts. From that, I think it gives me a distinct advantage because there are many lessons learned from the 2023 contract negotiation." — Carol Tomé, CEO

Assessment: The operating model is the right structural response to a company whose domestic network is now fixed and whose international network is not, and the global operations head's framing of exporting U.S. productivity techniques abroad is the most specific synergy claim attached to it. The market frame is useful as a ceiling and useless as a forecast, and its appearance alongside a refusal to give 2027 guidance until January is a reasonable trade. The labour clock is the item that has not improved: this is the third consecutive quarter in which a direct question about the 2028 negotiation was answered with something other than its cost, and a company that has now put a multi-year margin algorithm on the record has still never sized the wage escalator in dollars or in cents per piece.

Guidance & Outlook

The full-year changes are in the table under Key Topic 5. This section covers the fourth-quarter segment ranges, the implied requirement, the full-year segment arithmetic and where the guide sits against the Street.

Fourth-Quarter Segment Guidance

A consolidated fourth-quarter revenue and operating profit figure is new this quarter. For the previous two quarters the reader had to assemble the consolidated forward view from segment ranges, and that gap has now closed. Below the line, management guided approximately $190M of fourth-quarter expense and a 23% tax rate.

The segment guides reconcile to the consolidated guide, which is worth verifying rather than assuming. Applying each guided growth rate and margin to the prior-year segment base produces $25,683M of revenue and $3,261M of adjusted operating profit, against the $25,711M and $3,272M implied by subtracting nine-month actuals from the full-year guide. The two routes agree within $28M of revenue and $11M of profit, so the guide is internally consistent and the reader can hold management to the segment detail.

Implied Fourth Quarter

The earnings line cross-checks. Taking $3,272M of adjusted operating profit less the guided $190M of below-the-line expense, taxed at 23% over 852M diluted shares, gives $2.79 against the $2.76 residual, so the segment guide, the below-the-line guide and the earnings guide are mutually consistent within three cents.

Two observations about the shape of that requirement. First, the profit ask of 13.2% is almost exactly what the September quarter delivered on a comparable basis, 14.0%, which makes the fourth quarter a repeat rather than an acceleration. That is a materially easier test than the raw year-over-year figures imply, and it is the main reason we are close to the company's number rather than well below it. Second, the free cash flow requirement is the softest line in the guide: $1,880M against $2,726M produced in the same quarter last year, a 31% lower bar, in a year when nine-month capital expenditure ran 25.2% below last year against an unchanged $3.0B full-year plan. Either fourth-quarter capital spending jumps to roughly $779M or the capital number comes in light, and both paths leave room above the free cash flow guide.

The Full-Year Segment Arithmetic, and the Fact Nobody Led With

Carrying the fourth-quarter guides through to the full year produces the table that frames the whole of 2026, and it produces a result that no part of the release or the call stated.

Street at: the raised guide sits above where the Street was before the July print, when published full-year consensus was approximately $90.38B of revenue and $7.13 of adjusted earnings per share. The guide has moved from approximately $91.2B and $7.22 to approximately $92.6B and $7.28 since then, so for two consecutive quarters the company has guided ahead of that base. Published price targets rose across the board after the print and no rating changed, which is the signature of a quarter that confirmed the prevailing view rather than altering it. The target distribution runs from the mid-$90s to the mid-$130s with a cluster in the $112 to $131 band, and the last complete session closed below every point in that cluster.

Guidance style: unchanged across the two raises. Management raises the revenue line generously, raises the profit line by a tenth as much, keeps the segment ranges wide enough to land inside, and declines to extend anything into the next year until January. The one place that pattern breaks in the company's favour is free cash flow, where the fourth-quarter requirement is 31% below last year's actual, and the one place it breaks against the company is the Supply Chain Solutions margin, where the guide asks for a level the segment has not produced cleanly in any quarter of the published series. Our estimate takes the domestic guide at face value, shades International slightly and discounts Supply Chain Solutions.

Analyst Q&A Highlights

Free Cash Flow Into 2027 and Whether the Dividend Is Covered

The first question of the call went straight to the item the print had just changed, asking what the free cash flow step-up looks like next year once the Driver Choice cash requirement disappears, and whether dividend coverage follows. The answer did two useful things: it separated the business-performance portion of the raise from the pension-funding portion unprompted, and it named the specific amount that may have to be given back. The CEO then closed the exchange with the plainest dividend statement she has made.

Q: "If we think about the Driver Choice Program and take those costs out, it would imply a big step up, or maybe not big, but solid step up in free cash flow for 2027... just give some indication on how you're thinking about what free cash flow could look like or what the magnitude of that step up could look like for 2027, and just safety around the dividend or coverage of the dividend."
— Ari Rosa, Citigroup

A: "The Driver Choice cash requirement goes away, which will drive an improvement in free cash flow. Depending on how we fund the pension, we may then have to offset the $450 million in there. Obviously, look, we've always talked about the fundamental driver of free cash flow is getting the domestic margin to start moving in the right direction... we expect free cash flow to continue to move up and improve our dividend coverage as we grow back into the dividend payout ratio that we've talked about for a long time."
— Brian Dykes, CFO

Assessment: The most informative exchange on the call and the one that should change a model. Two numbers were volunteered that are not in the release: the $450M pension item that may reverse in 2027, and a Driver Choice cash figure of $1.4B for this year against the $1.2B of profit-and-loss exclusion the release carries. Taken together they mean 2027 free cash flow starts from a base roughly $450M lower than the guided $6.1B suggests and gains roughly $1.4B as the buyout cash stops, which is a genuine step-up and a smaller one than the headline arithmetic implies. The phrase "grow back into the dividend payout ratio" is also a quiet admission that the current ratio is not the target one.

The Earnings Algorithm Once the Glide-Down Is Behind the Company

A request to zoom out past 2026 and describe the growth algorithm produced the quarter's most expansive answer, and the most granular addressable-market structure the company has put on the record in our coverage period. The answer led with market size rather than with margin, which is itself a change in emphasis from a management team that has spent two years talking about cost.

Q: "I know it's a little early to talk about 2027, but what do you see as the growth algorithm for UPS as we go forward? Now that the glide down is done, you're focusing on B2B, pricing is good. How do we think about all of that, and what does it mean for earnings growth over the long-term?"
— Chris Wetherbee, Wells Fargo

A: "If you then roll that up to what we think is addressable, it's about $922 billion... the premium opportunity is about $535 billion. Here in the United States... it's about $150 billion. So you can see there's a lot of market for us to grow into, and the mathematical equation of what this means for our earnings is that we expect volume growth. We expect our RPP to grow faster than our cost per piece so that we have margin expansion."
— Carol Tomé, CEO

Assessment: The algorithm as stated has three terms, volume growth plus a positive price-cost spread plus a more efficient network, and only the second of the three has any evidence behind it. Volume has declined in every quarter of 2026 and is guided to decline in the fourth, and the company has just told the market the reported figure stays negative into the second half of 2027. A $922B addressable market is a ceiling, not a forecast, and its arrival in the same answer that declined to give 2027 guidance is a reasonable trade rather than an evasion. The sentence to hold management to is the middle one, because it is the only checkable claim in the paragraph and it is checkable every ninety days.

Whether the Cost Actions Produce Durable Incremental Margins

The most analytically precise question of the call used the implied second-half margin as a 2027 base and asked what incremental margins look like on top of it once the Amazon noise clears. The answer is the clearest statement management has given that the margin reset is structural, and it named the four components.

Q: "If we are looking at second half margins around 8.9%, I think that should be a base for 2027. I am just wondering, as you think about the follow-through from all the cost actions you have taken, how does that set you up for incremental margins looking forward?"
— David Vernon, Bernstein

A: "I absolutely believe that the changes that we're making are driving durable margin improvement here. These are structural, right? Start with the volume base. In the third quarter, our core volume, so the volume outside of Amazon, grew nearly 2%... That is structurally different volume than what we have been growing with over the last several years... we've completed now the network reconfiguration. We finalized the Driver Choice Program in the third quarter. We've completed the building restructurings. And we've had the hours come down commensurately with the volume."
— Brian Dykes, CFO

Assessment: Management did not dispute the 8.9% second-half figure as the 2027 base, which is an implicit endorsement of it and the most useful thing in the exchange, because it is a number the company has never guided to directly. Our own arithmetic from the fourth-quarter guide produces 8.86%, so the premise and the guide agree. Where the answer is weaker is on incremental margins specifically, which were not quantified: the reply listed completed actions rather than an incremental rate, and the forward margin content came back to the same 50 to 100 basis point spread. A business whose cost program is finished and whose volume is still declining has a harder incremental-margin question than this answer engaged with.

Whether Domestic Margin Can Escape Its Three-Year Range

The sharpest framing of the call observed that U.S. margins have sat around 7.5% for three years and asked, directly, whether a double-digit domestic margin is reachable or whether the fourth quarter's 30 basis points of expansion is the right magnitude to extrapolate. Both executives answered, and the answer was the same in substance: steady, small, every year.

Q: "If you look U.S. margins for the last three years, they have been right around 7.5%... are there big opportunities for margin improvement? Can we break out of this range and can we ever get back to a double-digit margin, or should we look at Q4 where you are saying U.S. margins are going to be up 30 basis points or so? Is that the right magnitude of margin improvement to think about going forward?"
— Scott Group, Wolfe Research

A: "There are no step changes, but we will be moving up out of the range as we continue to improve the business mix and drive productivity with some of the initiatives Carol outlined."
— Brian Dykes, CFO

A: "That's the way you should think about it, is every year improving margin. Every year improving the margin."
— Carol Tomé, CEO

Assessment: "No step changes" is the single most important forward sentence on the call and it should be taken literally. A 50 to 100 basis point annual spread on a flat-to-declining volume base compounds domestic margin by tens of basis points a year, not hundreds, which means a double-digit domestic margin is a decade-scale outcome rather than a cycle one. That is a lower trajectory than the June quarter's 400 basis point sequential step encouraged people to extrapolate, and management has now said so twice in two quarters. It also validates the full-year table above, where the domestic segment finishes 2026 with a 22 basis point margin decline despite the inflection.

Where Revenue Per Piece Goes When the Amazon Mix Benefit Laps

A two-part question on volume visibility and on the price-cost spread produced the most model-ready answer of the call, because it decomposed the forward revenue-per-piece line into its parts and said which part disappears.

Q: "When we think about CPP versus RPP next year, do you assume that you get, let's say 3% base price, and then you get some favorable mix on top of that? Like a point of mix on top? I am just trying to think about what keeps you above CPP next year and what are the pieces."
— Tom Wadewitz, UBS

A: "You'll notice that over the course of this year, the RPP level has been coming down... In the third quarter, it was 6.1% growth, about half of it was fuel, the other half is base rate and mix, right? As we wrap the Amazon declines, that mix benefit will come down, and it'll normalize to something that's closer to historic levels... We expect to continue to get that kind of 200-300 basis point base rate pricing and mix benefit as we go forward."
— Brian Dykes, CFO

Assessment: Build the 2027 model on the 200 to 300 basis points, not on this year's 6.1% or June's 9.3%. The answer is internally coherent and it is also a downgrade: a combined base-rate and mix contribution of 200 to 300 basis points, plus a published general rate increase of 5.9% that realizes well below its headline, against a cost line management says it now controls, describes a business growing revenue per piece in the low-to-mid single digits with a 50 to 100 basis point cushion over cost. What the decomposition still omits, for a third consecutive quarter, is the contractual labour escalator that lands on the cost half of the same equation before the next rate increase cycle completes.

When Reported Volume Turns, and Whether International Is Accelerating

A compound question on the volume inflection and the apparent International step-up drew the two forward disclosures that most changed our estimate: the duration of the reported-volume headwind, and the mechanism behind the approximately 20% fourth-quarter International margin.

Q: "Maybe just following up on some of the questions, maybe your thoughts on inflection to volumes. When do you expect them to turn positive?... It sounded like your international margins are actually accelerating, right? I think you said over 20%. So instead of gradual, that sounds like a bigger step up on the international side."
— Ken Hoexter, Bank of America

A: "We will have the wrap of the Amazon volume on the top-line volume number until we get through the second half of next year... On the international side, look, I think we had always contemplated a step up going into the fourth quarter. Look, we have a peak in international as well. It's not quite 25%, but it's close to 20%, a step up in volume, and that creates a surge in demand and related demand surcharges that help us improve the margins as we go into the fourth quarter."
— Brian Dykes, CFO

Assessment: Two answers, one favourable and one not. The International explanation is credible and specific, and it also makes clear that the approximately 20% fourth-quarter margin rests on peak demand surcharges rather than on a structurally recovered franchise, which is why we shade that line rather than take it. The volume answer is the harder one: the reported average daily volume figure stays negative into the second half of 2027, which is three quarters longer than the June framing implied, and it removes the first term of the three-term growth algorithm described earlier on the same call for another year. Core volume growth of roughly 2% is the right thing to watch and it is not what any screen will show.

Business-to-Business Volume and Amazon as a Competitor

The only question of the call that pressed on competitive risk asked about business-to-business volume and about the competitor's publicized pricing push into the same premium verticals. The answer disclosed a volume direction the release omitted and then gave a structural rather than reassuring account of the competitive overlap.

Q: "I want to come back to B2B trends in the quarter. I'm not sure if you guys addressed that, what your volumes were specifically, and then maybe more broadly about competition, because I think Amazon's obviously made a very big public push here with pricing going up. Are you seeing incremental competition for those premium verticals?"
— Brandon Oglenski, Barclays

A: "On the B2B side, we didn't disclose what the volume trends were, but they were actually down year-on-year, but 60% of the decline was Amazon related."
— Carol Tomé, CEO

After the CFO's remarks on the DAP program: "As we look at Amazon shipping, we haven't seen any measurable or meaningful impact to our business since the announcement they made earlier this summer... Our networks are just different. They have a linear network. We provide reverse logistics."
— Carol Tomé, CEO

Assessment: The competitive answer is better than the disclosure answer. Grounding the claim in service-level differences, reverse logistics and ground guarantees against a two-to-five-day linear network is a falsifiable argument rather than an assurance, and it is more useful than a relationship description. The disclosure half is a step backwards: business-to-business volume was down with the direction given only when asked, no penetration percentage for the first time in three quarters, and the 60% attribution to a customer whose size the company declined to state anywhere in the quarter. A reader cannot verify the 60% because the base is undisclosed.

Whether the Final Piece of the Cost Program Slips Into 2027

The last question of the call tested whether the remaining $1.2B of the $3B cost target can land in a peak quarter, or whether the resources peak demands push part of it into next year. The answer rested the commitment on a mechanism rather than on confidence.

Q: "From the $3 billion of cost savings, $1.8 billion through September, I know it was supposed to be back half loaded, but with 40% for the last quarter, and especially a peak quarter... Is it still kind of line of sight slam dunk, or is there a chance that it peaks better than expected and some resources are required for that some of that $1.2 billion of the final piece of that savings puzzle slips into the beginning of next year?"
— Jonathan Chappell, Evercore ISI

A: "No, we feel comfortable with it, Jonathan, because a lot of it is related to the hours that ramp down with the Amazon volume... As those hours exit, that volume comes out. That was always going to be weighted to the last quarter of the target, and we feel comfortable. Additionally, the driver buyout, the roughly 1,600 people finished exiting at the end of the third quarter, so you get the full wrap of that. Then obviously all the buildings now exited."
— Brian Dykes, CFO

Assessment: The mechanism is the right answer to this question, because hours leaving with volume is arithmetic rather than execution, and the final driver departures landing at the end of September means the fourth quarter captures their full benefit rather than a partial one. One detail cuts the other way and went unremarked: the release describes the full-year program benefit as "nearly $3 billion" where July said "approximately $3 billion," and the question was framed against the $3B figure. A company that is comfortable with a target does not usually soften the adverb in front of it in the same document.

What They're NOT Saying

1. Amazon, in any form: for the first time in the three quarters we have covered, neither a percentage of revenue nor a dollar. The June quarter gave 9%. This quarter gives only "the deliberate actions we took earlier in the year to reduce Amazon volume" as the explanation for a 3.3% volume decline, alongside a statement that the reported-volume drag continues into the second half of 2027. Nine percent of the guided revenue is roughly $8.3B, so this is not a rounding item, and no analyst asked.
2. Which Supply Chain reporting unit was impaired: $182M was written off against an unnamed unit, disclosed in one clause of the prepared remarks and one line of the reconciliation. The unit was not named in the release, the slide deck or the call, and the item drew no question. It is 6.1 percentage points of the segment's quarterly margin.
3. How Supply Chain Solutions gets from 9.2% to 12.2% in one quarter: the fourth-quarter guide would be the segment's best clean adjusted margin in the seven-quarter published series, the prior best being 10.3%, asked for immediately after a missed range and a write-down. No bridge appears anywhere in the materials.
4. That the full-year transformation exclusion range rose $100M: $1.3B to $1.5B in the July release became $1.4B to $1.6B in October, with the Driver Choice component moving from $1.1B back to $1.2B. Neither change is flagged as a change, and neither was mentioned on the call.
5. That the completion date moved: the March and June releases said the initiatives "are expected to conclude by 2027." The October release says the Efficiency Reimagined initiative is "not expected to conclude until the fourth quarter of 2027." That is a different statement and it was made without comment.
6. That the program-benefit wording softened: "approximately $3 billion" in July became "nearly $3 billion" in October, in a filed document, while the question in Q&A was asked against the $3B figure and answered with "we feel comfortable."
7. The operational counts, all of them: buildings closed, operational positions removed, hours eliminated and the share of U.S. volume flowing through automated buildings were each disclosed every quarter through June. None appears in the October release or on the call. June gave 45 of roughly 50 buildings, approximately 30,000 fewer positions and 68.5% automation. October gives a forward peak statement that nearly 70% of packages will flow through an automated process, which is not the same metric.
8. The bridge between the two Driver Choice numbers: the filed release sizes the Driver Choice element of excluded operating expense at $1.2B. The CFO sized this year's spend at $1.4B on the call. One is a profit-and-loss exclusion and the other is cash, and no reconciliation between them was offered.
9. Share repurchases, for a third consecutive quarter: zero in the first nine months against $1,000M a year ago, and the word appears nowhere in the release, the deck or the transcript. Total debt rose $1,951M over the same period.
10. Business-to-business penetration: given in March at 45.2% and in June at 43.8%, and omitted entirely in October. The only disclosure is that volume was down with 60% of the decline Amazon-related, which cannot be verified against an undisclosed base.
11. What the 2028 labour step-up costs: asked in substance for a third consecutive quarter and answered with the CEO's negotiating history. The contractual escalator applies to the workforce that remains after the reductions and it has never been quantified in dollars or in cents per piece, by a company that has now put a multi-year margin algorithm and a 2027 pricing framework on the record.
12. Where the prior-year gain and this year's impairment sit in the expense detail: without that, the "other expenses" line up 27.1% is uninterpretable from the filed table, and a reader is likely to conclude the cost base is deteriorating when the more probable reading is that it is roughly flat.

One item moved the other way and deserves credit. For the first time in three quarters UPS gave a consolidated forward guide, with fourth-quarter revenue of approximately $25.7B and adjusted operating profit of approximately $3.3B, and it moved the free cash flow and pension-contribution figures out of spoken remarks and into the filed release. The forward disclosure improved in the same quarter the historical operational disclosure thinned.

Market Reaction

• Pre-print setup: UPS closed at $100.32 on October 26, up 1.1% year to date against the S&P 500's 12.1%, up 12.4% over trailing twelve months and up 6.8% over the trailing thirty days. The 52-week closing range entering the print was $87.03 to $120.00, placing the stock at roughly 40% of its own year's range, against 92% entering the July print on the same calculation.
• Reaction session (before-open reporter, same-day reaction): opened at $101.54, a 1.2% gap up, traded a range of $99.61 to $106.59, and closed at $106.09, up 5.8% or $5.77. The S&P 500 rose 0.6% in the same session, so roughly 520bp of the move was stock-specific.
• Volume: 10.9M shares against a 4.9M thirty-day average, 2.2 times normal, against the 3.0 times that accompanied the July print.
• Intraday shape: the stock was quoted roughly 2.3% lower before the open, dipped to $99.61 or 0.7% below the prior close early in the session, and closed within 0.5% of the high and 93% of the way up the day's range. The session re-rated upward as it progressed.
• Sessions since the print: $105.62, $107.33 and $108.25 on the three following sessions, taking the cumulative move from the pre-print close to 7.9% over four sessions against the S&P 500's 1.3% decline, or 916bp of relative gain.
• Peer reaction: the principal competitor rose 2.2% on the day of this print without reporting, and 4.2% from the same pre-print close through the same four sessions, after reporting the identical calendar quarter the following evening with revenue of $22.86B against UPS's $22.853B and an adjusted earnings beat of 11.3% against UPS's 3.7%.

The session was the exact inverse of July's, and in both cases the shape was about the estimate base rather than the operating result. In July the stock gapped down 3.9% and closed near the low on three times normal volume, because the Street's models had not absorbed a charge the company had disclosed in April. In October it was quoted down before the open, traded briefly below the prior close, and then closed near the high, because the headline growth rates had not absorbed a gain the company had disclosed in a footnote. Two consecutive quarters, two violent sessions, and both resolved by reading something the company had already published.

The early weakness is legible and it is not irrational. The three numbers a screen carried at 7:00 in the morning were adjusted operating profit down 3.7%, adjusted operating margin down 100bp and adjusted earnings per share down 2.9%, with a Supply Chain Solutions segment whose profit had halved on 18.2% more revenue. Every one of those is correct, and every one of them inverts once the $330M prior-year gain comes out. The re-rating through the session tracks the call, which ran from 8:30 and spent its first twenty minutes establishing that the domestic segment had hit its guide, International had beaten its range, and the full-year outlook was going up on all four metrics.

The pre-print setup did the rest of the work, and this is the part worth carrying forward. The stock entered the print up 1.1% year to date against an index up 12.1%, at roughly 40% of its own 52-week range, having given back the entire July de-rating and more. There was no momentum position to unwind, no crowded long to shake out, and a 6.5% yield underneath. July's 6.6% decline happened to a stock at 92% of its yearly range; October's 5.8% advance happened to the same stock at 40%. The quarter was arguably less good than June's on the segment detail and the market reaction was better by twelve percentage points, which is a lesson about positioning rather than about parcels.

Street Perspective

Debate: Is the Domestic Inflection Compounding, or Already at Its Ceiling?

Bull view: The bull case being made on the Street is that two consecutive quarters of a positive price-cost spread, delivered on top of a completed network reconfiguration, a finished buyout program and exited buildings, is a permanently repriced business. Segment profit grew 14.3% on 3.3% fewer pieces, the contribution-per-piece arithmetic reconciles to the reported profit change within $13M for a second quarter running, and the second-half margin of roughly 8.9% is a base management declined to dispute when it was offered as the 2027 starting point.

Bear view: The bear camp contends that the ceiling is already visible in management's own words. The spread narrowed from 130bp to 60bp in one quarter and sits at the bottom of the committed band. Contribution per piece fell 15.7% sequentially, twice last year's seasonal give-back. Asked whether the segment can break out of a three-year 7.5% range, the answer was "there are no step changes." And the full-year table says domestic operating profit finishes 2026 within $5M of 2025 on 2.8% more revenue.

Our take: The bulls own the level and the bears own the slope, and the slope is what matters from here. The reset is real and we have stopped arguing about it: three independent checks, the contribution reconciliation, the ruled-out property gains and the delivered guide, leave no non-operational explanation standing. But a 50 to 100 basis point annual spread on a shrinking volume base compounds margin in tens of basis points, management has now said "no step changes" out loud, and the correct 2027 base is the roughly 8.9% second half rather than any extrapolation of the June quarter. Underwrite a permanently better business that improves slowly, which is what the current multiple is roughly paying for.

Debate: Is the Supply Chain Write-Down Isolated, or the Start of a Portfolio Problem?

Bull view: The bull framing is that $182M is 0.25% of total assets, the segment grew revenue 18.2% with forwarding up 17.5% and logistics up 17.2%, the digital businesses grew nearly 50%, and the charge almost certainly sits against a legacy unit management has already named as declining. Excluding the prior-year gain, segment profit grew 33%. An impairment that clears a non-core carrying value out of the way ahead of a 12.2% fourth quarter is housekeeping, not deterioration.

Bear view: The bear argument is about what the charge implies and what the response to it revealed. A goodwill impairment is management telling its auditor that an owned business will not earn what was paid for it, recorded in the same quarter the segment missed its own margin range, with the unit unnamed and the item drawing zero questions. Adjusted operating expense grew 36.3% against revenue up 18.2%, which is negative operating leverage. And the fourth-quarter guide asks for the best clean margin in the published series with no bridge.

Our take: The bears have the better of this and the market has not engaged with it at all, which is the more interesting fact. The size of the charge is not the point; the information content is. A segment that grows revenue 18% while expenses grow 36% is buying growth, and a write-down in the same quarter says at least one past purchase did not work. We model the fourth quarter at roughly 11.0% rather than 12.2%, which is still the best clean quarter in the series, and we treat a named reporting unit and a stated basis for the step-up as the disclosure that would resolve the debate. Until then this is where the risk in the guide sits, and it has rotated here from International.

Debate: Does the Free Cash Flow Raise Unlock Capital Returns, or Is It an Accounting Bridge?

Bull view: A growing consensus view holds that the constraint has broken. Nine-month free cash flow of $4,220M was up 53.8% and covered 104% of dividends paid against 68% a year ago. The full-year guide of approximately $6.1B takes the payout to 89% from 98%, the fourth-quarter requirement is 31% below what the business produced last year, capital intensity is running at 3.3% of revenue, and the Driver Choice cash requirement of $1.4B disappears entirely next year.

Bear view: The bear view is that only a quarter of the improvement is real and the headline is not. $450M of the $600M raise is a pension contribution funded with UPS bonds rather than cash, which the CFO said 2027 may have to give back. On the underlying figure the payout is 96% rather than 89%. Repurchases are zero for a third quarter against $1.0B a year ago, total debt rose $1,951M while equity fell $1,229M, and leverage went from 1.48x to 1.74x in a year of improving cash flow.

Our take: Both are right about their own number and the bull case wins on trajectory. The structural payout improvement is 98% to 96%, which is small, but the direction has reversed for the first time since we initiated and the 2027 arithmetic is genuinely better: a $1.4B cash outflow stops, offset by a possible $450M pension reversal, against a dividend held at $5.4B. That would put the dividend plausibly well under 90% of cash generation on an unmassaged basis. What the company has not done is pre-commit any of it, and a repurchase authorization remains the cleanest available signal. We are not waiting for it to stay constructive, and we would raise the multiple if it arrived.

Debate: Is the Fourth-Quarter Guide Conservative, or Does It Hide the Hard Part?

Bull view: The bull argument is that the fourth-quarter ask is a repeat rather than a stretch. The guide needs 13.2% adjusted operating profit growth and the September quarter just delivered 14.0% on a comparable basis. The segment ranges reconcile to the consolidated guide within $11M, so there is no hidden slack or hidden stretch. The domestic segment is asked for only 32bp of year-over-year expansion, the cost program's final tranche is mechanical hours leaving with volume, the final buyout departures landed at the end of September so the fourth quarter captures their full benefit, and the free cash flow requirement is 31% below last year.

Bear view: The bear case is that the aggregate hides the distribution. The guide requires 92bp of consolidated margin expansion, and the segment asks behind it are lopsided: the domestic segment, which is 66% of nine-month revenue, is asked for only 32bp of its own year-over-year expansion, while the other two are asked for 200bp and 189bp. International's 20.0% rests on peak demand surcharges that management named as the mechanism. Supply Chain Solutions' 12.2% would be a two-year high, one quarter after a 9.2% print and an impairment. And the whole guide depends on a peak that is being planned against the forecasts of roughly 100 customers.

Our take: The bears have correctly located the risk and overstated it. The 13.2% consolidated ask is comfortable because it equals what the September quarter produced, and the internal reconciliation within $11M means the reader can test each piece rather than trust the total. Our haircuts are deliberately where the bears point: 19.5% at International against the guided 20.0% and 11.0% at Supply Chain Solutions against 12.2%, which takes $74M off the implied fourth-quarter profit and leaves us $0.04 below the earnings guide. That is a smaller gap to the company than we carried in either of the last two quarters, and it is the measure of how much the forecasting record has improved.

Model Update & Valuation Framework

The table sets our revised estimates against both the company's raised guidance and the estimates published with the second-quarter recap.

Valuation. At the $108.25 close of the last complete session before publication, UPS trades at 14.9 times the company's raised full-year adjusted earnings guide of $7.28 and 15.0 times our $7.24, with a 6.1% dividend yield on the declared $1.64 quarterly rate. We are holding our multiple range at 15 to 17 times, which on $7.24 produces a $109 to $123 fair-value range with a midpoint near $116, about 7.2% above the current price. Adding the dividend, the twelve-month total return we underwrite is roughly 13%.

Why the multiple does not move. In July we raised the range from 14 to 16 times to 15 to 17 times, because the margin recovery had produced its first quarter of evidence in the form that is hardest to fake. This quarter is confirmation of that same fact rather than a new one: the spread held inside its committed band, the domestic segment delivered its guide, and the International pillar that had moved the wrong way moved back. Confirmation is worth an estimate increase, which it has received, and it is not worth another point of multiple. The three conditions that capped us at 17 times are all still in place. Volume is declining and is guided to keep declining in the reported figure into the second half of 2027. Consolidated adjusted margin finishes 2026 below 2025. And the dividend still absorbs 96% of free cash flow on an unmassaged basis with no repurchase.

Where we were wrong last quarter. We modelled International at roughly 14% for the full year and called the company's mid-teens target reachable only at its floor and only with a record fourth quarter. The segment printed 14.9% against a 13% to 14% guide and the full year now computes to 15.0% on management's own fourth-quarter number. That one-point miss is roughly $200M of operating profit and it accounts for most of the increase from $7.10 to $7.24. The judgment that was right was treating the June quarter's 130 basis point spread as flattered rather than as the new run-rate; 60bp is what arrived.

What would change the rating. To Hold: a fourth-quarter domestic adjusted margin below the guided approximately 10.5%, a second consecutive Supply Chain Solutions quarter below its guided range, or a 2027 initial guide that puts consolidated adjusted operating margin below 2026's 9.45%. To Underperform: any softening of the $5.4B dividend commitment, or a 2027 free cash flow guide below $5.4B once the Driver Choice cash stops. To add to the position: a repurchase authorization, a first Amazon revenue dollar disclosure, or a named Supply Chain reporting unit with a stated basis for the fourth-quarter margin step-up.

Thesis Scorecard Post-Earnings

The table grades the seven pillars established at initiation in April, in the same order and against the same wording, as carried in the thesis of record.

Overall: The thesis is modestly strengthened and its centre of gravity has shifted. Three months ago it rested on whether the domestic unit-economics reset was real; it is, and the argument is over. What the quarter adds is the first improvement in the capital-allocation pillar since initiation, and the correction of our own International error. What it subtracts is a new, unpriced item: a goodwill impairment in the smallest segment, unnamed and unquestioned, followed by a guide to that segment's best clean margin in the published series. The one fact that should temper any enthusiasm is the full-year table, where every segment's adjusted margin still declines and domestic operating profit dollars finish within $5M of last year. The inflection happened; 2026 is still a flat year.

Action: Hold the position and add on weakness below $105. All three of the downgrade triggers set in July were tested this quarter and none fired, though the ex-fuel spread test remains incomplete, which is the cleanest evidence available that the forecasting base has stabilized, so conviction rises to 7 from 6. We are not adding at $108.25, because the stock sits inside our fair-value range rather than below it and the next nine weeks carry a peak season and a Supply Chain margin step-up we do not fully underwrite. The January guide is the next event that matters, and a 2027 consolidated margin above 9.45% with a repurchase authorization attached is what would take this to a higher multiple rather than a higher estimate.