Mobility Missed, the M&A Window Closed, and Per-Share Growth Is Guided to Re-Accelerate: Upgrading to Outperform

Rating Action

We initiated coverage of Uber at Hold in Q1 2025, upgraded to Outperform in Q2 2025, maintained that rating through Q3 2025, Q4 2025 and Q1 2026, and returned to Hold in Q2 2026. The August downgrade was narrow and specific. We had underwritten two legs: operating leverage converting high-teens bookings growth into earnings growth at a multiple of that rate, and a programmatic buyback converting that into per-share growth at roughly twice the platform rate. The second leg had been redirected into a €14.2B acquisition bridge and a $10B autonomy program whose returns landed outside a twelve-month horizon, with no end date attached to either.

This quarter supplied the end date. It also supplied the first operational disappointment of the coverage arc. Those two facts pull in opposite directions, and the resolution is what the rating turns on. None of the three re-upgrade triggers our Q2 note named was met as written: there is still no dollar figure for the buyback, no income-statement sizing for autonomy, and Q3 bookings landed inside the guided range rather than above it. The upgrade rests instead on the M&A end date, which that note did not anticipate, together with the nine-month payout and the re-accelerating EPS guide.

The six Q2 commitments, scored. Ninety days ago we wrote down the forward promises management made on the Q2 call. This is how they landed.

Four of six landed at or near the commitment. The one clean failure is the U.S. Mobility acceleration, and the quarter's other negative, a broad international supply shortfall, was not a commitment at all. It is new.

Results vs. Consensus

Year-over-year comparison

Sequential comparison

Revenue assessment. The 10% reported growth rate should be ignored and the 20% bookings growth rate should not. Between them sits the $1.1 billion UK reclassification, which is presentation-only and nets to zero at every line below revenue, and which this quarter management finally sized cleanly in its own disclosure rather than leaving investors to reconstruct it from three documents. What is left after that adjustment is a second consecutive quarter of trip deceleration, and the first this year tied to a shortfall against management's own plan: trips grew 16% against 18% in Q2 and 20% in Q1, and the CFO split that two-point step roughly evenly between international Mobility supply constraints and the lapping of Trendyol Go in Delivery. The second half of that split is arithmetic that reverses on its own. The first half is the quarter's actual problem.

Margin assessment. This is the best margin quarter Uber has reported, and it was earned in the quarter when Mobility volume disappointed, which is the useful part. Non-GAAP operating income margin on gross bookings reached a record 3.9% from 3.4%, Mobility hit a record 8.0% segment margin on bookings from 7.4%, Delivery expanded to 3.70% from 3.48%, and corporate overhead fell to 1.81% of bookings from 1.93%. That overhead line is roughly $72 million of in-quarter operating leverage at current volumes. The one caveat is the widening gap between the GAAP and non-GAAP operating lines: 18bps of GAAP margin expansion on bookings against 50bps of non-GAAP, with the difference carried mainly by a $186M restructuring charge and a $50M legacy-award settlement that management did not discuss, plus higher acquisition costs. The legal and regulatory reserve was flat year over year and did not widen the gap.

EPS assessment. Non-GAAP EPS of $0.88 landed at the top of the $0.84 to $0.88 guide and a penny above consensus, and it did so through a 400bps increase in the non-GAAP tax rate and a 36% increase in interest expense. That is a higher-quality beat than the Q2 print, where EPS came in the upper half of its range while bookings cleared the high end. The deceleration our Q2 note worried about did not happen: growth held at 35% rather than stepping down to the low thirties, and the Q4 guide calls for 34% to 39%, which the CFO characterized as 37% at the midpoint. Against that, the Q4 guide sits below the single-source $1.06 consensus on this line alone, though a second estimate of $0.97 would put it in line, and any gap is structural rather than operational. Interest, tax and share count are now the binding constraints on per-share growth, and all three are consequences of capital-allocation choices rather than of trading conditions.

Segment Performance

The composition inverted from Q2. Three quarters ago Mobility was the segment doing the work and Delivery was the one the market under-appreciated. This quarter Delivery and Freight both beat management's plan while Mobility fell short of it, and Mobility still produced a record margin.

Gross bookings by segment

Revenue and take rate by segment

Take rate here is revenue divided by gross bookings. Management discloses the UK reclassification's effect on the Mobility revenue margin as roughly 350bps in Q3, consistent with the year-to-date impact, leaving roughly 160bps of the 511bps decline attributable to everything else: the deliberate investment in low-cost products that has been running all year, plus a rising share of two-wheeler supply cost in contested markets. Freight take rate is not a meaningful construct because Freight revenue is recognized gross. The sharper point is that Mobility revenue fell 2% while Mobility segment operating income grew 27%, which is only possible because the revenue decline is an accounting artifact and the cost decline underneath it is real.

Segment operating income

Mobility

Gross bookings grew 17% in constant currency against trips up 16%, and the segment missed management's own plan by 1 to 2 percentage points. Judge the quarter on the operating line rather than the revenue line, because revenue is where the UK reclassification lives, and the operating line is a record: $2,362M of segment operating income, up 27%, at 8.0% of bookings. That combination, a volume shortfall alongside a record margin, is the single most informative fact about the segment. It says the shortfall was not bought off with incentives.

Management's account of the shortfall is specific and geographically bounded, which makes it checkable next quarter.

"The large majority of our Gross Bookings deceleration came from outside the U.S., where consumer demand has remained healthy, but supply growth has not kept pace. In multiple international markets, elevated fuel costs, tighter local labor conditions, and competition for drivers constrained our ability to fully serve consumer demand. In India, driver supply growth lagged demand, while Brazil remained undersupplied, although conditions improved modestly from Q2. We also saw an impact in Canada from tighter immigration flows, and the market is currently meaningfully undersupplied." — Uber Q3 2026 prepared remarks

Two quantifications came with it. The Middle East conflict cost 75bps of gross bookings growth in Q3, and the Q4 outlook assumes a 100bps headwind on recent escalations during the busy travel season. That is the rare case of a company sizing a macro drag in both directions rather than invoking it.

The U.S. held up far better, and the mechanism our Bull 1 pillar rests on is still working. Insurance reform continues to flow through, California trip growth accelerated meaningfully and outpaced the rest of the country, and the product barbell is doing real volume: Wait & Save passed an $8 billion annualized bookings run rate and doubled year over year, the premium portfolio grew nearly 35%, and Uber for Business accelerated to nearly 50%. U.S. barbell products grew at three times the rate of the overall business. Against that, management conceded U.S. growth "moderated a bit," which breaks the Q2 commitment that acceleration would hold through the year.

Assessment: A supply shortfall in a business with no demand problem is the better of the two failures available, and Uber has an established playbook for it: reallocate the flexible incentive pool from riders to drivers and let the algorithms place the dollars. Management says supply-hours growth re-accelerated globally in October with notable improvement in Brazil and India, which is the claim the Q4 print will settle. The genuine concern is that two of the three causes, fuel prices and immigration policy, are outside Uber's control and will not respond to incentive spend. Watch whether Canada, which management describes as currently meaningfully undersupplied, is still being named in February.

Delivery

The strongest segment of the quarter and the one that carried the print. Bookings grew 23% in constant currency, revenue grew 23%, segment operating income grew 30% to $1,057M, and margin on bookings expanded 23bps to 3.70% with the take rate essentially flat at 19.23%. Management says the segment outperformed its own plan, and that organic growth accelerated for a second consecutive quarter even though the reported figure is still absorbing the Trendyol Go lapping drag.

The breadth is what makes this hard to dismiss. Strong acceleration in both the U.S. and Latin America, all-time-high category position in the UK, France and Türkiye, a fourth consecutive quarter of accelerating growth in Japan, and roughly 30% growth in Australia on expansion into less dense markets. Grocery and Retail bookings grew about 40% to what the CEO described as a $15 billion run rate, with Costco expanded to nationwide U.S. coverage and already among Uber's top merchants.

"And we are quite pleased in terms of US delivery, which accelerated nicely. it is the second quarter in a row that organic growth and delivery and US delivery accelerated. And, really, it is for a number of factors. 1 is mathematically, our grocer and retail business and direct business which are growing faster than our OFD business, are becoming a higher percentage of the portfolio so to speak. So you just have kind of a bigger, faster growing part of the business getting even larger. Our core restaurant business is also accelerating, and that is because of really merchant selection, the number of merchants that we added in The US. Was up nicely about 12% on a year-on-year basis." — Dara Khosrowshahi, CEO

The monetization layer is now material in its own right. Merchant spending across advertising and offers passed a $10 billion annualized run rate, and the 10-Q attributes $176M of the year-over-year Delivery revenue increase specifically to advertising revenue, $538M across the nine months. On a segment operating income base of $1,057M, an incremental advertising dollar carries far higher contribution than an incremental delivery dollar, which is most of the reason Delivery margin is expanding at all while the company reinvests aggressively in selection and affordability.

The same supply constraint that hit Mobility is now visible here too, and the CEO volunteered it rather than waiting to be asked: "we continue to invest in delivery and couriers to make sure that our courier growth keeps up with the consumer demand that we are seeing on the food side. That is right now the biggest challenge that we have got."

Assessment: Delivery is the segment where the bookings growth, the take rate, the segment margin and the competitive commentary all agree, and the second consecutive quarter of organic acceleration beneath a net inorganic drag is the hardest pattern to manufacture. The complication arrived six days before the print rather than inside it: DoorDash agreed to acquire Instacart for roughly $17 billion, creating a materially larger competitor in exactly the U.S. grocery category Uber is growing at 40%. Uber's own grocery business is scaling, its found rates hit an all-time high in September, and the Delivery Hero transaction brings a built quick-commerce and dark-store capability. But the cheapest route into U.S. grocery scale has now closed.

Freight

Bookings grew 35%, a third consecutive quarter of acceleration, and the operating loss narrowed to $33M from $40M. Freight is 3.0% of bookings and will not move the earnings line, but a business that spent nearly two years contracting is now growing faster than the platform, on Managed Transportation and Capacity Solutions adoption and a European expansion. Management expects Q4 growth broadly consistent with Q3.

Assessment: Immaterial to this year's earnings and increasingly relevant to the autonomy case, for the reason we have described for three quarters: Freight absorbs autonomous capacity in the hours when consumer mobility demand troughs, which is a utilization advantage no single-product autonomous operator can replicate. The segment's return to platform-rate growth is a small positive nobody is paying for.

Corporate overhead

Corporate G&A and Platform R&D grew 13% to $1,082M against bookings growth of 20%, so overhead fell to 1.81% of bookings from 1.93%. That 12bps is roughly $72M of operating leverage in the quarter, or close to $290M annualized at current volumes, and it is the visible output of headcount discipline that likely also explains this quarter's $186M restructuring charge. The charge is excluded from non-GAAP operating income; the savings are not.

Key KPIs

Two disclosure changes are worth flagging because they affect comparability. The Uber One member count returns to the disclosure set at approximately 55 million after being replaced by merchant counts last quarter, which resolves a flag we raised in August. In the other direction, the advertising metric changed basis: Q2's figure was a $2 billion-plus advertising run rate, and this quarter's $10 billion-plus figure combines advertising with merchant-funded offers. The two are not comparable, and the underlying advertising growth has to be read from the 10-Q's $176M year-over-year contribution instead.

Key Topics & Management Commentary

Overall Management Tone: Management led with the shortfall rather than burying it, named the markets, sized the macro drag in both directions, and pointed at an October data trend that will either confirm or contradict it in ninety days. That is the most falsifiable framing of a bad number this management team has offered in our coverage. The posture on capital was the inverse of Q2: where August answered "how much" with a framework and a promise to size things later, this call drew a line, closing the acquisition window through at least 2027 and naming buybacks as a priority alongside autonomy. The only place the confidence outran the disclosure was on who carries the autonomous fleet, where the long-run answer is third-party capital and the near-term answer is Uber's own balance sheet.

1. The Mobility supply shortfall, and why it is the better kind of miss

Uber has spent the coverage arc beating its own Mobility plan. This quarter it did not, and the deviation is small in magnitude and large in signal. Gross bookings growth came in 1 to 2 percentage points below internal expectations, with the large majority of the shortfall outside the United States. Four causes were named: elevated fuel costs in price-sensitive emerging markets, tighter local labor conditions, competition for drivers, and the Middle East conflict. A fifth, seasonality, compounds them.

"And if you zoom out and, look at the macro environment, with the escalation and the war in The Middle East, fuel cost impacts have been felt, globally. Where it shows up most materially for us is in emerging markets where there is a level of price sensitivity on both the consumer and the driver side. And, what we talked about last quarter in Brazil was specifically focused on motor supply from delivery competition in the market. But through the last quarter, we have seen the elevated fuel price impact in markets like Brazil and India in particular, quite meaningfully impacting the supply situation." — Balaji Krishnamurthy, CFO

The CFO's own framing of why this matters was unusually direct: "this is a good problem where we are at the scale that we are at. Operating at, and we still are running into supply constraints rather than demand constraints." The response is targeted spend on driver acquisition, retention and reactivation in the specific markets where demand is outrunning supply, and the early read is positive. Supply-hours growth accelerated globally in October, with notable improvement in Brazil and India.

Assessment: A supply constraint in a demand-rich marketplace is recoverable and a demand shortfall is not, so the distinction matters more than the two points. The uncomfortable part is that this is the second consecutive quarter in which an emerging-market supply problem has grown rather than resolved, and the Q2 note's explicit watch item was whether the Brazil dynamic would appear in a second market. It appeared in at least two, and the cause broadened from a competitive spillover into something closer to a general input-cost and labor-availability problem. The Q4 print is the test, and management has handed us a clean one: unchanged constant-currency guidance of 18% to 22% on a higher base, with October already called out as improving.

2. The acquisition window closes through at least 2027

This is the quarter's most consequential capital-allocation statement and the direct answer to the question our August downgrade was built on. Uber has announced five transactions in 2026: SpotHero for $617M in April, Getir's food delivery business for $361M in July, Blacklane for $554M in August, ezCater for roughly $2.3B announced October 5, and Delivery Hero at a $14.8B equity value. Management has now drawn a line under the list.

"Given the acquisitions we have already announced, including Delivery Hero and ezCater, we are now prioritizing integration and execution. Apart from investing behind our AV strategy, we expect minimal to no further M&A through at least 2027 as we prioritize delivering the growth and synergies from transactions already underway." — Uber Q3 2026 prepared remarks

The CEO was more explicit about what that frees up on the call: "We are going to shift to, I would say, more of the top 2, and we are going to pull back on M&A. Chiefly because we are really focused on integration of the deals that we have already announced." He closed with "next year, I think the focus is gonna be more on organic, autonomous, and then, of course, buying back stock. Opportunistically."

Management also restated the hurdle it applies, which is the right hurdle and the one it appeared to be ignoring in Q2: "we only acquire assets that can accelerate our organic strategy while delivering financial returns that exceed what we could earn by buying back our own shares." The CFO applied it to ezCater specifically, arguing the build-versus-buy comparison implies "a few hundred million-dollar profit swing in the build case versus buy" against a business with over $2.5 billion of trailing gross bookings and average orders above $400.

Assessment: Our Q2 downgrade said the capital reallocation had no end date and that nothing justifying it would resolve inside a twelve-month horizon. An end date now exists, it is inside that horizon, and the CEO ranked buybacks above M&A for 2027. This is the pillar whose downgrade drove the rating cut, and it has been substantially repaired. The remaining gap is that the commitment is stated as an absence of future deals rather than as a dollar figure for repurchases, and Uber announced a $2.3 billion acquisition in October, four weeks before saying it would stop.

3. The buyback: $750 million, and two true ways to read it

Uber repurchased $750M of stock in Q3, up 47% from $510M in Q2 and down 49% from $1,463M a year ago, with roughly $15 billion remaining on the July 2025 authorization. The diluted share count fell 3.5% year over year. Which of those numbers matters depends entirely on the denominator you choose, and both readings are defensible.

On a year-to-date basis, repurchases of $4,279M against nine-month free cash flow of $7,613M are 56% of free cash flow. That is the roughly-half-of-free-cash-flow framework being met, not abandoned. On a current run-rate basis, $750M a quarter annualizes to about 29% of the $10.4 billion trailing free-cash-flow base. The nine-month figure is flattered by a $3,010M first quarter; the quarterly figure understates a program that management has said it is rebuilding.

Assessment: We said in August that the market was not wrong to want a number, and it still does not have one. But the framework test, which is the test management set for itself, passes on the nine-month arithmetic, and the trajectory is up rather than down for the first time since Q1. More importantly, the competing use of capital has been explicitly shut off for a year. The honest summary is that capital return has gone from deteriorating with no end in sight to recovering with the obstacle removed, which is a different situation even though the current quarter's dollar figure is small.

4. Autonomy moves from partnerships to commercial execution

For six quarters the autonomous narrative has been about the breadth of Uber's partner base. This quarter it became about vehicles actually carrying paying riders with nobody in the front seat, with a dated schedule attached.

"In AVs, our focus is shifting from partnerships and testing driverless commercial service. We are excited for Zoox to launch on Uber in Las Vegas later this month, with Motional moving up to fully driverless operations soon thereafter. We expect approximately 700 fully-driverless vehicles on Uber by year end, and more than 2,000 by the end of 27. Just as importantly, we are focused on demonstrating utilization and economics that are needed to scale." — Dara Khosrowshahi, CEO

The schedule in the prepared remarks is specific enough to grade against. Las Vegas becomes the first U.S. city where riders can reach two fully driverless partners through the Uber app. Nuro and Lucid begin fully driverless testing this quarter ahead of a Bay Area commercial launch in Q1 2027 that operates fully driverless from day one, with Houston following. Avride adds driverless deployments in 2027; NVIDIA begins supervised deployment in 2027 and Rivian in 2028. Internationally, WeRide, Baidu and Pony.ai already operate fully driverless, Wayve runs supervised commercial rides in London, and the UAE alone is expected to carry nearly 200 fully driverless cars by year-end.

Management also stated, for the first time, that the open platform does not mean an evenly funded one: it intends to maintain an open platform "while concentrating capital, integration resources, and hands-on deployment support behind a smaller group of partners with the clearest paths to commercial scale and attractive economics at maturity."

Assessment: The 15-cities-by-year-end target from the Q2 call was not reported against, and the disclosure metric changed from cities to vehicles. That is a legitimate criticism, and it is also the more useful metric, because a city with four cars and a city with four hundred are not the same thing. The substantive progress is real and the dates are near enough to be tested within the rating horizon. The deliberate re-concentration behind fewer partners is worth watching, since the multi-partner strategy was the answer to the single-partner dependence we flagged as a bear point in August.

5. The first real numbers on autonomous unit economics

The question that has sat unanswered since the $10 billion commitment was disclosed is what Uber gets for it. This quarter produced an answer in the form of a utilization comparison rather than a P&L figure, and the comparison is favorable enough to matter.

"Consider an illustrative 30% increase in paid trips per vehicle. Holding fixed costs constant, that improvement in utilization reduces fixed cost per trip by approximately 23%, creating room for higher fleet profitability even after paying Uber's marketplace fee. We are already seeing encouraging evidence of this utilization advantage. Today, we are driving ~30% higher trips per vehicle per day on average for our U.S. AV partners compared with first-party deployments nationwide." — Uber Q3 2026 prepared remarks

The CFO sharpened it on the call with the comparison that actually settles the argument, because it strips out the favorable mix in the first-party base: "what you are seeing on the ground today is exactly that. Both for Waymo and Tesla, their trips per vehicle per day in the new markets that they have launched this year is in the low teens. Versus high twenties that we are seeing on our network in these hybrid markets." He added two further sources of advantage, advantaged vehicle pricing from scale and shared charging, maintenance and fleet infrastructure across multiple partners, and framed the stated goal plainly: "Our goal is for AV trips to deliver margins comparable to human driven trips over time."

Assessment: This is the disclosure that moves the autonomy debate from faith to evidence. A roughly 100% utilization advantage in newly launched markets is a commercial moat argument that does not depend on anyone's view of self-driving technology, and the shared-infrastructure analogy to hyperscalers against on-premise data centers is the right framing of where the operating leverage sits. What is still missing is the number management promised in August: how much of the $10 billion lands above the operating line and how much below it. An illustrative fixed-cost calculation is not a P&L sizing, and this is the second consecutive quarter the sizing has been deferred.

6. Autonomy is not zero-sum, with evidence from the markets where it already runs

The structural bear case on Uber has always been that autonomy disintermediates the marketplace. The prepared remarks attacked that directly with data from the markets where autonomous vehicles have operated longest, and it is the most persuasive argument in the document.

"In U.S. markets where AVs operated throughout Q3, Uber's aggregate trips growth exceeded our overall U.S. growth rate, and those same markets grew faster than they did a year ago. More so, in mature U.S. AV markets, the ridesharing category itself has grown 7 percentage points faster than the rest of the country for the past several quarters. This continues to underscore our core belief: ridesharing is a supply-led business, and like all forms of new supply, AVs expand the addressable market." — Uber Q3 2026 prepared remarks

Two further details complete the argument. Uber's category position in those mature autonomous markets increased rather than eroded, and Mobility delivered its record 8.0% operating margin in the same quarter. The CFO's sparse-market statistic reframes the geography of the threat: autonomous deployments are concentrated in the densest top cities, and less than 10% of eligible consumers in U.S. sparse markets have used Uber in the past twelve months against 50% in dense markets.

Assessment: If this data holds up, it is the most important thing Uber disclosed this quarter, because it converts the autonomy question from existential to commercial. A category growing 7 points faster where autonomous supply has been added, with Uber's share rising inside it, is the empirical shape of a supply-led market rather than a disintermediated one. The caveat is sample: "mature U.S. AV markets" is a small set that management did not name, and the longest-running AV markets are also the markets where ridesharing penetration and consumer affluence are highest. One or two more quarters of the same pattern in newer markets would make this conclusive.

7. Who carries the autonomous fleet, and for how long

The capital-intensity question got its clearest answer yet, and the answer is that Uber's balance sheet goes first. The CFO described the intended end state as third-party financed, pointing to the $1 billion Santander arrangement set up earlier in the year for traditional fleet partners as the template, and said the near-term capital requirement is modest because the vehicle counts are small. The CEO then explained why the usual financing structures do not yet apply.

"And I think the residual value of the AV vehicles is not established yet. And so as a result, in the early days, we are gonna take these vehicles on our balance sheets. We have very strong balance sheet, are capable of doing so. And I think a lot of the fleet financing is gonna be based on revenue kind of revenue potential of these fleets. And what we are able to demonstrate is that cars on our network get very, very high utilization and actually higher utilization on cars that are not on our network." — Dara Khosrowshahi, CEO

The sequencing is explicit: limited capital against roughly 700 vehicles this year and 2,000 next, then third-party capital as deployments scale past ten thousand vehicles in 2028. Management says announcements on both infrastructure financing and vehicle financing should come in the next few months.

Assessment: This confirms the bear point we added in August rather than resolving it. A platform whose multiple rests on being asset-light has now said on the record that it will own vehicles until the residual-value market develops, and the honest version of that is that Uber is underwriting an unpriced residual risk on behalf of an industry that has not yet established one. The mitigation is genuine, because the utilization premium is exactly the collateral a revenue-based financing structure needs, and the dollar amounts are small until 2028. But the sequencing means the capital-intensity risk peaks after the point where management currently has visibility, and the financing structures that are supposed to absorb it have been described rather than announced for two consecutive quarters.

8. The capital structure got substantially larger, and the Q4 guide shows the cost

Uber issued €4.5 billion of euro-denominated senior notes on September 15, its first euro bond offering, in five tranches maturing between 2029 and 2046 with coupons from 3.75% to 5.25%, raising net proceeds of about $5.2 billion. It replaced its revolving credit facility on August 6 with a larger one, $7.7 billion of commitments against $5.0 billion before, and entered a €4.0 billion term loan that reduced the Delivery Hero bridge commitments by the same amount. Nothing is drawn on either the bridge or the term loan. The $2.0 billion 2026 term loan drawn in June was fully repaid in September.

The effect on the balance sheet is large. Aggregate debt principal stands at $15.7 billion against $12.6 billion at June 30 and $10.6 billion at December 31, up roughly 48% in nine months, while unrestricted cash and short-term investments rose to $10.1 billion from $5.4 billion at the end of Q2. Interest expense grew 36% to $152M, and that figure carries only two weeks of the new notes. Management reiterated an expectation of gross leverage below two times with an investment-grade balance sheet.

Assessment: This is the mechanism behind the Q4 guide's shape, and it is why Adjusted EBITDA guidance landed in line with consensus while non-GAAP EPS guidance came in up to 8% below the single-source $1.06 estimate. The euro notes add approximately $57 million of interest expense per quarter at a blended coupon near 4.45%, the non-GAAP tax rate has gone from 19% to 23%, and the share count is shrinking only 3.5%. None of those is an operating problem and all three are permanent. A $1.06 estimate is hard to reconcile with the company's own EBITDA translation and published cost-structure guidance; the company's $0.95 to $0.99 broadly is.

9. Delivery's monetization layer, and the competitor that just got bigger

Merchant spending across advertising and offers passed a $10 billion annualized run rate, combining ads that extend merchant reach with offers that subsidize consumer pricing. The CEO's framing is that these are merchant dollars funding consumer value: "these are dollars that we are giving to our consumers and the return for merchants is new consumers and kind of higher penetration in the marketplace." Grocery and Retail is the category absorbing most of the investment, at roughly 40% growth on a $15 billion run rate, with Costco expanded nationwide, 1,800 Tesco stores added in the UK, more than 375 Wakefern supermarkets in the U.S. Northeast, plus REWE in Germany and Decathlon in Australia.

Against that, the competitive structure of U.S. grocery delivery changed on October 29, when a $17 billion combination was agreed between two of Uber's largest competitors in the category. The CEO's response was to play down the direct effect and play up the merchant reaction.

"So this potential deal takes 1 of those partners off the table, and we have already had kind of reach-outs from our partners, you know, both who are very worried about the potential transaction and asking us, what do you think? and then 2, wanting to actually work with us more. Wanting to build their business with us so that they are not overly dependent on a single platform if DoorDash and Instacart happens, which will be a very, very large percentage of the online markets." — Dara Khosrowshahi, CEO

On first-party fulfillment, the answer was selective rather than dismissive. Management will not build dark stores broadly, but it acquires the capability through Delivery Hero, which has run quick commerce for seven to eight years, hit breakeven and is growing. It can also build on its Türkiye position, which now includes the Getir business, and an organic experiment in Taiwan.

Assessment: The merchant-choice argument is the right one and it is not merely rhetorical: a grocer facing a single dominant online channel has a strong incentive to fund a second. But a 40%-growing $15 billion business now competes with a combination that will hold a large share of U.S. online grocery, and the capability Uber is counting on to answer it arrives through a transaction that does not close until the second half of 2027. This is the one place where the Delivery Hero timeline, which we criticized in August for landing outside any useful horizon, is genuinely inconvenient rather than merely slow.

10. Agentic commerce, answered with four trips

Asked how third-party AI agents change the distribution of on-demand services, management gave the most quantitatively humbling answer of the call, and it is the right answer. The strategic position is that ridesharing and delivery differ from purely digital marketplaces because completing the transaction requires real-world execution, so Uber's coordination layer retains value regardless of where discovery happens. The commercial reality today is negligible.

"So I think just to put numbers to it, we had 1 of our largest weeks last week because it was Halloween. So we did just around 320 million trips on Uber and Uber Eats. The volumes coming from AI agents for Uber Eats were around 500. 500 orders total. And for our Rides business, it was a grand total of 4 trips that we have fulfilled from agents." — Balaji Krishnamurthy, CFO

The defensive moat is friction, measured. The median time to book a ride is about 50 seconds, 75% of rides are single-tap, and the median time for an Eats consumer to reorder a favorite is about 90 seconds. Against the disintermediation case, 85% of riders actively adjust their ride option before checkout and more than one in three Eats orders comes from a restaurant the consumer has not ordered from in the past year, both of which argue that visual discovery has value an agent cannot replicate. Uber's own AI products are showing early traction, with Cart Assistant baskets running roughly twice the size of an average Grocery and Retail order.

Assessment: Four rides out of 320 million in a peak week is the correct way to size a risk the market has been pricing as imminent, and management deserves credit for disclosing it rather than hiding behind a strategic answer. The negotiating position is strong for the structural reason the CEO identified: in most markets there are two large mobility and two large delivery providers against eight or more assistants, so the scarcity sits on Uber's side of the table. The honest risk is longer-dated and not addressed, which is what happens to the Uber One relationship if agents eventually aggregate demand at scale.

11. Uber One returns to the disclosure set

Approximately 55 million members, up from 50 million disclosed at Q1, with members accounting for more than half of combined Mobility and Delivery gross bookings and 70% of Delivery gross bookings specifically. Management flagged that Mobility penetration is the opportunity, and that richer rewards scaling with ride and order frequency arrive in the coming months.

Assessment: We flagged the disappearance of this metric in August as the kind of substitution worth noticing even when the trend underneath is fine. It is back, the trend was fine, and the 70%-of-Delivery figure, up from roughly two-thirds a year ago, is more useful than the headline count, because it quantifies how much of the highest-growth segment now sits behind a subscription relationship. That is also the asset most exposed to agentic intermediation, which makes the two topics the same topic.

Guidance & Outlook

Implied sequential ramp. The Q4 bookings midpoint of $64.25B is 7.4% above the Q3 actual of $59,819M. The comparable seasonal step a year ago was 8.8%, and Uber delivered 3.1% sequentially in Q3. So the guide asks for a sharper sequential step than Q3 produced and a slightly softer one than Q4 2025 delivered, which is consistent with the one-point currency headwind and the 100bps Middle East assumption sitting inside it. Holding the constant-currency band at 18% to 22% on a base that is $10 billion larger, in the quarter immediately after a supply shortfall, is a more confident guide than it looks.

Where consensus sits. Bookings and Adjusted EBITDA guidance are both essentially at consensus, at midpoints of $64.25B against $64.4B and $3.18B against $3.19B. Non-GAAP EPS is the exception: $0.95 to $0.99 against a $1.06 consensus that we could source only once, a shortfall of roughly 7c even at the top of the range. A second, independent estimate of $0.97 sits at the guide midpoint, so the size of the gap is uncertain. Measured against $1.06, a guide that is in line on operating profit and up to 8% light on earnings per share is not an operating warning. It would mean the estimate had not modelled a €4.5 billion bond issue, a 23% tax rate and a share count shrinking only 3.5%.

The guide reconciles and the consensus did not. Uber's non-GAAP operating income includes stock-based compensation, so it bridges from Adjusted EBITDA by deducting stock compensation and non-GAAP depreciation. Applying that bridge to the Q4 EBITDA midpoint using management's own full-year color, stock compensation of $2.00B to $2.10B and non-GAAP depreciation of $500M to $550M against nine-month actuals, implies roughly $499M and $143M respectively in Q4, and therefore non-GAAP operating income near $2.54B. Carry that through interest of roughly $40M net, a 22% to 24% tax rate and about 2.04 billion diluted shares, and you arrive at $0.93 to $0.96. The company's range overlaps the top of that one. A $1.06 estimate requires a materially different cost structure than the one Uber has published.

Guidance style. Uber has now held the same 18% to 22% constant-currency bookings band for three consecutive quarters while the base grew roughly 20%, and it landed inside it each time, twice above the midpoint. The EPS band has less historical slack because it is absorbing a rising tax rate, rising interest expense and a share count that no longer shrinks at the Q1 pace. On the established pattern the bookings number should land at or above $64.5B.

What is not in the guide. No full-year framework was restated. No dollar figure was attached to the buyback. No income-statement sizing was given for the autonomy program, for the second consecutive quarter. No financial contribution was assigned to Delivery Hero or ezCater, which is correct because neither has closed, and no acceptance-level update was given on Delivery Hero. Full-year color was supplied where it matters least and withheld where it matters most: stock compensation narrowed to $2.00B to $2.10B, non-GAAP depreciation narrowed to $500M to $550M, the full-year non-GAAP tax rate set at 22% to 24%, and the cash tax rate guided to the mid-to-high single digits this year and 10% to 15% over the next few years as historical tax assets are used.

Analyst Q&A Highlights

Nine analysts asked questions on a Q&A-only call, with prepared remarks published separately. The distribution tells you what the Street cared about: autonomy came up in four of the nine, Delivery competition in two, capital allocation in two, and only one analyst asked directly about the Mobility supply shortfall that was the quarter's actual news. Nobody asked about the $186M restructuring charge, the $484M legal and regulatory reserve, or the Australian court ruling that landed the day before the print.

Autonomous fleet financing and whose balance sheet carries the vehicles

The opening question went straight to the uncomfortable part of the autonomy program: how much balance sheet and free cash flow stands behind it, and how the returns are being underwritten. The answer separated the long-run intent from the near-term reality, and the near-term reality is that Uber owns the cars. The reason given is that autonomous vehicles have no established residual value, so conventional fleet financing, which prices residuals, does not yet apply. The proposed substitute is financing against revenue potential, with Uber's utilization premium as the collateral.

Q: "I want to come back to the parts of the letter and prepared remarks around EV fleet financing. Can you go a little bit deeper in terms of how we should be thinking about the balance sheet and incremental free cash flows that you want to put behind these efforts? And how should we be thinking about rates of return or aligning the economics with the commercial opportunity?"
— Eric Sheridan, Goldman Sachs

A: "So our long run expectation is that this should be financed by third party capital, not by Uber's balance sheet. And in the near term, as you have seen with the announcement, that we had in the release that people have about 700 driverless vehicles on our network this year. 2,000 by the end of next year. The capital deployment against those is fairly limited. But as we start scaling out into the 10 thousand+ in 2028 and beyond, we want to bring in third party capital again."
— Balaji Krishnamurthy, CFO

Assessment: Management answered the question honestly and the honest answer is unwelcome. The capital requirement is small while the fleet is small, and the structures meant to absorb it at scale are being worked on rather than announced, for the second consecutive quarter. What makes this better than a dodge is that the collateral argument is specific and testable: if the utilization premium holds, revenue-based fleet financing is a real product and the asset-light model survives. If it does not, Uber owns depreciating cars.

What changed in international driver supply, and how persistent it is

The most important exchange of the call, and the one that pressed hardest. The questioner noted that last quarter's problem was described as Brazil-specific and this quarter's spans several markets, and asked directly whether the causes, immigration among them, are structural. The answer conceded the broadening and separated the drivers: fuel costs in price-sensitive emerging markets, immigration flows in Canada, and seasonal peak demand landing on a marketplace already short of supply. The offered evidence of recovery was an October data point rather than a forecast.

Q: "I wanted to follow-up on the international supply dynamic you called out. Just want to make sure, did something sort of change through the quarter? I think last time we had talked about Brazil. But it seems like there is a few more markets that are now reflecting this. So if you could just elaborate on what you saw through the quarter, would be helpful. Then also, as we look forward, should we be thinking about the supply side investments required Because some of the variables called out like immigration feel a little bit more persistent."
— Nikhil Devnani, Bernstein

A: "Now, this is a good problem where we are at the scale that we are at. Operating at, and we still are running into supply constraints rather than demand constraints. And we know exactly how to address those kinds of challenges with the history that we have demonstrated We are making targeted supply investments. So this is focused on driver acquisition, retention, and engagement. And as we go through that kind of investment, we are already encouraged with the early signs of supply that we saw in October, and supply hours growth globally accelerated in October. Including notable improvements in Brazil and India."
— Balaji Krishnamurthy, CFO

Assessment: The persistence question was not really answered. A claim to know "exactly how to address those kinds of challenges" is a claim about capability, and incentive dollars do not change the price of fuel or a country's immigration policy. What was answered, and answered well, is the direction of travel: supply hours re-accelerated globally in October with the two worst markets improving, which is the right leading indicator and one that will be visible in the Q4 print. Management also supplied the mechanism later in the call, describing a flexible incentive pool reallocated from riders to drivers by algorithm, which is a genuine structural advantage over a smaller competitor facing the same input costs.

Whether the record Mobility margin survives the supply investment

A short, pointed question about whether the 8% Mobility margin is a level or a peak, given that the fix for the supply problem is spending money in the supply side of the marketplace. The answer drew the distinction that matters for modelling: the supply investment will show up in the take rate, not in the margin, because the dollars are incentives netted against revenue rather than operating costs.

Q: "I just wanna ask 1 question on the profit profitability of the mobility segment, these 8% margins you had this last quarter. With the supply initiatives that you are introducing in international markets, does that mean you can maintain the profitability performance that you have had? Or is it-- do we have a detour? Is it structural, or is it a temporary detour?"
— Mark Mahaney, Evercore

A: "I would just take you back to our philosophy as always, which is that we want to drive margin expansion for the total company on a year-on-year basis. And, ideally, we want to do the same for the segments as well. But on a quarterly basis, there is some variability in the business both from you know, just the investments we typically make in supply in holiday season, etcetera. So I do not. I would not call out any kind of meaningful changes in the margin trajectory for the business, and, you know, we would have some color on just the Q4 supply investments for mobility that would have an impact on the take rate, but not necessarily from a margin standpoint, anything meaningful."
— Balaji Krishnamurthy, CFO

Assessment: This is the most model-relevant answer of the call and it is easy to miss. Mobility take rate should fall further in Q4 by design, and reading that as deterioration would be a mistake for the second consecutive quarter. The commitment is to year-over-year company margin expansion with quarterly variability permitted, which is the right framing and a slightly softer one than a promise of a record margin every quarter. The CFO then pivoted unprompted to the per-share line and the 37% Q4 midpoint growth rate, which is where management clearly wanted the conversation, and on this quarter's evidence that preference is justified.

How durable the utilization advantage is

The sharpest analytical question of the call: the illustrative 30% utilization uplift is attractive, but is it durable, or does it compress as first-party operators mature? The answer argued the opposite, that the comparison currently understates Uber's advantage because the first-party base is flattered by San Francisco, and that the gap widens in markets launched this year.

Q: "So in the illustrative example that you gave, 30% increase in paid trips, with a 23% decrease in fixed costs. And even if partners were to pay the transaction fee or take rate to Uber, that would still be that would still be additive to them. So my question is, how durable is that 30% increase in paid trips that you would add? How durable is that advantage as you think long term in AV?"
— Shweta Khajuria, Wolfe Research

A: "And these mature 1P deployments have the advantage of mix in that they include San Francisco, which is a uniquely favorable for AV deployments today. As that deployment scale broadens out, we expect that 1P deployments will likely have utilization that would be worse than what they are seeing in San Francisco. And in fact, what you are seeing on the ground today is exactly that. Both for Waymo and Tesla, their trips per vehicle per day in the new markets that they have launched this year is in the low teens. Versus high twenties that we are seeing on our network in these hybrid markets."
— Balaji Krishnamurthy, CFO

Assessment: This is the strongest single argument Uber has made for its position in the autonomous value chain, because it is a like-for-like comparison in newly opened markets rather than a blended average. High twenties against low teens is roughly a doubling, and it arises from exactly the asset Uber cannot be stripped of: pre-existing demand density across the whole day. The caveat is that the figures are Uber's, drawn from its own network and from external observation of competitors, and no independent source publishes them. Treat the direction as well supported and the magnitude as management's.

The geographic pace of the driverless rollout

A straightforward schedule question, answered with an unusually long and specific partner-by-partner calendar running into 2028. The notable feature of the answer is the argument for why the ramp should be fast once vehicles exist: Uber is already operating in these cities, so no demand has to be built.

Q: "I want to ask a little bit more on the fully driverless rollout from 700 this year to 2,000 next year. Can you just help us understand a little bit on the geography pace of rollout? I think you said 50/50 in the U.S. and international and how you are thinking about that."
— Ronald Josey, Citi

A: "So you know, they. The partner strategy and the panoply of all these partners are really going to start adding up, and we are quite confident of getting to 2,000 vehicles in 2027, but the real goal is to get to tens of thousands of vehicles in 2028 and beyond. And really scaling the ecosystem. And, you know, we are in market in these markets. We have relationships with the cities. We have operations in those markets. We do not have to build demand. The demand is already there. So we think we can launch very quickly at very high utilizations for our partners which makes us even more optimistic about the opportunity here."
— Dara Khosrowshahi, CEO

Assessment: The partner list is now long enough that no single supplier failure is thesis-relevant, which resolves the dependence risk we flagged in August. The observation that Uber does "not have to build demand" is the whole commercial argument compressed into one sentence, and it is correct. What the answer does not address is the tension with the prepared remarks' stated intention to concentrate capital and deployment support behind a smaller group of partners. Breadth is the risk mitigant and concentration is the economic strategy, and the two pull against each other.

Grocery investment and the case for first-party dark stores

A question about whether competing in U.S. grocery and retail eventually requires Uber to own fulfillment. The answer was a clear no as a general matter and a clear yes in selected markets, with the capability arriving through acquisition rather than being built.

Q: "Maybe 1 on grocery. Just talk to us a little bit philosophically about the investments you need to make in grocery and retail to sort of further accelerate and compete in that business in The U. S? How do you think about the need to have a larger 1P presence? So, you know, your own dark stores or your own sort of fulfillment operation."
— Brian Nowak, Morgan Stanley

A: "Now we are very excited as it relates to Delivery Hero. They have built a substantial quick commerce and dark store business. And the benefit of dark stores is that you have a much better handle on inventory Usually, your pick-and-pack success percentages. Your found rates are much higher, so quality goes up. As a result of the vertical integration. And, you know, this is a part of the delivery hero hopeful potential acquisition that we are very excited about. That team is been building the quick commerce business for you know, well over, you know, 7 to 8 years. It has hit breakeven. It is growing at very attractive rates, and we are very excited to add that as part of our portfolio once we close on Delivery Hero."
— Dara Khosrowshahi, CEO

Assessment: Management called this a "fit-for-purpose sort of solution," which is the right posture: dark stores have genuinely better found rates and genuinely worse returns on assets, so selectivity is the answer. The problem is timing rather than strategy. The capability is inside a transaction that closes in the second half of 2027 and migrates in 2029, while the competitive structure of U.S. online grocery consolidated in October 2026. Uber is answering a 2026 problem with a 2028 asset.

The DoorDash and Instacart combination

The competitive question the market most wanted addressed, since the deal was the proximate cause of the stock's slide toward its 52-week low in the week before the print. The answer leaned on merchant behavior rather than on market share, arguing that a single dominant online channel creates its own counterweight.

Q: "I wanted to maybe talk about the potential combination of DoorDash and Instacart. Uber has talked a lot about the cross-platform you have experienced from your Eats to your mobility business. So I was wondering if you could talk about the potential opportunity and or risk that you see created by this combination."
— Michael Morton, MoffettNathanson

A: "You know, when we talk to our merchant partners, they want choice. They wanna develop direct channels They want plenty of partners including us and DoorDash and Instacart to work with. So this potential deal takes 1 of those partners off the table, and we have already had kind of reach-outs from our partners, you know, both who are very worried about the potential transaction and asking us, what do you think? and then 2, wanting to actually work with us more."
— Dara Khosrowshahi, CEO

Assessment: The merchant-counterweight argument is real and the early evidence, inbound interest from worried retailers, is the most useful thing in the answer. But the response opened by saying the deal is not "gonna affect us too much," which is a stronger claim than the facts support for a category Uber is growing at 40% and now contests against a combination that will hold a large share of U.S. online grocery. The closing line, that it will be up to the government to decide whether one alternative or two is better, is the most honest sentence in the exchange, and it points at the real variable: the deal's regulatory path.

U.S. Mobility resilience and the drivers of the organic Delivery acceleration

A two-part question seeking the U.S. versus international split on Mobility, with specific interest in markets carrying heavy autonomous penetration, and the restaurant-versus-grocery composition of the Delivery acceleration. The Mobility half produced the clearest statement that the shortfall was not a U.S. event, alongside the concession that U.S. growth moderated.

Q: "First, could you talk about domestic or U.S. mobility trends, maybe what you saw in 3Q and in 4Q, you talked about improving mobility growth. But I want to see if we can touch a little bit upon what you are seeing in The U. S. Versus some of the international markets you have talked about. And in particular, I think there is a market focus on continued on California, maybe even in some markets in Texas where there is been heavier AV penetration."
— Ken Gawrelski, Wells Fargo

A: "On the U.S. for mobility, we are seeing relatively resilient growth in The US, versus the international trends we talked about earlier in the call. You know, growth moderated a bit in Q3, but not as much as in the international markets. You know, elevated gas prices are an issue globally, and there is some level of supply pressure that we are seeing in The US as well, but we are responding with the targeted investments, supply investments that we talked about. Earlier. I think beyond the supply question, what we are seeing in The US that continues to be true is the insurance tailwinds continue to build. We talked about the meaningful acceleration in California trip growth."
— Balaji Krishnamurthy, CFO

Assessment: Two things are confirmed here and one is new. Confirmed: the insurance mechanism underpinning our first bull pillar is still working, and California, the market where the reinvestment is heaviest, is accelerating and outpacing the country. New: there is "some level of supply pressure" in the U.S. as well, which was not in the prepared remarks and makes the problem less neatly international than the written disclosure suggests. The question's premise about autonomous penetration in California and Texas went unaddressed in the answer and was handled instead in the prepared remarks, where the category-growth data is considerably more favorable than a dodge would imply.

AI agents, and the volume they actually produce

The closing question asked how building agent technology into the platform differentiates Uber, and how to think about distribution through third-party assistants. The strategic answer was about owning the direct relationship; the useful answer was a number that arrived only after a follow-up prompt.

Q: "Hoping to get your latest thoughts on AI agents Can you just talk about some of the goals and differentiation that you can provide by building agent tech directly into the Uber platform?"
— Douglas Anmuth, JPMorgan

A: "And the good news here is in most markets, you have 2 big mobility providers, 2 big delivery providers. You know, I think I have counted now. there are 8 different assistants, and I think there will be more than 10 assistants. So I think, you know, the content players in our space in the physical space, we are kind of the fulfillment. The demands on the services that we build are very large. I think the demands on our services are to be very, very high. So something we are really excited about now it is pretty small."
— Dara Khosrowshahi, CEO

Assessment: The bargaining-power observation is the substance: two providers per market against eight or more assistants puts the scarcity on Uber's side, and the terms of trade are being negotiated from that position. The volume disclosure that followed, 500 Eats orders and four rides out of roughly 320 million trips in a peak week, is the most deflationary data point available on a risk the market has been pricing as near-term, and management volunteered it. What remains unaddressed is the longer-dated question of what agentic aggregation does to a subscription relationship that now carries 70% of Delivery bookings.

What They're NOT Saying

1. A $186 million restructuring charge appeared with no explanation. Restructuring and related charges ran $186M in the quarter against $3M a year ago, and a further $50M was booked for the settlement of legacy awards. Neither was mentioned in the prepared remarks or on the call, and nobody asked. Corporate overhead fell to 1.81% of bookings in the same quarter, so the cost of achieving that leverage is visible in the reconciliation and absent from the narrative.
2. A $484 million legal and regulatory reserve, almost all of the year's total. The nine-month figure for legal, non-income tax and regulatory reserve changes is $496M, of which $484M landed in Q3. The same line carried $479M in Q3 2025, so a large third-quarter charge has precedent, but no cause was named on the call or in the prepared remarks. The 10-Q does say the accrual for the Australian payroll-tax matter below is recorded as of September 30.
3. An adverse High Court ruling in Australia the day before the print. The 10-Q discloses that on November 3, 2026, the High Court of Australia ruled against Uber on New South Wales payroll-tax assessments covering July 2014 to June 2020, and that the company then evaluated the impact on other Australian states and territories given "the unprecedented nature of this ruling." An accrual is recorded as of September 30, but no amount is disclosed, and the matter was not raised on a call held the following morning.
4. No income-statement sizing for the autonomy program, for a second consecutive quarter. In August management said it would "size that for investors clearly" as deployment scaled. This quarter supplied vehicle counts, a utilization premium and an illustrative fixed-cost calculation. The question of how much of the $10 billion lands above the operating line remains unanswered, and it is now more pressing because the CEO has said the vehicles go on Uber's balance sheet first.
5. No dollar figure for the buyback, and no restated framework. The nine-month payout ratio happens to satisfy the roughly-half-of-free-cash-flow framework, but the company neither restated the framework nor committed to a forward amount, in a quarter where it told investors M&A is closed through 2027. That is the quarter to give the number.
6. The 15-autonomous-cities target was quietly retired. Q2's target was 15 live cities by year-end 2026, with 7 live and 8 named. This quarter reports vehicles instead of cities. The new metric is more informative and the substitution was not acknowledged.
7. The advertising run rate changed basis without a bridge. Q2 disclosed advertising past a $2 billion annualized run rate. Q3 discloses merchant spending across advertising and offers past $10 billion. The second is not a growth rate on the first, and the only comparable number is the 10-Q's $176M year-over-year advertising revenue increase in Delivery.
8. No Delivery Hero acceptance update. The minimum acceptance threshold of 50% plus one share was the gating item in August. Four months on, the timetable is restated and the acceptance level is not. Uber's equity-method carrying value of $3,372M also sits above the $3.1 billion market value of the 24.5% stake, which is disclosed in the 10-Q and was not discussed.
9. Free cash flow guidance is absent in the quarter it would matter most. Trailing-twelve-month free cash flow is a record $10.4 billion, and Q4 absorbs an expected $1.5 billion payment on disputed UK VAT assessments. On our estimate, full-year 2026 reported free cash flow lands near $9.1 billion to $9.2 billion, below 2025's $9,763M. Management flagged the payment and encouraged investors to evaluate free cash flow annually, without noting that the annual figure will decline.
10. No comment on interest expense trajectory. Aggregate debt principal rose roughly 48% in nine months to $15.7 billion, €4.5 billion of notes were issued seven weeks before the call at coupons up to 5.25%, the revolver was enlarged by 54%, and interest expense already grew 36% year over year carrying only two weeks of the new paper. It is the single largest driver of the gap between the Q4 EBITDA guide and the Q4 EPS guide, and nobody raised it.

Market Reaction

• Pre-print setup: UBER closed at $66.60 on November 3, down 18.5% year to date and 29.7% over twelve months, against an S&P 500 up 10.6% year to date. The stock was 4.1% lower over the trailing 30 days and sat roughly 2% above the bottom of a 52-week closing range of $65.25 to $99.72. The final leg of that de-rating came in the week before the print, when two of Uber's largest competitors in U.S. grocery delivery agreed a roughly $17 billion combination on October 29 after Uber's own talks with the target had lapsed. This was a maximally de-rated setup.
• Reaction session (November 4, before the open): gapped up 2.5% to open at $68.28, traded as high as $72.50 (+8.9% from the pre-print close), and closed at $69.87, up 4.9%. Volume of 37.0 million shares was 2.3x the 30-day average. The S&P was up 1.6% on the session, so roughly 3.3 points of the move was stock-specific.
• Peers: the closest ride-hailing comparable closed up 12.4% and the closest delivery comparable up 8.3% on the same session, both larger moves than Uber's own. The read-across split cleanly by segment, with Uber's Mobility shortfall taken as relief for its ride-hailing competitor and its Delivery strength taken as confirmation for the delivery group.
• Day after (November 5): UBER closed at $68.78, down 1.6%, on a session in which the S&P fell 4.4%, its worst day since the April 2025 tariff shock. This was a market-wide risk-off session rather than a delayed verdict on the quarter, and Uber outperformed the index by 2.8 points inside it.

The reaction was a relief rally against a very low bar, and the shape of it is more informative than the magnitude. A company that preannounced nothing, missed its own Mobility plan by 1 to 2 points, and guided earnings per share below at least one consensus estimate should not add 4.9% on a day the market rose 1.6%. It did so because the pre-print price had already discounted a worse quarter than the one that arrived, and because the two things the market had stopped believing in, capital discipline and autonomous commercialization, both produced evidence rather than framework.

That the stock gave back roughly half of an 8.9% intraday gain into the close is the market working through the Q4 earnings guide, which is the one genuinely disappointing number in the package and the one that required the most work to interpret. The sell-side response captures the ambiguity precisely: targets were cut almost without exception, in a $70 to $100 band, and not a single rating moved in either direction. That is the signature of a Street that believes the business and has to lower its arithmetic, which is what happens when a company issues €4.5 billion of debt and slows its buyback inside the same two quarters.

The November 5 session deserves separate weight, because it was an unusually clean natural experiment. On a day the index lost 4.4% on concerns about capital-intensive artificial-intelligence buildout, Uber lost 1.6%, about a third as much. A business trading near 20x current-year earnings with no hyperscale capital exposure, growing per-share earnings close to 40%, behaved like a defensive holding during the worst session in eighteen months. For a stock whose twelve-month de-rating is the central fact of the investment case, that relative resilience is the first evidence we have seen that the de-rating has found a floor.

Street Perspective

Debate: Is the Mobility shortfall cyclical or structural?

Bull view: The bull case being made on the Street is that a supply constraint in a marketplace with healthy demand is the most recoverable problem a platform can have, that Uber has a long record of solving it with reallocated incentive dollars, and that the evidence is already in hand with supply hours re-accelerating globally in October and the two worst markets improving. The unchanged constant-currency guidance band on a 20% larger base is offered as management's own confidence vote, and the record Mobility margin earned in the same quarter is offered as proof the shortfall was not bought off.

Bear view: The bear camp contends that two of the three named causes, fuel prices and immigration policy, are entirely outside Uber's control and will not respond to incentive spend, and that a problem described as Brazil-specific one quarter and as spanning India, Brazil and Canada the next is broadening rather than resolving. The concession that there is "some level of supply pressure" in the United States too, which did not appear in the written disclosure, is read as the problem being less neatly contained than presented. On this reading the fix is a permanent increase in the cost of supply, visible first in the take rate and eventually in the margin.

Our take: The bulls have the better of this on evidence and the bears have the better of it on causation. October supply hours are a real leading indicator and the unchanged guidance band is a real commitment, so the near-term recovery is well supported. But fuel and immigration are genuinely exogenous, and the honest position is that Uber can out-compete for a shrinking pool of drivers without being able to grow the pool. That is a share argument rather than a growth argument, and it is why we treat this as the quarter's genuine negative even while rating the stock higher.

Debate: Does the autonomy program create value or consume it?

Bull view: A growing consensus view holds that this quarter settled the question. A roughly 30% utilization premium against first-party deployments nationwide, high-twenties trips per vehicle against low-teens in markets opened this year, a 7-point category-growth premium in mature autonomous markets with Uber's share rising inside it, and a record Mobility margin in the same quarter, are collectively the empirical case that autonomy expands Uber's market rather than disintermediating it. On this view Uber is acquiring the commercialization layer of an entire industry for less than one year of free cash flow.

Bear view: The skeptics answer that nothing about the capital question improved and one thing got worse. The chief executive said on the record that Uber will carry autonomous vehicles on its own balance sheet until residual values are established, which is fleet ownership in a business whose entire multiple rests on being asset-light. The income-statement sizing promised in August was deferred again. The third-party financing structures meant to absorb the risk have been described for two consecutive quarters and announced in neither. And the capital requirement scales into 2028, beyond management's stated visibility.

Our take: Both sides are right about different halves, and this quarter moved the commercial half decisively while leaving the financial half where it was. The utilization data is the most important disclosure in the package because it is a like-for-like comparison in new markets rather than a blended average, and it supports a durable advantage that does not require any view on whose software wins. We would still treat the $10 billion as a real cost against the option value until a financing structure is announced rather than described, which is why this supports an upgrade rather than a higher-conviction rating.

Debate: Has capital allocation been fixed, or just paused?

Bull view: Some desks argue capital allocation has been fixed rather than paused. The acquisition window is explicitly closed through at least 2027, buybacks were named above M&A in the 2027 priority stack, nine-month repurchases run at 56% of nine-month free cash flow, the quarterly figure rose 47% sequentially, and management restated the correct hurdle that an acquisition must beat buying back its own stock. The per-share consequence is already in the guide, with earnings growth re-accelerating to 37% at the Q4 midpoint.

Bear view: The bear camp notes that a company announcing a $2.3 billion acquisition on October 5 and declaring an M&A pause on November 4 has demonstrated how much the pause is worth. The buyback is $750 million against a $10.4 billion free-cash-flow base, there is still no forward dollar commitment, aggregate debt principal rose roughly 48% in nine months, Q4 gives up $1.5 billion of cash to a disputed tax assessment, and the €14.2 billion of Delivery Hero financing has not yet been drawn. On this view capital return is a residual claim behind two pending acquisitions and an autonomous fleet.

Our take: We are closer to the bull case than we were in August and not all the way there. The decisive change is not the $750 million; it is that the competing use of capital now has an end date inside our rating horizon, which is precisely the resolution our downgrade said would not arrive. The remaining complaint is a disclosure complaint rather than an allocation one, and the nine-month arithmetic shows the framework being honoured. We would still like the number, and we would be more comfortable if the pause had not arrived four weeks after a $2.3 billion acquisition.

Model Update Needed

Valuation impact: The operating changes are a small net positive and the capital changes are now a net positive too, which is the reverse of last quarter. Lower Mobility growth takes something out of the bookings path; higher margin on bookings, a lower tax rate and a materially larger buyback put more back into the per-share line. Our full-year 2026 non-GAAP earnings estimate, built on the Q4 guide midpoint, is approximately $3.37 against $2.45 in 2025, and at the reaction close of $69.87 the shares trade near 20.7x that figure. For a platform compounding bookings above 20% with record and still-expanding margins, record free cash flow, and an autonomous option whose commercial evidence turned positive this quarter, that is a low multiple, though not the lowest of the quarter: the pre-print close was 19.8x the same estimate. The net effect is a higher fair-value range and a narrower one, because one of the two previously unsized capital programs now has a stated end date.

Thesis Scorecard Post-Earnings

Overall: the thesis is weaker on operations and materially stronger on structure, which is the exact inverse of last quarter, and structure is what the rating was priced on. Bull 4, the pillar whose downgrade drove the August cut, returns to ON TRACK. Bear 2 de-escalates. Bull 5's commercial evidence arrived even though its capital question did not. Against that, Bull 1 moves to AT RISK on a broken commitment, Bear 4 materializes beyond the market we were watching, Bear 3 is now confirmed by the chief executive rather than inferred by us, and a new competitive risk appeared in the week before the print. Three pillars improved and three deteriorated, and the improvement landed squarely on the capital-allocation pillar whose downgrade drove the August cut.

Action: Upgrading to Outperform from Hold. Our coverage arc runs Initiated Hold at Q1 2025, upgraded to Outperform at Q2 2025, maintained through Q3 2025, Q4 2025 and Q1 2026, downgraded to Hold at Q2 2026, and now returns to Outperform. The August downgrade rested on one sentence: an operating story was being funded by a capital-return story that had been redirected into commitments with no end date, and nothing justifying that redirection would resolve inside twelve months. Management has now put the end date inside our horizon, the nine-month payout satisfies the framework it set itself, and per-share growth is guided to re-accelerate to 37% rather than decelerate to the low thirties. The quarter's genuine negative, an international driver-supply shortfall, is the more fixable of the two problems a marketplace can have, and the leading indicator management handed us for it is already positive. We are being paid to accept that risk: the stock is 29.7% lower over twelve months, trades near 20.7x our estimate of this year's earnings, and on the worst index session in eighteen months fell about a third as much as the market. Watch items for the next 90 days: whether international supply growth converts October's improvement into a Q4 Mobility number inside the guided band; whether the buyback finally gets a dollar figure now that M&A is closed; whether an autonomous fleet-financing structure is announced rather than described; whether Canada is still being named as undersupplied; and what the $186M restructuring charge and $484M reserve were actually for.