Audience and Frequency at All-Time Highs, $20B Buyback Reframes Capital Return, AV Ecosystem Compounds — Upgrading to Outperform

Rating Action

We initiated coverage at Hold in Q1 2025 on a name we acknowledged was operationally excellent but where three structural questions kept us from underwriting the bull case: (i) whether Uber's autonomy strategy would prove durable as Waymo and others contemplated direct-to-consumer alternatives, (ii) whether the membership flywheel was over-indexed to Delivery and would stall in Mobility, and (iii) whether the cash flow generation would actually translate to per-share value via buybacks rather than be diluted by SBC and incremental investment cycles.

Q2 2025 resolved each of these, and did so with specificity. On autonomy, the partnerships announced this quarter (Baidu, Lucid, Nuro, Wayve) plus the existing Waymo/WeRide/Avride/Volkswagen/Pony/Momenta footprint position Uber as the aggregator across hardware, software, and geographic stacks. Khosrowshahi's framing — "we are a supply-led company" with humans and AVs treated as the same supply pool — is the cleanest articulation yet of why the network model wins regardless of which AV stack ultimately scales. The 99th-percentile Waymo utilization on Uber's network is the empirical proof point: AVs deployed on Uber complete more trips per day than the top 1% of human drivers, and the consumer halo effect (people excited to try AV trips boosting overall system frequency) is showing up cleanly in Austin.

On membership, Uber One's growth to 36M (+60% YoY) is now explicitly being extended into Mobility via Surge Savings — a product Khosrowshahi described as "the #1 product that our Mobility consumers have asked for." This addresses the bear-case argument that membership penetration in the higher-frequency Mobility business has trailed Delivery. Management acknowledged Mobility is still "not nearly as optimized" as Delivery (~80% of where it ultimately lands), but the product roadmap is explicit and the run-rate is already +60%.

On capital return, the $20B authorization combined with the ~$3B remaining from the 2024 program is the largest commitment in Uber's history at ~12% of market cap. CFO Mahendra-Rajah's explicit "50% of free cash flow to buybacks on a multi-year basis" framing converts a rhetorical commitment into a programmatic one, and the -1% YoY share count reduction in Q2 is the first quarter the engine produced visible per-share value.

Layered on top, the operating print delivered the strongest growth signal of the past year (+18% trips, +18% gross bookings, all-time highs in audience and frequency, record adjusted EBITDA/GAAP OI/FCF), and the Q3 guide of low- to mid-30s EBITDA growth on high-teens bookings growth confirms operating leverage is widening, not compressing. The risk-reward has shifted clearly favorable, and we move to Outperform from Hold.

Results vs. Consensus

Uber delivered a quality print across nearly every line. Trips and gross bookings both grew 18% YoY — symmetric growth that signals neither take-rate compression nor pricing-driven inflation; rather, broad-based volume expansion. Audience (Monthly Active Platform Consumers) grew 15% YoY with +10M sequential adds, and frequency hit an all-time high. Adjusted EBITDA, GAAP operating income, and free cash flow were all record quarters.

Quality of Beat

• Volume vs. Pricing: Trips +18% matched gross bookings +18% — meaning the growth is volume, not price. In the U.S., management is explicitly passing through insurance-cost savings to consumers (lower fares), and profit per ride is still up YoY. The growth is not bought with margin.
• Audience expansion: +10M MAPC sequential add is a step-change from trend. Khosrowshahi attributed it to barbell strategy — low-cost products (Moto +40% to $1.5B, Reserve growing 60%) and premium products (now $10B+, +35%) reaching demographics outside the core 25–45 urban user. In top-10 markets, only ~20% of the 18+ population uses Uber monthly — the runway is meaningful.
• Operating leverage: Q3 guide of low- to mid-30s adjusted EBITDA growth on high-teens gross bookings growth implies sustained EBITDA-to-bookings spread of ~15–17pp — a level that bear case framing assumed would compress as growth moderated. It is not compressing.
• Cash flow: Record FCF combined with $20B incremental buyback authorization signals management has internal visibility on multi-year FCF that comfortably funds capital return AND incremental AV investment AND opportunistic M&A (Trendyol Go was cited as a recent example).

Segment Performance

Mobility — Pricing Pass-Through Driving Volume Reacceleration

Mobility's Q2 was anchored on the U.S. insurance-cost dynamic: insurance increases that had pressured pricing and demand for several quarters are now decelerating, and the savings are being passed to consumers. Profit per ride is still up YoY (the savings are insurance-funded, not margin-funded), and the lower price points are driving session-conversion improvements PLUS delayed return-frequency benefits.

"What we're seeing in the U.S. is that there are some session benefits or in-session benefit, but there's a delayed benefit to consumers then coming back with the app more and that's absolutely showing up in July, where transaction growth accelerated nicely versus Q2." — Dara Khosrowshahi, CEO

July transaction growth accelerating versus Q2 is the most actionable forward datapoint of the call — a within-quarter signal that Q3 Mobility growth is building, not stabilizing. Khosrowshahi expects "the same trends are going to show up in the balance of the quarter" with continued acceleration into Q4.

The barbell-product strategy is the medium-term growth engine. Premium (now $10B+ gross bookings, +35% YoY), Reserve (+60%), Uber for Business (+30% YoY), and lower-cost products (Moto $1.5B+, +40%; UberX Share/Wait & Save) are all expanding simultaneously. The premium tier is the higher near-term margin contributor; the lower-cost tier is the larger long-term TAM expander but currently runs at a margin investment management explicitly considers appropriate for the growth opportunity.

"In terms of the barbell strategy, it is both. It is low cost and premium. I would say that premium is easier to execute on... I would also say that our Uber for Business is also growing nicely over 30% on a year-on-year basis. So I think premium near term in terms of profitability is — has higher potential, call it, over the next 3 years." — Dara Khosrowshahi, CEO

Assessment: The Mobility setup heading into Q3 is the strongest in over a year. Insurance pressure unwinding to consumer benefit, July transaction acceleration confirmed, barbell strategy producing simultaneous premium and low-cost growth, and the AV halo effect adding incremental engagement in launch markets. The segment is firing across multiple vectors.

Delivery — Membership Flywheel Driving Frequency

Delivery's Q2 strength was anchored on Uber One. Members spend 3x more than non-members; Uber One reached 36M in Q2 (+60% YoY). The Trendyol Go acquisition (closed Q1) is contributing to growth, and the platform-strategy focus on cross-promoting between Mobility and Delivery is adding incremental frequency for the ~20% of consumers who already cross-engage.

The cross-platform economics are the quietly important Delivery story. Cross-engaged users generate 3x the gross bookings AND 35% higher retention than single-business users. Today fewer than 1 in 5 consumers cross-engage. Andrew McDonald's COO appointment with both Mobility and Delivery reporting to him structurally addresses what was previously an org-design impediment to cross-platform optimization — both teams now work toward a unified user journey rather than independently maximizing their own business.

"The Mobility app drives $10 billion of bookings or delivery kind of bookings on the Mobility app. It's about 12% of annualized delivery gross bookings. So in some ways, we're slowly kind of moving towards a super app of source." — Dara Khosrowshahi, CEO

Assessment: Delivery is benefiting from membership flywheel compounding, M&A integration (Trendyol Go), and the early stages of cross-platform optimization. The organizational restructure under McDonald is a multi-quarter tailwind — the impact won't show in a single print but compounds via better cross-promotion targeting.

Freight — Stable Through-Cycle

Freight was not specifically called out in management's prepared remarks or Q&A, consistent with its through-cycle role as a smaller-margin contributor that is neither a structural driver nor a structural detractor at present. The consolidated +18% gross bookings growth print absorbs Freight's flatter trajectory cleanly given Mobility and Delivery scale.

Assessment: The lack of Freight commentary is itself the takeaway — the segment is operating as expected and is not a swing factor for the consolidated thesis. Watch for any shift in management commentary in subsequent quarters as a leading indicator of either acceleration or write-down risk.

Key Topics & Management Commentary

Overall Management Tone: The most confident Uber call posture since the post-pandemic recovery period. Khosrowshahi opened by listing all-time records, framed AV as Uber-led, and announced a $20B buyback in the prepared remarks — an unusually forward-leaning Q2 setup. Mahendra-Rajah's CFO commentary was structured (clear framing on capital return, AV investment posture, M&A philosophy), reinforcing the disciplined-growth-with-disciplined-return narrative.

Autonomy — Uber as the Aggregator, Not the Disintermediated

The autonomy commentary on this call was the strongest articulation yet of why the Uber network model wins regardless of which AV stack scales. Three points anchored it:

1. The supply philosophy. Khosrowshahi: "we are a supply-led company. The more drivers there are in the ecosystem that we can amalgamate with our platform, the better our service becomes, and that applies with robotic and autonomous drivers as well." Humans and AVs are the same supply pool to Uber's marketplace.
2. The empirical utilization data. The average Waymo on Uber's network is "busier than 99% of our drivers in terms of completed trips per day." This is not a forward projection — it's current production data. AVs deployed on Uber's network are at 99th-percentile utilization out of the gate.
3. The partnership panoply. Q2 added Baidu, Lucid, Nuro, and Wayve to a partnership list that already included Waymo, WeRide, Avride, Volkswagen, Pony, and Momenta. The Lucid/Nuro deal is the first capital-committed vehicle-and-software stack — up to 20,000 vehicles over multiple years, with Mahendra-Rajah explicitly stating the deal is expected to yield better economics than asset-light arrangements.

"I wouldn't conclude that by purchasing vehicles the economics are going to be worse. In fact, our analysis is that the Nuro, Lucid deal will probably yield better economics for us because it's a collective deal and we just get the efficiencies of having the vehicle and having the software integration in there." — Prashanth Mahendra-Rajah, CFO

Mahendra-Rajah outlined the three business models that will coexist in Uber's AV marketplace over the next five years: a merchant model (Uber pays partner per-trip or per-day, takes monetization risk), an agency model (rev-share like the existing driver model), and a capital model (Uber or financing partner owns the assets and licenses the software). This framework converts a previously fuzzy "Uber and AV" narrative into a structured marketplace design.

Assessment: The narrative around AV has shifted decisively in Uber's favor over the past two quarters. The Atlanta Waymo launch, the Lucid/Nuro capital-committed deal, the breadth of partnership announcements, and the empirical 99th-percentile utilization data have collectively reframed AV from "existential threat" to "Uber-led aggregation play." Tesla's robotaxi deployment so far is "very, very small" per Khosrowshahi and has not measurably affected Uber's San Francisco or Austin trends.

$20B Buyback — The Capital Return Inflection

The $20B authorization is in addition to the ~$3B remaining from the prior 2024 program, for a combined ~$23B available over the coming years. At Uber's current market cap, this represents approximately 12% of equity value. Mahendra-Rajah committed to a programmatic ~50% of FCF being deployed to buybacks on a multi-year basis, with opportunistic incremental purchases during dislocations.

"We've been allocating around 50% of our free cash flow to buybacks. I think that's a fair sort of way for you to think about how we will execute the capital return over the coming years... we made a commitment last year in our Investor Day that we were going to turn the curve and start reducing our share count. And now in the second quarter, we've actually taken share count down 1%, and you'll see that trend continue for the next couple of years." — Prashanth Mahendra-Rajah, CFO

Assessment: This is the cleanest capital-return commitment Uber has ever made. The programmatic 50%-of-FCF framework is the kind of disciplined allocation policy that earns multiple expansion at quality compounders. The first-time-ever Q2 share count reduction is the empirical proof that the engine is now live, not just rhetorical.

Membership — Mobility-Optimized Product Roadmap

Uber One's 36M members (+60% YoY) is impressive, but the structural thesis update on this call was the explicit acknowledgment that membership is "highly optimized for Delivery, call it, 80% of where we're going to get to" but "not nearly as optimized" for Mobility. The product roadmap response is Surge Savings — what Khosrowshahi called "the #1 product that our Mobility consumers have asked for."

"Surge is necessary for us to improve reliability across the network, but a lot of our users don't like it to be frank. So surge savings is an opportunity for members to save during those surges and kind of pay prices closer to what they're used to, and we think that can drive membership as well." — Dara Khosrowshahi, CEO

Assessment: Surge Savings is exactly the right Mobility-tailored membership benefit — it monetizes the structural gap between member and non-member willingness-to-pay during high-demand periods, while improving member satisfaction with peak pricing dynamics. If the Mobility membership flywheel reaches even 60–70% of Delivery's optimization, the Uber One run-rate has multi-year compounding ahead of it, well beyond the +60% YoY pace.

Platform Strategy — Cross-Engagement and AI Personalization

Khosrowshahi's commentary on platform — explicitly the focus of the new COO Andrew McDonald — addressed an under-appreciated structural asset. Cross-engaged users generate 3x the gross bookings and profits of single-business users with 35% higher retention. Today fewer than 1 in 5 consumers cross-engage. The implied uplift if cross-engagement penetration were even modestly increased is meaningful.

"For a subset of consumers we can just pay more than anyone else can. And that structural advantage is going to continue to get better over a period of time." — Dara Khosrowshahi, CEO

The AI/personalization angle ties directly into platform: Uber's models are increasingly capable of context-aware promotional targeting (the example Khosrowshahi used: a Starbucks promotion timed to the morning commute). The "right-time, right-person" promotional capability is what allows Uber to economically support cross-platform marketing in ways monoline competitors structurally cannot.

Assessment: Platform is the most quietly important narrative on this call. The 3x bookings / 35% higher retention statistic from cross-engaged users is a structural moat that is currently underexploited. McDonald's COO appointment institutionalizes the org structure to lift that share. The compounding flywheel: more cross-engagement → better personalization data → better targeting → more cross-engagement.

Externalizing Tech Capabilities — The Optionality Layer

Khosrowshahi closed the call with commentary on externalizing Uber's platform capabilities, citing two examples: (i) the advertising business plus the on-demand fulfillment business (think Uber Direct), where Uber separates audience-monetization from fulfillment-as-a-service, and (ii) Uber AI Solutions, where drivers and specialists earn through data labeling, translation, map labeling, and algorithm tuning for AI training partners.

"One way to look at Uber is, it is a platform for work. We've got almost 9 million people who are earning on our platform, and they can earn in different ways other than driving or transporting things... it's using the core Uber capability, which is sending out tasks to earners all over the world, you're just going to see a different kind of earner that is going to work for kind of the really exciting kind of AI developments." — Dara Khosrowshahi, CEO

Assessment: Uber AI Solutions is small today but represents structurally interesting optionality — Uber's task-distribution infrastructure to 9M earners globally is a marketable platform capability in the AI training-data economy. This is not a near-term needle-mover for the model, but it's a long-duration option on Uber's labor-platform infrastructure that costs nothing to maintain.

Guidance & Outlook

The Q3 guide is the strongest forward statement Uber has made in over a year. High-teens gross bookings growth maintains Q2's pace (no deceleration), and low- to mid-30s adjusted EBITDA growth implies operating leverage widening rather than mean-reverting. Khosrowshahi's "we're expecting more of the same strong performance in Q3" framing is unusually direct.

The forward setup beyond Q3 is supported by multiple compounding drivers: insurance pass-through compounding into Q4, AV deployments ramping, Surge Savings rolling out to Mobility members, the McDonald-led platform reorganization beginning to optimize cross-engagement, and the Trendyol Go integration adding to Delivery growth. None of these are step-change catalysts — they're small compounders that collectively support sustained high-teens growth and widening operating leverage through 2026.

Analyst Q&A Highlights

Topic: Platform Strategy & Super App Question

• Eric Sheridan, Goldman Sachs: Asked Khosrowshahi how cross-platform behavior is unlocked — consumer knowledge, supply array, or affordability — and whether Uber should consolidate to a single super app under one brand. Khosrowshahi framed cross-platform as "easy to talk about, but actually much harder to execute," cited the 35% retention uplift and 3x bookings/profit advantage of cross-engaged users, and argued for a "best of both worlds" approach: highly tuned standalone apps with intelligent cross-promotion. Mobility app already drives ~$10B (12%) of delivery bookings as a partial super app.
  Assessment: The clearest articulation of the platform strategy yet. The "second inning" framing implies multiple years of compounding ahead from cross-platform optimization — a quietly large unmodeled tailwind.

Topic: MAPC Acceleration & AV Utilization

• Brian Nowak, Morgan Stanley: Asked about the +10M MAPC sequential add and the +6M Uber One add — what drove the faster growth and durability into 2H. Also asked for AV deployment quantification beyond the 1.5M-trips figure cited prior quarter. Khosrowshahi attributed the audience strength to barbell strategy (low-cost Moto, premium tier, demographic-targeted products like teens and elder), noted "no signal whatsoever that audience growth is slowing," and on AV cited the 99th-percentile Waymo utilization stat plus positive halo effect on overall system engagement.
  Assessment: The 99th-percentile Waymo utilization datapoint is the single most important AV data disclosure of the call — converts AV from a forward speculation to a current production reality on Uber's network.

Topic: AV Capital Allocation & Asset Ownership

• Michael Morton, Moffett Nathanson: Probed concerns that the Lucid/Nuro deal signals capital intensity creep and reads negatively on the Waymo relationship. Asked how Uber thinks about owning vs. divesting AV assets. Khosrowshahi framed Uber as a supply-led company indifferent between human and AV drivers, emphasized the deal does not signal Waymo deterioration, and noted Uber's cash flow allows aggressive AV investment alongside meaningful capital return ("not either/or"). Mahendra-Rajah outlined the equity-stake-in-software / capital-in-vehicles / cash-flow-funded-by-minority-stake-recycling framework.
  Assessment: The clearest framework yet for AV capital deployment. Mahendra-Rajah's framing that the Lucid/Nuro deal yields better economics than asset-light arrangements is a meaningful update from prior commentary that emphasized asset-light positioning.

Topic: OEM Partnerships & Tesla Competitive Read

• Justin Post, Bank of America: Asked about the OEM partnership pipeline and the Tesla competitive impact on San Francisco / Los Angeles. Khosrowshahi noted hardware (not software) is the harder commercialization gate at scale and confirmed Uber is talking to "all of the major OEMs," with more announcements coming. On Tesla: deployment is "very, very small" and has not measurably moved Uber's trends in Austin or San Francisco.
  Assessment: The "more OEM announcements coming" forward signal plus the "no measurable Tesla impact" reassurance both reduce risk on the AV competitive thesis.

Topic: Mobility Pricing Pass-Through & Buyback Cadence

• Doug Anmuth, JPMorgan: Asked Khosrowshahi about consumer response to U.S. Mobility pricing decreases (insurance pass-through) and Q3 trip acceleration confidence; asked Mahendra-Rajah for buyback timeline color given the 2024 program ramped slowly. Khosrowshahi confirmed July transaction growth accelerated nicely versus Q2 and that the same trends will persist through Q3 and into Q4. Mahendra-Rajah laid out the ~$23B combined available, ~50% of FCF programmatic framework, and "active every quarter" execution posture.
  Assessment: The most actionable pair of forward datapoints on the call — July Mobility acceleration (current-quarter signal) plus the programmatic 50%-of-FCF buyback (multi-year capital-return signal).

Topic: AV Vehicle Commitment Sizing & Waymo Expansion

• Ross Sandler, Barclays: Asked for dollar sizing on the OEM-partnership vehicle commitments and how important Waymo expansion to additional cities is. Mahendra-Rajah declined to size the dollar commitment but reiterated 50% of FCF goes to buybacks with the remainder funding AV ecosystem build. Khosrowshahi noted Austin and Atlanta Waymo deployments are going well, both highly utilized, and that he'd like to expand but won't pre-commit to timing or geography.
  Assessment: The non-commitment on Waymo geographic expansion is interesting — suggests the Waymo relationship is on Waymo's pace, not Uber's, but the existing markets are producing the empirical data Uber needs to recruit additional partners.

Topic: Mobility Barbell & Data/AI Externalization

• Benjamin Black, Deutsche Bank: Asked which side of the U.S. Mobility barbell (premium vs. low-cost) carries the bigger incremental opportunity, and pressed on what "externalizing technical capabilities" means in practice (data licensing for AV?). Khosrowshahi confirmed both barbell sides are growing and that premium has higher near-term profitability while low-cost has greater long-term TAM. On externalization, he pointed to advertising/Uber Direct separation, AV data collection (not for profit but for ecosystem speed), and Uber AI Solutions (data labeling, translation, algorithm tuning) leveraging the 9M-earner platform.
  Assessment: The Uber AI Solutions commentary is the most novel disclosure of the call — Uber's task-distribution infrastructure as a marketable AI-training-data platform is structurally interesting optionality.

What They're NOT Saying

1. Specific quantification of AV capital commitments. Mahendra-Rajah explicitly declined to size the dollar amount Uber will commit to vehicle and software partner stakes over the next several years. The "modest redeployment of FCF" framing is comforting but unspecific. Investors should expect periodic capital deployments to AV partners that are not pre-announced.
2. Waymo expansion timeline. Khosrowshahi was explicit that he "won't comment on when and if" Waymo expansion to additional cities happens. The implication: Waymo's pace, not Uber's. This is consistent with the Waymo direct-deployment strategy in some markets, and the silence is itself a soft signal that the relationship is healthy but not on Uber's preferred geographic timeline.
3. FY25 full-year guidance update. Management did not formally update FY25 framework numbers despite the operating beat. The Q3 guide implies trajectory, but the absence of a multi-quarter FY25 framework refresh leaves the Street to do the math.
4. Detailed Uber AI Solutions sizing. The business is "growing very quickly off of a smaller base" but was not quantified. Could be $50M or $500M; the operating leverage profile is unknown. Worth tracking as it scales.
5. Freight commentary entirely absent. Neither prepared remarks nor Q&A addressed Freight performance, outlook, or strategic positioning. Could simply reflect Q2 stability, or could foreshadow strategic restructuring in subsequent quarters. Watch for any change in commentary cadence.
6. Tesla robotaxi competitive update specifics. Khosrowshahi's "very small" framing is reassuring but the Austin Tesla robotaxi deployment was less than two months old at the print — the competitive read-through is necessarily preliminary. Q3 and Q4 will produce more data.
7. Insurance pass-through duration. Khosrowshahi confirmed insurance savings are being passed through "for the balance of the quarter" but did not quantify how long the structural insurance tailwind lasts. The insurance market has cyclical and regulatory dynamics; the tailwind window may close in 2026.

Market Reaction

• Pre-print setup: UBER had been trading near recent highs heading into the print, with the Street looking for confirmation that the Q1 Mobility deceleration on insurance-cost pressure was unwinding. Pre-print expectations had bracketed gross bookings growth in the mid-teens, with EBITDA growth in the high-20s to low-30s. The bar entering the print was constructive but not heroic.
• Initial reaction: The +18%/+18% trips and bookings prints, combined with the $20B buyback authorization announced in the prepared remarks, generated immediate positive momentum. The market absorbed the AV partnership announcements, the Atlanta Waymo launch, and the Q3 guide of low- to mid-30s EBITDA growth as collectively bullish. The first-time share-count reduction provided empirical proof of the capital-return inflection. Volume was elevated; institutional flow was net buying.

The market reaction reflected the compounded positive content: not a single headline moved the stock, but the cumulative weight of the +18% growth print, the buyback authorization, the AV partnership flurry, the share count decline, and the Q3 acceleration commentary collectively reset expectations. Pre-print bears who anchored on insurance-cost pressure or AV-disintermediation risk were squeezed by the Q3 acceleration commentary and the Atlanta Waymo launch respectively. The forward setup is cleaner than at any point in the past 12 months: high-teens growth pace confirmed for Q3, programmatic buyback live, AV ecosystem expanding under Uber's network gravity.

Street Perspective

Debate: Is Uber Operating Leverage Sustainable, or Margin-Capped Near Current Spread?

Bull view: The Q3 guide of low- to mid-30s adjusted EBITDA growth on high-teens gross bookings growth implies a sustained 15–17pp spread, and the underlying drivers compound from here. Uber One scaling (3x spend per member with +60% YoY membership growth), cross-engagement compounding (3x bookings, 35% higher retention from sub-20% penetrated cohort), advertising margin (high-margin, fast-growing), and Uber Direct as a separable margin layer all support widening leverage through 2026. The bear case for margin compression assumes growth slows; growth is not slowing.

Bear view: Uber operating margins are already at structural highs. AV capital deployment (Lucid/Nuro vehicle commitments, equity stakes in software partners, data-collection investment) drains margin in the medium term even if accretive long-term. Mobility pricing pass-through is currently insurance-funded but the insurance cycle will normalize, removing the volume-without-margin-erosion dynamic. The 15–17pp spread is peak, not run-rate.

Our take: The bull case is winning the data, but with measured caveats. Operating leverage will widen modestly through 2026 driven by Uber One Mobility penetration, advertising scaling, and cross-engagement compounding. AV capital is a real medium-term margin headwind but is being explicitly funded by minority-stake recycling rather than pure FCF dilution per Mahendra-Rajah's framework. We model Adjusted EBITDA growth running ~10–12pp ahead of bookings growth on average through 2026, narrower than the Q3 guide implies but wider than the bear scenario.

Debate: Does the AV Ecosystem Settle Around Uber, or Does Waymo Go Direct?

Bull view: The breadth of partnerships announced this quarter (Baidu, Lucid, Nuro, Wayve added to the existing roster) plus the empirical 99th-percentile Waymo utilization on Uber's network plus the consumer halo effect plus Uber's supply-led aggregation thesis all point to Uber as the structural aggregator. AV operators benefit from Uber's demand, capital, and operational learning; Waymo's direct-deployment in some markets is complementary, not zero-sum. The "panoply of models" Khosrowshahi described is healthy ecosystem behavior, not fragmentation risk.

Bear view: Waymo's direct-to-consumer deployment in San Francisco, Phoenix, and other markets demonstrates that the most valuable AV operator can bypass the aggregator. As Tesla, Waymo, and others scale direct robotaxi services, Uber's role compresses to a marginal-supply role for less-utilized AV players. The 10% NFL-style alignment Disney secured is exactly what Uber has not secured with Waymo — Uber lacks structural deal protection against Waymo going direct in additional cities.

Our take: The ecosystem settles into a plurality, not a winner-take-all. Waymo will deploy directly in some cities and through Uber in others, depending on local economics and market entry timing. Uber's value proposition — demand aggregation, supply balance, multi-stack platform — is genuinely differentiated for the long tail of AV operators (the 9-10 partners now signed up) even if Waymo selectively goes direct. The empirical 99th-percentile Waymo utilization on Uber's network is itself the proof: even Waymo gets more utilization in Atlanta on Uber's network than via direct deployment, which is why Atlanta launched exclusively with Uber. The bear view requires Waymo to scale direct AND maintain pricing AND avoid utilization compression in markets where they bypass Uber — we don't think all three hold.

Debate: Is the $20B Buyback the Right Use of Capital vs. AV Investment?

Bull view: Mahendra-Rajah's "not either/or" framing is correct. Uber's FCF generation comfortably funds ~50% capital return AND aggressive AV investment AND opportunistic M&A. The first-time share count reduction in Q2 demonstrates the capital return engine is live. Programmatic 50%-of-FCF buyback is the right policy for a quality compounder generating record cash flow.

Bear view: $20B is a large authorization at a stock trading near recent highs. AV capital-deployment requirements are uncertain — if Uber needs to commit materially more to vehicle ownership and software equity stakes to defend against Waymo direct or to support struggling AV partners, the buyback envelope shrinks. Buying back stock at 30x earnings is less compelling than buying back at 20x.

Our take: The programmatic 50%-of-FCF framework is the right policy regardless of stock-price level — it removes timing-discretion risk. If Uber's stock trades to 40x next year, the policy still produces share-count reduction at a slower pace. If it sells off to 20x, the policy produces faster share-count reduction. The discipline IS the value. We'd prefer Uber maintain the 50%-of-FCF anchor rather than try to time the buyback.

Model Implications

Valuation impact: The combined effect of higher growth (FY25 bookings +17% vs. +15%), wider operating leverage (EBITDA +30%+ vs. +25–28%), and faster share-count reduction (down 1.5–2% vs. flat) compounds into a meaningfully stronger per-share earnings trajectory. Our 12-month framework moves Outperform-side: the AV optionality is no longer a 2027+ question but is producing live data on Uber's network today, the Uber One Mobility roadmap is explicit, and the buyback engine is empirically live. Bull-case path: bookings growth holds high-teens through 1H26, Uber One reaches 50M+, AV trip volume grows materially as Lucid/Nuro deploys. Bear-case path: AV capital deployment surprises to the upside, insurance pass-through runs out in 2026, Mobility growth normalizes earlier than current trajectory.

Thesis Scorecard Post-Earnings

Overall: The thesis has materially strengthened from the Q1 Hold initiation. Five of the six bull points are confirmed (vs. the three pending in Q1), with three new structural bulls added (capital return inflection, cross-engagement TAM, externalization optionality). The bear points remain persistent but contained — Waymo direct is mitigated by empirical utilization data, insurance pass-through is cyclically bounded but supportive through the relevant forecast horizon, AV capital intensity has an explicit framework cap, and the operating-leverage and Tesla bear cases are actively rebutted by the Q3 guide and current Tesla deployment scale.

Action: Upgrading to Outperform from Hold. The Q1 Hold was anchored on three open questions: (1) AV ecosystem durability against direct-to-consumer competition, (2) Uber One penetration into Mobility, and (3) capital return actualization. Q2 resolved each: AV partnerships expanded with Uber as the structural aggregator and 99th-percentile Waymo utilization confirming the network value, Surge Savings is the explicit Mobility-tailored membership product on the roadmap, and the $20B buyback combined with the first-ever share count decline proves the capital return engine is live. The +18%/+18% trips and bookings prints, the all-time-high audience and frequency, the Q3 guide of widening operating leverage, and the AV/buyback strategic announcements collectively reset the risk-reward. The valuation case is no longer "wait for execution" — it's "underwrite the compounding off the now-empirical base."

Upgrade triggers (further): Uber One reaching 50M+ exit FY25 with material Mobility-membership penetration; AV trip volume on Uber's network reaching 10M+ cumulative by year-end with positive economics disclosure; programmatic buyback exceeding 50% of FCF on opportunistic dislocation. These would each support a higher conviction multiple.

Downgrade triggers: Waymo announcing direct-to-consumer expansion in additional major markets without Uber participation; AV capital commitments materially exceeding the minority-stake-recycling framework, dragging FCF allocation away from buyback; Mobility growth decelerating to single digits as insurance pass-through runs out without offsetting volume drivers.