CAPITAL THESIS

Independent equity research · UBER · full edition

Uber TechnologiesThe network is compounding. Acquisitions must earn their cash.

Uber’s demand and earnings engine remains attractive, but new acquisition commitments reduce the cash available for share retirement. A transparent model tests the five-year return after those competing claims.

Research date: 15 October 2026 · USD common equity · Reference price: USD 69.29 · Observation: 2026-10-14

Price-disciplined watchlist

The investment decision

At $69.29, our revised base scenario delivers 19.6% annualised over five years. The ezCater agreement warrants slower buybacks; the cash-flow cross-check and unresolved deal funding argue for a buffer.

Reference price$69.2914 Oct 2026, 20:00 UTC; recent public quote [M1]
Q2 Gross Bookings$58.0bnQ2 2026 actual; 22% constant-currency growth [S1]
2026E normalised P/E21.0xHouse $3.30 EPS; retains SBC
Base / no-rerating CAGR19.6% / 18.5%Five-year conditional scenarios
Conditional 20% entry$68.08Arithmetic threshold; no execution buffer

Edition update. October update: incorporates the 6 October ezCater agreement, defers net share retirement and charges foregone cash interest. Q2 actuals remain latest; core operating forecasts retain the September vintage after review. New source-labelled exhibits and reproducible calculations; no invented acquired earnings.

01. Investment view and what changed

Price-disciplined watchlist. Uber has a strong operating franchise, but acquisitions are now the important test of shareholder compounding. At $69.29, the revised central scenario returns about 19.6% a year over five years, close to a demanding 20% hurdle with little execution buffer. The lower quote does not make the business automatically cheaper after committed cash uses rise.

The meaningful October development is the agreement to purchase ezCater for $2.3bn in cash. We therefore delay base-case net share retirement by a year and charge foregone cash interest, while giving no credit to undisclosed target earnings. This is a conservative treatment of missing information, not a forecast that ezCater earns nothing. The full acquisition funding bridge remains incomplete. [S6]

Our operating assumptions still originate in the 22 September model: 2026 estimated bookings of $233bn, $3.30 normalised EPS and 2.050bn diluted shares. We reviewed the current issuer releases; the latest published financial statements remain Q2, with Q3 due in November. We have not invented a fresh quarterly result or current analyst consensus. This edition adds a reproducible model, sourced operating exhibits and a clearer test of what investors actually receive. [S8] [S9]

02. Network economics and the competitive debate

Uber sells access to a two-sided marketplace: customers value short waits and reliable selection; drivers, couriers and merchants value demand. Density can improve utilisation and lower the incentive needed to match a transaction. A customer using both Mobility and Delivery is potentially cheaper to retain than two separately acquired users. These advantages matter only if the platform retains enough economics after incentives, insurance and support.

Gross Bookings is the most useful demand anchor. In Q2, Mobility represented $28.988bn, Delivery $27.463bn and Freight $1.571bn. Revenue is a less stable growth comparator because principal-versus-agent and other business-model changes can alter reported sales without the same change in customer spending. We model earnings per dollar of bookings, and then divide by the shares investors actually own. [S1]

The moat remains contestable. Riders and restaurants can use competing platforms; local density is more useful than a worldwide brand in an underserved market. Autonomous fleets could benefit from Uber demand but could also negotiate lower fees, demand financing or bypass the app. We assign no robotaxi valuation premium and no presumed monopoly margin. Insurance inflation and worker-status rules can transfer value away from shareholders even while trips rise. [S10]

Mobility remains the established profit engine. Delivery can add contribution through density, retail and advertising, while Freight is not yet a material source of group profit. Expansion should be judged on incremental cash after compensation and the capital used, not on a bigger addressable-market slide.

03. Operating results and earnings quality

USD bn unless statedQ2 2025Q2 2026
Gross Bookings46.75658.022
Revenue12.65114.191
GAAP operating income1.4501.890
Non-GAAP operating income1.5342.143
Non-GAAP EPS (USD)0.600.81
Trips (bn)3.2683.867

Source: issuer Q2 release. Normalised earnings retain stock compensation; adjusted EBITDA is not a substitute for owner earnings. Investment marks inflated the latest GAAP net result, and earlier tax-valuation releases also distort a trailing P/E. We use estimated normalised EPS rather than a low headline GAAP multiple. The 2026 EPS anchor is a house estimate, not full-year company guidance. [S1] [S2]

At $69.29, $3.30 estimated normalised EPS implies 21.00x earnings, a 4.8% earnings yield. That yield is not added to the model CAGR: earnings fund growth, deals and repurchases, while the scenario already includes their effect on EPS. A mechanical PEG would conceal acquisition financing and the distinction between core growth and share retirement.

The next results need to validate both demand and earnings conversion. The published Q3 ranges are $58.25–60.25bn of bookings and $0.84–0.88 non-GAAP EPS. Beating a bookings target while cash conversion deteriorates would be weaker evidence than growth with resilient per-trip economics. [S1]

04. Cash, debt and the cost of maintaining ownership

USD bnH1 2025H1 2026
Operating cash flow4.8885.213
Cash capital expenditure0.1630.135
Issuer free cash flow4.7255.078
Stock compensation0.9101.023
FCF less SBC: economic proxy3.8154.055
Common repurchases3.1483.529

The last-but-one row is our compensation-aware diagnostic, not an issuer measure or cash legally available for distribution. H1 cash flow included $0.830bn from increased insurance reserves and a $0.771bn deferred-tax addback. Neither is inherently bad accounting; both make simple extrapolation of conversion unreliable. The platform must ultimately settle claims and finance its obligations. [S3]

June debt carrying amounts totalled $12.723bn, against approximately $5.391bn of unrestricted cash and short-term investments. Operating lease liabilities were $2.008bn and require a separate operating-capacity assessment. Restricted insurance investments are not spare takeover funding. September euro borrowing initially increases cash and debt together; adding the debt while omitting the proceeds would overstate the immediate net-debt change. [S3] [S7]

Repurchases only help continuing holders after stock issuance is covered. Our scenario share paths are net of employee dilution. We retain SBC in normalised earnings, and use FCF less SBC only as a separate diagnostic rather than subtracting it a second time in the EPS model. No extra dividend, buyback yield, investment-portfolio windfall or residual cash balance is added to terminal shareholder wealth.

05. Acquisitions: strategic logic must earn its funding

Delivery Hero expands geographic reach, but an agreed takeover is not already-owned profit. The €41.50 offer implies $14.8bn for all target equity, or $13.7bn adjusted for earlier stake purchases; these headlines are not a complete remaining-cash requirement. Existing interests, derivatives, minority claims, target liabilities and separate sales of overlapping markets complicate the bridge. The tender deadline is 5 November and closing remains conditional. [S4] [S5]

The revised scenarios credit Delivery Hero earnings only through an explicit percentage overlay on core net income: −8%, +5% and +10% in year five. These are house assumptions after financing, tax, integration and ownership costs, not full merger accounts. The base deliberately stays below management’s stated high-single-digit EPS accretion ambition by year three. We do not add target bookings to the core forecast and then also count its full earnings.

ezCater brings workplace catering and large orders, with disclosed trailing bookings above $2.5bn and qualitative positive non-GAAP operating income. No quantified net-profit, purchase-accounting or detailed financing bridge was supplied in the announcement. We recognise the $2.3bn commitment through a delayed-buyback sensitivity and deduct $64m a year of assumed after-tax foregone interest: $2.3bn × 3.5% × 79%. The assumed share-retirement path is not derived from an integrated cash budget; this calculation does not demonstrate a funded purchase-price reserve or prove the remaining buybacks affordable. There is no incremental ezCater earnings uplift in this model. [S6]

This treatment is intentionally asymmetric until evidence improves. It is not evidence that the transaction destroys $2.3bn of value. A complete underwriting would compare acquired cash flows with the cash paid and financing risk. Future disclosure could support higher earnings, or reveal larger integration costs and less repurchase capacity.

06. Five-year scenarios and the base operating path

House assumption / outputBearBaseBull
Annual core booking growth path10% → 6%16% → 12%21% → 17%
Year-five core net income / bookings2.00%3.25%3.60%
Year-five deal earnings / core earnings−8%+5%+10%
Cumulative net share reduction0.0%4.4%8.2%
Year-five normalised EPS$3.04$7.70$11.49
Exit P/E14x22x26x
Year-five share value$42.57$169.41$298.74

All scenario inputs are Capital Thesis assumptions. No probabilities are assigned. Base requires sustained mid-teens activity growth, an improvement in net profitability and buybacks resuming after acquisition demands ease. Bull requires stronger growth and a premium exit valuation. Bear reflects weak margins despite continued activity growth; it is not a worst-case floor.

Base annual model20272028202920302031
Core bookings (USD bn)270.3310.8354.3400.4448.4
Normalised net income (USD bn)7.979.4911.2413.1815.24
Average diluted shares (bn)2.0502.0502.0402.0141.979
Normalised EPS (USD)3.894.635.516.547.70

Base ending shares retire at 0%, 0%, 1%, 1.5% and 2% annually. EPS uses the average of beginning and ending shares each year. Core net margin moves linearly from the 2026 estimate to the stated terminal level; this is an explicit simplifying assumption, not a quarterly forecast.

Bear case: five-year total-wealth CAGR is -9.3%, from $69.29 to $42.57 per share with no cash dividends.

Base case: five-year total-wealth CAGR is 19.6%, from $69.29 to $169.41 per share with no cash dividends.

Bull case: five-year total-wealth CAGR is 33.9%, from $69.29 to $298.74 per share with no cash dividends.

No-rerating case: holding the starting 21.00x normalised P/E constant gives $161.69 per share and a five-year total-wealth CAGR of 18.5%.

Total-wealth CAGR is ((terminal share value + cumulative cash distributions) / reference price)^(1/5) − 1. No dividends are assumed; if cash distributions existed they would be held at zero reinvestment return. This endpoint CAGR is not an IRR and does not imply the same gain every year.

07. Cross-check the multiple before trusting the headline

At unchanged operating performance, valuation can materially change the result. Holding the base EPS fixed, 18x terminal earnings produces $138.61 and 14.9% CAGR; 26x gives $200.21. The 22x base is only modestly above the current 21.00x estimate, but even that small expansion contributes about one percentage point to annualised return.

Equity discount rate10%12%14%
Existing-core DCF per current share$93$70$56

The independent core cross-check starts at $7bn of annual economic equity cash flow, grows 14% for five years and 8% for another five, then 3% perpetually. Shares remain fixed at 2.050bn. These are house assumptions deliberately below the earlier LTM FCF-less-SBC proxy. The model uses equity cash flows, so existing net debt is not subtracted again. It excludes the takeover cash and acquired cash flows together; it is not a complete post-acquisition DCF.

The approximately $70 central cash-flow value is much less emphatic than the earnings scenario. That disagreement argues for an entry buffer and better cash evidence. Discount-rate precision cannot solve uncertain capital allocation or prove that a growth company is grossly mispriced. No current consensus target or peer PEG is used as a substitute for this work.

08. Entry discipline and disconfirming evidence

Conditional 20% entry: $68.08 equals the base terminal value of $169.41 divided by 1.20 to the fifth power. This is a model-dependent threshold, not a margin of safety or a promised return.

The hurdle holds modeled operating results, net share-retirement paths and terminal multiples fixed. Changing today’s entry price does not automatically change future corporate repurchase prices or funding capacity. The website’s quote sensitivity changes the entry denominator only; it does not rerun the company’s business or financing model.

At the current reference price, 20% annually requires $172.42 in five years. An additional 8% discount to the arithmetic entry threshold is about $62.64. A roughly $62–64 research range offers some buffer in this model, provided the investment thesis remains intact; a falling quote alone is not evidence of value.

MonitorEvidence that strengthens the caseReason to reconsider
Bookings and economicsMid-teens or better core growth; durable net profitabilityTwo quarters below roughly 12% constant-currency growth, or persistent margin erosion
Cash and ownershipCash conversion supports strategic spending and net share retirementRepurchases fail to offset dilution; insurance or cash taxes consume expected funds
DealsClear retained-scope profits, cash commitments and financingHigher costs, restrictive remedies or weaker acquired cash earnings
AutonomyUseful distribution economics without outsized capital supportFleet bypass, fee compression or greater asset funding
Next dated events4 November earnings; 5 November tender deadlineGuidance misses or transaction changes without credible explanation

Thresholds are our monitoring rules, not company promises. Current short interest, borrow cost and a verified real-time ownership register were not obtained; we do not infer a squeeze or forced buying. Management incentives matter through capital allocation and dilution, not through an unsupported claim about market positioning. [S5] [S8]

09. Evidence, model boundaries and decision

Uber remains a credible compounding business, but today’s valuation requires successful execution. The revised base case is still near the 20% hurdle because a modestly lower entry price broadly offsets more cautious capital allocation. That coincidence is not a reason to preserve the old recommendation mechanically.

Actuals run through June 2026; announcements were checked through 15 October. Forecast years 2027–2031 approximate a five-year investment horizon rather than exact October fiscal stubs. The model is reproducible but does not include a complete debt maturity waterfall, target purchase-price allocation, deal-level working capital or future regulatory settlements. Latest-price repricing changes entry mathematics, not these assumptions.

Reported facts, management guidance, historical house assumptions and new October judgments are labelled separately. The prior report’s sparse external consensus and peer table are not repeated as current data. USD returns exclude fees, investor taxes and currency translation; an Australian investor’s AUD outcome can differ materially. This is public-source research, not personalised sizing or leverage advice.

Financial exhibits

Exact observations accompany every chart. Actuals, management guidance and Capital Thesis assumptions are labelled separately; the chart shapes do not imply probabilities.

Chart 01 · USD bn

Both consumer marketplaces are growing

Mobility and Delivery provide almost all bookings; Freight remains small.

Q2 2025 actualQ2 2026 actual
Both consumer marketplaces are growing (USD bn). Exact figures in the table below.051015202530MobilityQ2 2025 actual: 23.762 USD bn23.762Q2 2026 actual: 28.988 USD bn28.988DeliveryQ2 2025 actual: 21.734 USD bn21.734Q2 2026 actual: 27.463 USD bn27.463FreightQ2 2025 actual: 1.26 USD bn1.26Q2 2026 actual: 1.571 USD bn1.571
Exact values · USD bn
ObservationQ2 2025 actualQ2 2026 actual
Mobility23.76228.988
Delivery21.73427.463
Freight1.261.571

Issuer Gross Bookings. Segment definitions differ from revenue; no acquisition pro forma added. Sources: [S1]

Chart 02 · USD bn

Cash grows, but compensation is an economic claim

H1 free cash flow less SBC grew more slowly than the strongest headline earnings measures.

Operating cash flowIssuer free cash flowFCF less SBC proxy
Cash grows, but compensation is an economic claim (USD bn). Exact figures in the table below.0123456H1 2025Operating cash flow: 4.888 USD bn4.888Issuer free cash flow: 4.725 USD bn4.725FCF less SBC proxy: 3.815 USD bn3.815H1 2026Operating cash flow: 5.213 USD bn5.213Issuer free cash flow: 5.078 USD bn5.078FCF less SBC proxy: 4.055 USD bn4.055
Exact values · USD bn
ObservationOperating cash flowIssuer free cash flowFCF less SBC proxy
H1 20254.8884.7253.815
H1 20265.2135.0784.055

Actual cash-flow statement. FCF = OCF less cash capex; FCF less SBC is Capital Thesis economic proxy, not distributable cash or an issuer measure. Sources: [S3]

Chart 03 · USD per share

What the central earnings case requires

The investment depends on operating earnings growth while early net buybacks pause.

House normalised EPS
What the central earnings case requires (USD per share). Exact figures in the table below.024682026EHouse normalised EPS: 3.3 USD per share3.32027EHouse normalised EPS: 3.8884 USD per share3.88842028EHouse normalised EPS: 4.6275 USD per share4.62752029EHouse normalised EPS: 5.5089 USD per share5.50892030EHouse normalised EPS: 6.5439 USD per share6.54392031EHouse normalised EPS: 7.7005 USD per share7.7005
Exact values · USD per share
ObservationHouse normalised EPS
2026E3.3
2027E3.8884
2028E4.6275
2029E5.5089
2030E6.5439
2031E7.7005

Capital Thesis forecasts, not company guidance. October model applies ezCater cash opportunity cost, lower net share retirement, and explicit Delivery Hero earnings assumptions. Sources: [S9] [S6]

Source register and research basis

Evidence checked for this edition on 15 October 2026. Dates below identify source vintages. Company guidance and house scenarios are conditional; source links and the downloadable model allow the assumptions to be inspected.