Uber Technologies ($UBER) Stock Hub 2026: First-Ever Autonomous Rides Launch In London With Wayve, As Delivery Hero’s Board Recommends The €41.50 Takeover Offer

01 What Uber Reported, And The Number That Is Not What It Seems

Uber reported net income attributable to the company of $2.394 billion for the second quarter of 2026, up 77 per cent, and diluted earnings per share of $1.17 against $0.63. Those are the headline figures, and they are the least useful numbers on the page.

Inside them sits a pre-tax benefit of roughly $1.6 billion from revaluing shareholdings in other companies. The line is called other income, and it swung from negative $19 million a year earlier to positive $1.342 billion. Strip the mark-to-market and the quarter still made money, but the profit is closer to the $1.890 billion of GAAP operating income than to the $2.394 billion at the bottom.

The operating business is the part worth reading. Revenue of $14.191 billion grew 12 per cent. Gross bookings, the total value of everything transacted on the platform, reached $58.022 billion and grew 24 per cent. Trips rose 18 per cent to 3.867 billion, and monthly active platform consumers rose 16 per cent to 208 million. Cash from operations was $2.862 billion in three months against capital expenditure of $70 million.

That last pair is the whole reason this page sits next to Hertz and Avis rather than in a technology section. In the same three months in which Uber spent $70 million on property and equipment, Hertz spent billions buying cars. Both companies put a stranger in a vehicle and charge for the ride. Only one of them owns the vehicle.

02 Executive Summary

The figures below come from the Form 10-Q and the results release, both filed on August 5, 2026, for the quarter ended June 30, 2026.

The quarter. Revenue of $14.191 billion against $12.651 billion. GAAP operating income of $1.890 billion against $1.450 billion. Net income attributable to Uber of $2.394 billion, of which a large part is the revaluation of equity stakes. Diluted earnings per share of $1.17.

The platform. Gross bookings of $58.022 billion, up 24 per cent. Trips of 3.867 billion, up 18 per cent. Monthly active platform consumers of 208 million, up 16 per cent. Revenue is 24.5 per cent of gross bookings, a Merlintrader calculation from the two reported figures.

The three segments. Mobility produced $28.988 billion of gross bookings and $2.215 billion of segment operating income. Delivery produced $27.463 billion and $1.055 billion. Freight produced $1.571 billion and a $24 million loss. Delivery is now within $1.5 billion of Mobility on bookings.

Cash. Operating cash flow of $2.862 billion in the quarter and $5.213 billion in the half. Free cash flow, as the company defines it, of $2.792 billion, after $70 million of capital expenditure. Cash and equivalents of $4.870 billion plus $521 million of short-term investments, against total debt of $12.723 billion.

The stakes. Didi at $1.900 billion, Grab at $2.020 billion, Aurora Innovation at $1.763 billion, plus $3.076 billion of other holdings. Approximately 61 per cent of the Aurora Class A shares held were pledged as collateral for the 2028 exchangeable notes at June 30, 2026; the position was reduced by the August 17 placement, and no later filing updates that percentage.

The buyback. 6.9 million shares repurchased for $510 million in the quarter, 46.6 million for $3.5 billion in the half, with roughly $15.7 billion of authorisation remaining under the $20 billion programme approved in July 2025. There are 2,042,560,121 shares outstanding at July 31, 2026, and no dividend.

The guidance. Third quarter gross bookings of $58.25 billion to $60.25 billion, growth of 18 to 22 per cent at constant currency, and non-GAAP earnings per share of $0.84 to $0.88, which the company says translates to adjusted EBITDA of $2.86 billion to $2.96 billion.

03 The Delivery Hero Takeover Offer, As The Offer Document Sets It Out

On the morning of August 27, 2026 Uber International Technologies II Corporation, an indirect wholly owned subsidiary of Uber Technologies, published the offer document for its voluntary public takeover offer for all Delivery Hero shares it does not already hold, after approval by Bafin, the German financial supervisor. The announcement of intent dates from July 16, 2026. What changed on August 27 is that the document exists, the terms are binding, and shareholders can act on them.

The price and the premium. EUR 41.50 in cash per no-par registered share of Delivery Hero SE, ISIN DE000A2E4K43. The bidder states the price is a premium of approximately 108 per cent to the unaffected closing price of May 8, 2026 and approximately 127 per cent to the unaffected volume-weighted average Xetra price over the three months up to and including that date. Uber put the equity value at $14.8 billion implied for the whole company, or $13.7 billion adjusted for the stakes it had already bought, on 314 million fully diluted shares.

The calendar, which is the part with a deadline. The acceptance period runs from August 27, 2026 and ends on November 5, 2026 at 24:00 Frankfurt time, 6:00 pm in New York. The bidder warns in the same release that custodian banks may set earlier internal deadlines, so the last useful date for a given shareholder is a matter for that shareholder’s bank rather than for the timetable. Settlement, meaning payment of the cash consideration, is expected in the second half of 2027. Between tendering and being paid there is more than a year, and that gap is a feature of the structure, not an accident.

The threshold, read carefully. The 50 per cent plus one share is measured on the share capital other than the treasury shares held by Delivery Hero, and it counts the shares Uber already owns. Adding the direct 24.77 per cent to the 16.68 per cent Prosus has irrevocably committed gets close to the line before a single other shareholder decides anything. The 11.74 per cent of derivative exposure is a different thing: depending on their terms those instruments settle in cash or permit physical delivery subject to regulatory approval, so it is economic interest rather than votes in hand.

What has not happened yet

Two things are still open; a third, the joint reasoned statement, was resolved on September 2, 2026 and is covered in Latest News above.

• The joint reasoned statement, now published. On September 2, 2026 the management board and the supervisory board of Delivery Hero published their joint reasoned statement recommending shareholders accept the offer, backed by fairness opinions from J.P. Morgan and UniCredit. In a German takeover that document is where the target boards tell their own shareholders whether the price is adequate; both boards said it was.
• Merger control. The offer is conditional on clearances that are not listed in the press release. The transaction touches 50 markets on the Uber side and overlaps with Uber Eats in several of them, which is the reason 14 markets were carved out to SSW Partners in the first place.
• The acceptance count. Nothing is known until the bidder publishes the interim notices required during the acceptance period.

The part Uber is not buying

Delivery Hero agreed separately to sell its businesses in 14 markets to SSW Partners, a New York investment firm, for approximately $1.6 billion. Those are largely the markets where Uber Eats and Delivery Hero already overlap: foodora in Austria, Czechia, Norway and Sweden; efood in Greece; Foody in Cyprus; Glovo in Moldova, Poland, Portugal, Romania and Spain; PedidosYa in Chile and Ecuador; and Yemeksepeti in Tuerkiye. The sale is separate from the takeover offer but conditional on it closing, and Uber states it will not acquire control over those businesses.

How it is being paid for

Uber will fund the offer from existing cash and new debt, and has executed a committed bridge facility of approximately EUR 14 billion. The company states the structure is designed to keep its investment grade rating with gross leverage below 2 times, and that the existing capital allocation framework, buybacks included, is unchanged. For a company that generated $2.792 billion of free cash flow in a single quarter against $70 million of capital expenditure, the debt is serviceable on the arithmetic; what it removes is the option value of that cash for something else, for as long as the bridge is outstanding.

What the company says it gets

The combination takes the platform to 99 markets with combined pro-forma gross bookings of $236 billion in 2025, and nearly doubles the markets where Uber offers both mobility and delivery, from 34 to 58. Uber’s own argument for why that matters is a cross-selling one: it states that cross-platform users generate roughly three times the gross bookings and profits of single-product users. The company expects the transaction to be accretive to non-GAAP earnings per share upon close and high-single-digit percentage accretive by year three.

Uber has also committed to keep Delivery Hero’s Berlin headquarters and make no changes to its Berlin workforce until at least 2029, and to invest EUR 2 billion in Germany over five years. Those commitments are the political price of a foreign takeover of an MDAX company, and they are worth reading as a cost rather than as a courtesy.

What can go wrong, stated plainly

• The threshold is not met. The arithmetic looks comfortable, but comfort is not the same as a counted result.
• Merger control imposes remedies. Conditions attached to a clearance can subtract from the assets acquired or add obligations, and the carve-out to SSW shows the overlap was already recognised as a problem.
• The gap to settlement. Expected settlement in the second half of 2027 means the bridge, the integration planning and the regulatory files all run for a long time in a delivery market that has repriced twice in five years.
• Integration. Fifty markets, several brands and a workforce commitment in the largest of them. Uber has bought marketplaces before, but not at this scale in one document.

None of that is a view on whether the offer is good or bad for either set of shareholders. It is the list of things that decide whether the transaction described in the document is the transaction that eventually closes.

04 How The Marketplace Works: Bookings, Revenue, Take Rate

Three numbers describe a marketplace, and confusing them is the most common way to misread this company.

Gross bookings are the meter. They are the total value of what happened on the platform: the fare a rider paid, the food order a customer placed, the freight load a shipper booked, including taxes and fees where applicable. In the second quarter that was $58.022 billion. Uber does not keep this money. Most of it goes to the driver, the courier, the restaurant or the carrier.

Revenue is what stays. $14.191 billion, which is 24.5 per cent of gross bookings on the reported figures. That percentage is often called the take rate, and it has to be handled with care: revenue recognition differs by market and by offering, because in some jurisdictions Uber records the gross fare and in others only its own commission. A rising ratio can therefore reflect an accounting change in one country rather than a pricing decision.

Operating income is what survives the cost of running it. $1.890 billion on a GAAP basis, after paying for support operations, marketing, engineering, insurance and administration. That is 13.3 per cent of revenue and 3.3 per cent of gross bookings, both Merlintrader calculations.

Two derived figures make the scale concrete. Across 3.867 billion trips, gross bookings work out to about $15.00 per trip and revenue to about $3.67. And with 208 million monthly consumers, the average user took roughly 6.2 trips a month during the quarter. Neither number is disclosed by the company; both follow from dividing the disclosed ones, and they are stated here as arithmetic rather than as company metrics.

The point of the structure is that Uber sells access to a network rather than a product. Nothing that produces the $58 billion belongs to it: not the cars, not the kitchens, not the trucks. What belongs to it is the software that matches a request to a supplier in seconds, and the payment layer that settles it.

05 Second Quarter 2026 In Numbers

The reported quarter, with the same quarter of 2025 alongside.

Read the operating income line and the net income line as two different statements about the same company. Operating income rose 30 per cent because the platform grew and the cost of running it grew more slowly. Net income rose 77 per cent because of a $1.342 billion swing in a line that has nothing to do with rides, deliveries or freight.

The cost lines deserve a look of their own. Cost of revenue grew only 3 per cent against gross bookings up 24 per cent, which is the operating leverage the whole model is built on. General and administrative expense grew 40 per cent and research and development 24 per cent, both faster than revenue. In a marketplace, the first is a warning to watch and the second is usually a choice.

06 Mobility: Three Growth Rates From One Business

Mobility is the ride-hailing business, and in this quarter it produced the strangest pair of numbers in the release.

Gross bookings up 22 per cent, revenue up 1 per cent, segment profit up 28 per cent. Three different directions from one business in one quarter.

The explanation is in how revenue is recognised across markets. Where Uber acts as principal it books the whole fare as revenue and the driver payment as cost; where it acts as agent it books only its own commission. When the mix of markets or of business models shifts, reported revenue moves in a way that has little to do with how many rides happened or how profitable they were. Gross bookings up 22 per cent describes the activity. Segment operating income up 28 per cent describes the economics. Revenue up 1 per cent describes an accounting geography.

Segment operating income of $2.215 billion is 7.6 per cent of Mobility gross bookings, against 7.3 per cent a year earlier, a Merlintrader calculation. That ratio, profit against the money that flows through the segment, is the cleanest way to compare the segment with itself over time, because it is not disturbed by principal-versus-agent accounting.

07 Delivery: Closing The Gap With Mobility

Delivery is food, groceries and retail, and it is close to becoming the larger half of the company.

Delivery gross bookings are now $1.525 billion below Mobility, against a gap of $2.028 billion a year earlier. On the current growth rates the two would cross within a few quarters, though growth rates are not promises and the company has not guided to a crossover.

Segment operating income of $1.055 billion is 3.8 per cent of Delivery gross bookings, against 3.5 per cent a year earlier. That margin is half the Mobility figure and the reason is structural: a delivery involves a restaurant or a store as a third party, a courier, and a customer, and the value of the basket is split three ways before Uber takes anything. A ride is split two ways.

What makes the segment matter beyond its own profit is that it uses the same app, the same account, the same payment method and often the same courier network as the rest. A company that has already acquired a customer for one purpose and sells a second one to them is buying growth cheaply, which is what the 25 per cent increase in sales and marketing is meant to keep doing.

08 Freight: Small, Loss-Making, And Reported Differently

Freight is the third segment and the smallest, and its numbers are worth stating precisely because they are so often skipped.

Revenue slightly exceeds gross bookings, which looks like an error and is not: in Freight, Uber is generally the principal in the transaction, so it records the shipper’s payment as revenue and the carrier’s payment as cost, and the two definitions do not align the way they do in the other segments.

The segment lost $24 million in the quarter, an improvement of $2 million on a year earlier. Freight brokerage is a business with thin margins in the best conditions and negative ones in a soft trucking market, and the American freight market has been soft for a prolonged period. It is 2.7 per cent of group gross bookings and does not change the group result in either direction.

09 The Metric Change: From Adjusted EBITDA To Operating Income

Something changed in how Uber asks to be measured, and it is stated in the release itself: adjusted EBITDA is no longer the key metric used by management, replaced by non-GAAP operating income.

The difference between the two non-GAAP measures is depreciation, amortisation and, above all, stock-based compensation treatment. Moving from adjusted EBITDA to operating income narrows the distance between the number the company highlights and the number the accounts report: $2.143 billion against $1.890 billion of GAAP operating income, a gap of $253 million, where adjusted EBITDA sat $929 million above GAAP.

A company that voluntarily switches to a stricter measure is usually signalling that it no longer needs the looser one. That is a reasonable reading and it is not a fact, so it belongs in the interpretation column rather than in the numbers. What is a fact is that all three measures rose, and that the gap between the flattering one and the audited one narrowed.

10 The Stakes In Didi, Grab And Aurora

Uber holds shares in other companies, and those holdings are large enough to decide what the bottom line says in any given quarter.

Marketable holdings are revalued at their market price every quarter and the change runs through the income statement. That is why other income swung by $1.361 billion year over year and why net income rose 77 per cent while operating income rose 30 per cent. In a quarter when Grab and Aurora fall instead of rising, the same mechanism will run the other way, and it will have nothing to do with rides or deliveries.

One detail in the filing deserves its own line. Approximately 61 per cent of the Aurora Class A shares Uber holds are pledged as collateral for the 2028 exchangeable senior notes. A pledged asset is still owned and still marked to market, but it is not freely available: it is securing a debt, and its price affects the terms of that security.

The Aurora position changed after the quarter end

The table above is dated June 30, 2026 and one of its lines is already out of date. On August 17, 2026 Neben Holdings, a wholly owned subsidiary of Uber, sold 72,000,000 shares of Aurora Innovation Class A common stock to a financial institution in a block sale at $6.55 per share. That is approximately $471.6 million of gross proceeds, a Merlintrader multiplication of the two figures rather than a disclosed total. Uber reported the sale in Amendment No. 6 to its Schedule 13D on Aurora, filed on August 19, 2026 and signed by chief financial officer Balaji Krishnamurthy.

After the sale Uber holds 186,473,411 Aurora Class A shares, approximately 10.9 per cent of that class, calculated on the 1,708,146,085 Class A shares Aurora reported outstanding at July 22, 2026. The filing states that Uber periodically reviews the investment and may dispose of part or all of it, and that it is not in discussions with Aurora management about control or strategy. The $1.763 billion in the table is therefore the June 30 mark on a larger position than the one held today, and the September quarter will carry both a realised element and a smaller remaining stake. The roughly 61 per cent pledge against the 2028 exchangeable notes is the figure disclosed at June 30; the Schedule 13D amendment does not restate it.

The practical instruction for reading future quarters is simple. Start at operating income, not at net income. The stakes are real assets and their movements are real, but they describe the value of other companies rather than the performance of this one.

11 Cash Generation, Buybacks And Debt

The cash statement is where the asset-light model becomes visible.

Seventy million dollars of capital expenditure against $2.862 billion of operating cash flow means the business converts 97.6 per cent of the cash it generates into cash it can deploy. There is no fleet to replace and no property to build, so the money goes to buybacks, to acquisitions, or onto the balance sheet.

In this quarter it went mostly to buybacks: 6.9 million shares for $510 million, after 46.6 million shares for $3.5 billion across the first half. Roughly $15.7 billion of authorisation remains under the $20 billion programme the board approved in July 2025. There are 2,042,560,121 shares outstanding at July 31, 2026, and the company pays no dividend.

The balance sheet holds $4.870 billion of cash and $521 million of short-term investments against total debt of $12.723 billion, which rose from $10.521 billion at the end of 2025. A company generating close to $3 billion of free cash flow a quarter carries that debt comfortably; the reason to watch it is that the increase funded the buyback rather than the business.

12 Autonomous Vehicles: What The Filing Actually Says

The autonomous vehicle question is the one every reader arrives with, so here is precisely what the filings say and what they do not.

What the 10-Q states. That Uber has invested, and may continue to invest, substantial amounts in companies with which it partners to offer autonomous driving technology on the platform, and that these commercial partnerships currently support the development and deployment of autonomous vehicle technology capable of operating in the United States and globally.

The one partner named. Waymo. The filing says Waymo has introduced a commercialised ride-hailing fleet of autonomous vehicles on its own platform, in addition to a fleet of autonomous vehicles that it makes available through Uber’s platform. No city is named and no date is given in the document.

What the 10-Q does not contain. No mention of Nvidia, Nuro, WeRide or Pony.ai. Tesla and Zoox appear, but as competitors in autonomous development rather than as partners. Anyone reporting a partnership with those names is not reading it from the filing. Company press releases issued after August 5 are a separate source, and three of them name partners the filing does not.

What the press releases add, after the filing date

• Zagreb, August 19, 2026. Uber, Verne and Pony AI announced the launch of autonomous rides on the Uber app in Zagreb, which the release calls the first time riders in Europe can book an AV through Uber. Pony.ai supplies the autonomous driving system, Verne owns the fleet and operates the service, Uber provides the demand layer. A licensed operator sits behind the wheel at launch, with fully driverless operation described as a later phase.
• Dubai, announced August 19, 2026. Baidu’s fully driverless Apollo Go vehicles became available to riders on the Uber platform in Dubai, with New Horizon Luxury Transport as fleet operator.
• Tokyo, August 13, 2026. Uber Japan signed an operational partnership with the taxi operator Hinomaru Kotsu to run depot operations, maintenance and charging for an autonomous vehicle pilot scheduled to launch in late 2026, built on the March 12, 2026 memorandum with Wayve and Nissan. Japanese law requires an authorised taxi company to operate passenger transport, which is why the fleet sits with Hinomaru Kotsu rather than with Uber.

The pattern in all three is the same one the filing describes in abstract terms: someone else owns the vehicles and carries the operating risk, and Uber supplies the passengers. That is the position worth judging, because it is the one the company is actually taking.

Why the structure matters more than the headlines. Uber’s position is that of the demand layer. If autonomous fleets scale, the vehicle owner still needs someone to bring it passengers, handle payment, manage disputes and absorb the demand peaks, and Uber has 208 million monthly consumers already doing that. The counter-argument is equally clear: a fleet owner with enough scale can build its own demand layer, as Waymo is doing on its own platform, and then Uber becomes a competitor rather than a channel.

The Aurora stake sits underneath this debate. It was marked at $1.763 billion on June 30, 2026, with about 61 per cent of the Class A shares pledged against the 2028 notes, and it was reduced on August 17 by the sale of 72,000,000 shares at $6.55 described in section 10. It is a financial position in autonomy rather than an operating one, and the filing describes it that way.

13 The $1.8 Billion VAT File And Driver Classification

Three regulatory files carry real money, and two of them have a number attached.

The Dutch GDPR fine, August 21, 2026. The Autoriteit Persoonsgegevens, the Dutch data protection authority, imposed a fine of €824,990,000 on Uber for taking fully automated decisions about drivers. Between 2018 and 2022 software tracked driving behaviour and customer ratings and deactivated accounts, temporarily on a suspicion of fraud and permanently on persistently low ratings, with no human assessment and without informing drivers adequately. The regulator ruled this breached the GDPR prohibition on solely automated decision-making. The case reached the Dutch authority through the one-stop-shop mechanism because Uber’s European headquarters are in the Netherlands, after 171 French drivers complained to the Ligue des droits de l’Homme, which took the matter to the CNIL. The AP states the infringements have stopped. Uber has said publicly it strongly disagrees and will appeal, and points to current policies that include human review and a route to challenge a suspension. For scale: the amount is roughly 44 per cent of the $1.890 billion of GAAP operating income Uber booked in the second quarter alone, and it is the second-largest GDPR fine on record after the €1.2 billion levied on Meta. Nothing is payable while the appeal runs.

The United Kingdom VAT assessments. After Uber changed its UK operating model in March 2022, becoming the merchant of transportation, HMRC issued a series of assessments disputing the application of the VAT Order 1987 for the period from March 2022 to September 2024. The total claimed is approximately $1.8 billion, about £1.4 billion. Uber has paid the assessed amounts in order to appeal, recording them as receivable pending the outcome, and expects further assessments covering 2023 to 2025. From January 2, 2026 Uber UK ceased applying the VAT Order 1987 following a change in the law.

That is the largest single quantified regulatory exposure in the filing, and the mechanism is worth understanding: paying first and appealing later is normal in the United Kingdom, so the payment is not an admission and the receivable is not a certainty.

Driver classification. Three live matters appear. In California, the action brought in May 2020 by the Attorney General together with San Francisco, Los Angeles and San Diego remains pending for the period before Proposition 22, which passed in November 2020 and led to the preliminary injunction being dissolved in April 2021. In Switzerland, several social security authorities have classified drivers and couriers as employees; a Federal Court ruling of March 21, 2023 addressed 2014 drivers, a dispute with SVA Zurich for 2014 to July 2020 was resolved in the first quarter of 2025, and matters covering 2014 to 2022 and Geneva remain open. In France, URSSAF issued observations in December 2024 and an assessment in June 2025, which Uber has appealed.

Uber does not quantify the amounts accrued for these matters in the 10-Q. That absence is itself information: the company records them within accrued and other current liabilities without a separate figure, so an investor cannot size them from the public document.

14 Guidance For The Third Quarter

The company gave a range for the third quarter of 2026, in its own words.

Gross bookings of $58.25 billion to $60.25 billion, representing growth of 18 to 22 per cent year over year on a constant-currency basis. The release adds that the outlook assumes roughly a one percentage point currency headwind to total reported growth.

Non-GAAP earnings per share of $0.84 to $0.88, representing growth of 28 to 35 per cent year over year. The company states that this translates to adjusted EBITDA of $2.86 billion to $2.96 billion.

Two things follow from the shape of that guidance. The first is that the midpoint of the bookings range, $59.25 billion, is only 2.1 per cent above the $58.022 billion just reported, which is the normal seasonal step from the second to the third quarter rather than an acceleration. The second is that guiding to earnings per share rather than to a profit total quietly folds the buyback into the target: a shrinking share count helps the metric independently of the business.

Note also what is not guided. There is no revenue range, which is consistent with a company that has just told the market its revenue line is distorted by principal-versus-agent accounting across countries.

15 Uber Against Hertz And Avis: Who Owns The Car

This is the reason an internet company sits in a travel section next to two car rental groups. All three take a person who needs a vehicle and connect them to one. Everything after that is different, and the differences are the whole lesson.

The asset is the difference. Hertz and Avis buy vehicles, finance them with debt secured against those vehicles, rent them by the day and sell them a year or two later at a price nobody controls. Their profit is the gap between what a car earns each month and what it loses in value each month, multiplied by how much of the fleet is out on rent. Their balance sheets carry $18.7 billion and $25.9 billion of debt and negative book equity, because the assets are financed and the accumulated losses are real.

Uber never buys the car. The driver does. The capital, the depreciation, the insurance on the vehicle, the residual value risk and the financing all sit outside the company. What Uber supplies is demand, matching and payment, and what it keeps is a share of each transaction. That is why $70 million of quarterly capital expenditure supports $58 billion of transactions, and why equity is positive $27.3 billion instead of negative.

The trade-off is not free. A company that owns nothing controls nothing. Hertz can decide the size of its fleet, the price of a rental day and which cars to buy; Uber cannot make a driver work on a wet Tuesday, and it competes for that driver with every other platform. The rental companies own a depreciating asset and the pricing power that comes with owning it. Uber owns neither.

And the regulatory bill lands in different places. A rental company’s risk is the used car market. Uber’s risk is whether a court decides its drivers are employees, and whether a tax authority decides its model owes VAT on the whole fare rather than on the commission. The second is already a $1.8 billion number in the United Kingdom.

16 Risks And Red Flags

The bottom line moves for reasons unrelated to the business. Net income of $2.394 billion includes roughly $1.6 billion pre-tax from revaluing stakes in Didi, Grab and Aurora. The same mechanism reverses when those shares fall.

Revenue is not a clean measure. Mobility revenue grew 1 per cent while its gross bookings grew 22 per cent, because principal-versus-agent accounting differs by market. Comparisons across quarters and against competitors need gross bookings and segment operating income, not revenue.

Regulatory exposure is quantified in one place and unquantified in several. The United Kingdom VAT assessments total approximately $1.8 billion, paid pending appeal. The driver classification matters in California, Switzerland and France are disclosed without amounts.

Cost lines are growing faster than revenue. General and administrative expense rose 40 per cent and research and development 24 per cent, against revenue up 12 per cent.

The autonomous question cuts both ways. Waymo makes a fleet available through Uber’s platform and also runs its own. A supplier that can reach customers directly is not permanently a supplier.

Debt rose to fund returns. Total debt went from $10.521 billion to $12.723 billion in six months while $3.5 billion of stock was repurchased.

A pledged stake is not a free one. About 61 per cent of the Aurora Class A shares held were pledged as collateral for the 2028 exchangeable notes at June 30, 2026; the position was reduced by the block sale of August 17 and no filing restates the ratio since.

Competition is structural. Drivers and couriers work on several platforms at once, and consumers switch at the price of one tap.

17 Bottom Line

Uber earned $1.890 billion of operating income on $58.022 billion of transactions in three months, and spent $70 million on property and equipment to do it. That sentence is the entire investment case and the entire risk: enormous throughput, almost no assets, and a profit that depends on keeping a share of other people’s work.

Next to Hertz and Avis the contrast is instructive rather than flattering. The rental companies own the fleet, carry the debt that finances it and take the residual value risk, and both report negative book equity as a result. Uber owns none of it, reports $27.3 billion of positive equity, and converts almost all of its operating cash into free cash flow. In exchange it gives up control of supply and inherits a different kind of exposure, the kind decided by courts and tax authorities rather than by used car prices.

What to watch, in order. Gross bookings and segment operating income, not revenue, because revenue is distorted by accounting geography. The gap between Delivery and Mobility bookings, now $1.525 billion and closing. Operating income rather than net income, so the equity stakes do not do the talking. The United Kingdom VAT appeal, which carries approximately $1.8 billion. And the language on autonomous partnerships in each new filing, because today it names exactly one partner and gives neither cities nor dates.

Related Research On Merlintrader

These pages sit alongside it in the Merlintrader travel section.

• Hertz Global Holdings ($HTZ) Stock Hub — the opposite model: 539,118 vehicles owned, $18.7 billion of debt and negative book equity.
• Avis Budget Group ($CAR) Stock Hub — 664,638 vehicles, 35.3 million shares and a $3.4 billion equity deficit.
• Merlintrader Travel Pub — the travel index, with every airline, cruise, hotel and mobility hub and its own update date.