Uber has tripled its free cash flow to roughly $10 billion a year over the past three years. Yet the stock trades below where it did when Shawn O'Malley and Daniel Mahncke first pitched it on the show 15 months ago, while Waymo just raised money at a $126 billion valuation — nearly the same market cap as all of Uber, on a fraction of the revenue and no profit at all.
"Waymo is either grossly overvalued or Uber is grossly undervalued."
O'Malley and Mahncke have carried Uber as a live position in their own portfolio since that first pitch, and they're revisiting the call now that Waymo's expanding valuation, its move to end exclusive city partnerships with Uber, and a wave of more than 20 rival autonomous-vehicle deals have all landed inside a single earnings cycle.
I listened to the full episode so you can skip it. 1 hr 5 min of audio, 17 minutes of reading.
Here are the 11 takeaways that matter.
🎙️ Hosts: Shawn O'Malley and Daniel Mahncke, who co-host The Intrinsic Value Podcast and manage the show's real-money Intrinsic Value Portfolio, in which Uber has been a holding since their first pitch 15 months ago
📰 Published: 9 September 2026 on YouTube
🔴 YouTube | ⏱️ 1 hr 5 min | ✅ Time saved: 48 min
Key Takeaways
Uber's operating margin swung from -43% to +12% in under six years, a 55-point turnaround
Advertising alone now runs at a $2 billion-plus annual pace, growing more than 50% a year
Post-pandemic insurance inflation is turning into a margin tailwind
This year's carrier renewals came in at low single-digit increases, the most benign in years
Uber One's 50 million members now drive roughly half of Uber's gross bookings
Waymo raised $16 billion at a $126 billion valuation — nearly Uber's entire market cap on a fraction of the revenue
Uber's flexible driver supply is built for demand that swings roughly 4-to-1 in a single day, which a fixed robotaxi fleet can't match
Waymo is ending its exclusive Uber partnership in Austin and Atlanta when its contracts expire in early 2028
Only about 9% of Uber's profit is genuinely exposed to near-term robotaxi competition, rising to 18% once the suburbs are included
Uber has gone from 14 autonomous-vehicle partners a year ago to more than 20 today
Nuro, Lucid, Rivian, Nvidia, Baidu, WeRide and Pony.ai are all now running or launching on Uber's network
Uber wants outside capital, not its own balance sheet, to own the robotaxi fleets
It has already committed roughly $10 billion to leases, investments and offtake deals with AV builders
Uber's $14.8 billion bid for Delivery Hero is the largest acquisition in its history
It expands the number of markets where Uber runs both rides and delivery from 34 to 58
The board raised its buyback authorization to $20 billion while still funding every AV partnership and the Delivery Hero deal
1. Uber In 60 Seconds
Uber is really three businesses under one marketplace model: mobility (the rides business), delivery (Uber Eats, now covering groceries and general retail as well as restaurants), and a smaller freight brokerage that O'Malley and Mahncke set aside because it isn't material to the investment case. Uber owns no cars and, in most markets, employs no drivers — they're contractors. The company matches riders with drivers and diners with couriers across roughly 70 countries and takes a cut of every transaction, called the take rate.
Roughly 200 million monthly riders and eaters spent about $190 billion through Uber's apps over the past year — what the company calls gross bookings. Uber's own revenue is roughly a 20% slice of that figure, after paying out drivers and restaurants. O'Malley said Uber not owning the cars or restaurants on its platform is "really the point of the entire Uber thesis" — especially against a company like Waymo that has to own the fleet itself.
2. The Margin Inflection
Mahncke recalled his original hang-up on Uber: the business looked like it would always carry Walmart-style economics rather than Google's, because every incremental ride means paying a driver and covering insurance. Operating margins were about 6% when they first covered the company; they have since roughly doubled on a quarterly basis. From 2020 through today, Uber's operating profit margin has swung from -43% to +12% — a 55-percentage-point turnaround in under six years for a company already doing billions in revenue.
Three things are pushing margins further, per O'Malley:
Advertising, a business that didn't exist a few years ago, now runs at more than $2 billion in annual revenue and is growing more than 50% a year — ad dollars are close to pure profit against the core business. Placements include in-app ads on the map, sponsored restaurant listings, and search-style ads where restaurants bid for placement.
Uber Eats' expansion beyond food. Uber partnered with Ulta Beauty — once the hosts' first-ever holding in the Intrinsic Value portfolio — to deliver cosmetics within 45 minutes, a speed advantage Mahncke argued Amazon "is not designed to match" today. Uber now has more than 1.5 million merchant partners globally, spanning grocery, cosmetics, sporting goods, alcohol and florists.
Insurance, covered next.
3. Insurance Turns Tailwind
Since the pandemic, sharply higher vehicle prices had pushed up insurance premiums, and that became a major cost headwind for Uber — one Mahncke said was hard to fully appreciate the first time they researched the company. The US, still Uber's biggest market, saw the steepest increases.
That's now reversing. Uber renegotiates its insurance rates every March, and this year's renewals came in at low single-digit increases — the most benign in years — on top of hundreds of millions of dollars in savings from state-level insurance reforms. For a few years, that inflation had actually slowed Uber's US rides growth even as Europe and Latin America kept growing around 30%, because Uber passed the higher costs on to riders. Mahncke noted that Uber's own management "described it as an accidental AB test on price elasticity with America as the test group." With insurance now normalized, Uber is passing the savings back into lower prices, and US growth is reaccelerating.
Looking ahead, the hosts see autonomy compounding the tailwind: a fleet that's part drone or robot doesn't carry driver-style insurance costs. Uber just announced a partnership with drone-delivery company Zipline to bring drone delivery to millions of Americans by the end of 2029; Zipline already operates across four continents with 135 million autonomously flown miles and 2.7 million deliveries. Uber's press release said it is "building the world's most flexible hybrid delivery network, seamlessly integrating couriers, sidewalk robots, and drones to match every delivery with the best mode of transportation." A Zipline co-founder added: "It's becoming part of everyday life. Every great transportation revolution has changed where people live, how businesses operate, and how economies grow. Together with Uber, we are taking the next step toward building a world where getting what you need is as fast and as effortless as sending a text, no matter where you are."
4. Uber One's Flywheel
Uber One, the company's membership program, now has 50 million members — up 14 million from a year earlier. In the US it costs about $10 a month and includes free Uber Eats delivery and 6% credits on rides; Mahncke pays roughly €5 in Germany and isn't sure he gets the same perks. Subscription revenue itself is immaterial next to a $150 billion company, but the real value is loyalty: Uber One members now drive roughly half of Uber's total gross bookings and about 23% of delivery bookings. Dara Khosrowshahi, Uber's CEO, has compared it to Netflix — pay one fee, and the platform with the most content wins, except Uber's "content" is cars, couriers, restaurants, groceries, hotels and now parking.
In February Uber acquired parking app SpotHero, covering more than 13,000 garages — a small business on its own, Mahncke said, but another reason to keep the app open. The bigger prize is cross-selling: multi-product users spend more than three times what single-product users do, letting Uber grow existing customers rather than chase expensive new ones.
5. Waymo By The Numbers
In February, Waymo — Alphabet's self-driving subsidiary — raised $16 billion at a $126 billion post-money valuation, from backers including Sequoia and Andreessen Horowitz, with Alphabet remaining the majority owner. Waymo has now driven more than 100 million fully autonomous miles, completed 15 million paid rides in 2025 alone (triple the prior year), and is running on the order of 500,000 rides a week. It plans to expand to more than 20 new cities this year, including Tokyo and London.
Set against that: Uber does more than 3 billion trips a quarter and generates $10 billion a year in free cash flow — and the private market values the two companies at roughly the same price. "Waymo is either grossly overvalued or Uber is grossly undervalued," O'Malley said. He called Waymo's technology "incredibly impressive" but argued there's a difference between novel technology and a scaled, profitable business that can support a nine-figure valuation — noting that a chunk of Waymo's own completed rides were actually booked through the Uber app.
6. Why Fixed Fleets Lose
The hosts' core structural argument, replayed from their original pitch: Uber's flexible, contractor-driven supply naturally adjusts to demand, while a fixed robotaxi fleet cannot. "The peak-to-trough ratio within a single day's demand for rides is something like 4:1," O'Malley said, and a fleet sized for peak demand sits idle in the troughs, while a fleet sized for the troughs leaves riders stranded at rush hour. In Austin, Uber's own data found that Waymo vehicles operating on its network were busier than 99% of human drivers — evidence, the hosts argued, of just how much demand-aggregation liquidity Uber's network provides.
Mahncke added that even today, opening the Waymo app directly in California can mean an 18-minute wait versus a few minutes on Uber — a gap he doesn't think is lost on Waymo. Longer term, he expects Waymo will keep wanting to plug into Uber's platform if it proves the best way to monetize its hardware, even as it tries to go it alone for now. What gives him pause is that Waymo, backed by Alphabet's capital, doesn't need to be efficient or profitable in the near term — it can tolerate poor fleet utilization for years if that's what displacing Uber takes. Mahncke pointed out that Uber wrote that exact playbook itself, subsidizing rides for a decade before turning profitable.
7. The Uber-Waymo Breakup
Uber stock hit a 52-week low in late July after Waymo formally notified Uber it will end their exclusive partnerships in Austin and Atlanta — two cities where Waymo's robotaxis had been available only through the Uber app — when the existing contracts expire in early 2028. Waymo plans to launch its own standalone app in both cities. The hosts stressed this doesn't mean every city partnership is fraying, but tensions are real and intensifying.
The Financial Times reported the two companies have begun lobbying regulators for opposing frameworks. Uber wants rules that enshrine hybrid networks of human drivers and robots working together; in New Jersey, its lobbyists proposed requiring any robotaxi platform to have human drivers provide at least 85% of rides during a three-year pilot program — clearly aimed at Waymo. Mahncke said "things have actually gotten pretty petty": Waymo has accused Uber of not taking good care of its vehicles, and Uber has pointed back at safety issues on Waymo's end.
8. How Exposed Is Uber, Really
Uber's CFO has said that over the next five years, AVs are "relatively immaterial to Uber's volume" — and the numbers back that up. Waymo does about 500,000 rides a week; Uber does 40 million trips a day. The entire global autonomous-vehicle industry combined does roughly 50 million trips a year, while Uber adds about 3 billion trips a year just in growth — meaning autonomous rides are around one-tenth of 1% of global ride-share volume today. Even the fastest-growing AV deployments are at best tripling their volumes annually, versus the 9-to-10x annual growth Uber itself posted in its early hypergrowth years.
Mahncke walked through the exposure math: mobility is roughly 60% of Uber's operating profit, the US is roughly 60% of mobility, and the top 20 US cities — the only places robotaxis can realistically operate at scale in the foreseeable future — account for about a quarter of US mobility profit. That works out to roughly 9% of Uber's total profit genuinely exposed to robotaxi competition in the near-to-medium term (O'Malley put that window at five to ten years), rising to about 18% if the suburbs, one of Uber's fastest-growing areas, are included.
The evidence so far cuts against a zero-sum read. Austin and Atlanta — two of Waymo's showcase markets — have actually been among the fastest-growing Uber markets in the US. In San Francisco, Waymo's most mature market with a genuine 15-to-20% share of rides, Uber's own trip growth still accelerated in 2025. The hosts' explanation: robotaxis are expanding the category, converting people who used to drive themselves or take transit, faster than they're taking share from Uber. Ride-share today is still less than 1% of the roughly 3 trillion miles Americans drive every year, leaving a lot of room for that number to grow as autonomy pulls the cost of a ride closer to owning a car.
9. Betting On Every Horse
Uber's response to the AV wave has been to sign as many partners as possible rather than bet on one winner. A year ago it had 14 autonomous-vehicle partners; today it has more than 20, and the deals have moved from abstract to concrete:
Nuro and Lucid — a minimum of 35,000 Lucid Gravity SUVs running Nuro's self-driving system, launching commercially in the San Francisco Bay Area later this year and in Houston by mid-2027. Rental company Hertz has spun up a new affiliate, Oro Mobility, to handle charging, cleaning, maintenance and depots for these fleets, and has already leased a 50,000-square-foot depot in Houston.
Rivian — 10,000 autonomous R2 robotaxis starting in San Francisco and Miami in 2028, with an option to scale to 50,000 vehicles across 25 cities by 2031. Uber is investing up to $1.25 billion in Rivian tied to technical milestones.
Nvidia — launching its full self-driving software stack on Uber in Los Angeles and San Francisco in the first half of 2027, targeting 28 cities globally by 2028.
WeRide — already running a fully driverless fare-charging service with Uber in Dubai and Abu Dhabi, with Zurich and Madrid coming soon.
Baidu's Apollo Go — coming to Uber in Dubai; Baidu's robotaxis reportedly cost under $30,000 to build.
Pony.ai — partnering with European fleet operator Verne, which Uber is also investing in, to launch Uber's first commercial robotaxi service in that market.
Tesla and Amazon's Zoox remain the wild cards outside that coalition. Tesla is scaling its own robotaxi service and, Mahncke noted, has the manufacturing base to potentially flood the market with lower-cost purpose-built vehicles, while still refusing to use LiDAR sensors — whose cost, O'Malley noted, has fallen from about $100,000 to roughly $25,000 per vehicle in a few years. Zoox will run vehicles on Uber's network in Las Vegas and Los Angeles but is also building its own app; Amazon's devices chief reportedly asked, "why should we give that up," when discussing owning the customer relationship directly. Amazon has about 260 million Prime members against Uber's 50 million Uber One members and a long history of being willing to burn cash for market share.
Uber is also selling picks and shovels: a new unit called Uber Autonomous Solutions offers insurance, customer support, fleet management and remote assistance to any AV operator, and Uber is putting sensor kits on its regular human-driven cars — which see 40 million trips a day worth of edge cases — to collect and sell up to 2 million miles of training data a month to AV partners by year-end.
10. Who Owns The Fleet
Per Uber's CFO, the company sees the autonomy ecosystem as five layers: the consumer-facing marketplace (Uber itself), the AV software developers (Nuro, Waymo), the automakers building the cars, the fleet operators running depots and charging (Hertz), and — the piece the hosts flagged as most interesting — third-party financing that actually owns the vehicles and carries the balance-sheet risk. The model Mahncke described is closer to how Marriott runs hotels without owning the real estate than to Uber's traditional setup, where individual drivers own their own cars.
For now, Uber is taking on more capital intensity than usual to get there — signing offtake agreements to buy cars off assembly lines, leasing depots, and investing directly in Verne, Rivian and Nuro. Altogether, roughly $10 billion has been committed to AV builders so far — a sum the hosts said would have looked enormous a few years ago but now reads as modest next to what the largest technology companies are spending on AI infrastructure.
11. The Delivery Hero Deal
While attention was on the Waymo fallout, Uber quietly pursued the largest acquisition in its history: Delivery Hero, a Berlin-headquartered food-delivery company (with, oddly, no business left in Germany after selling its domestic operations) that owns leading platforms across the Middle East, Asia and Latin America, including Talabat in the Gulf and Baemin, the dominant app in South Korea.
Uber built up roughly a 25% stake in Delivery Hero before launching a formal offer on July 16th valuing the company at about $14.8 billion (roughly $13.7 billion net of the stake it already owned). Dutch investment firm Prosus irrevocably committed to tender its shares, pushing Uber past 50% ownership and effectively guaranteeing the deal will close, regulatory approval permitting. To pre-empt antitrust concerns, Delivery Hero is selling its operations in 14 overlapping markets to a third party for about $1.6 billion; the deal should close in the second half of next year.
The logic, per O'Malley: the deal is about expanding the number of markets where Uber can sell its full range of products. Uber currently runs both rides and delivery in 34 markets; after the deal, that jumps to 58, opening the door to selling Uber One and cross-selling rides and food in about two dozen new markets alongside 50 million new Delivery Hero customers. It also marks a shift from Uber's earlier playbook of taking passive stakes in local winners it couldn't beat outright — DiDi in China, Grab in Southeast Asia — to a full acquisition and integration.
There's a real margin case, too: Delivery Hero has a higher take rate than Uber but lower margins, largely because it spends far more on tech as a share of bookings — it lacks Uber's single global delivery platform. Delivery Hero also monetizes about 3% of its gross merchandise value through advertising, ahead of Uber's own delivery ad penetration; if Uber's ads business converges toward that level across a combined delivery business exceeding $100 billion in bookings, O'Malley said that implies several billion dollars of incremental, high-margin revenue over four or five years. Mahncke added a strategic angle: the deal is also an autonomy hedge, since AV adoption will roll out more slowly in most international markets than in the US, diversifying Uber's revenue away from its most AV-exposed market.
Bonus Insights
Kroger, the second-largest US grocer, added roughly 2,700 stores to Uber Eats nationwide in January — notable to Mahncke because he'd expected large grocers like Kroger, Walmart or Target to want to control the customer relationship directly rather than plug into a middleman.
Uber partnered with Expedia to sell more than 700,000 hotel properties inside the Uber app, with Uber One members earning 10% back in credits (Mahncke wasn't sure the offer extends outside the US).
Uber acquired Blacklane, a premium chauffeur service operating in more than 500 cities, complementing the luxury end of its offering.
Uber added a "woman preferences" feature letting women riders match only with women drivers, which the hosts see as a meaningful safety-driven reason for more people to use the app.
The board's buyback program has grown from $7 billion — a size that drew criticism, including from the Financial Times, that Uber was underinvesting in growth — to $20 billion, funded without stretching the balance sheet even while Uber makes every AV and Delivery Hero commitment described above. Reading from a letter Uber CEO Dara Khosrowshahi wrote defending the earlier buyback, O'Malley quoted: "I believe Uber's best days are ahead. We have a large utility-like business that is still in the early days of penetrating its market." Mahncke read the next line: "We are taking the humble investment route of dollar cost averaging over what we hope will be multiple years."
O'Malley closed with a quote from Ben Graham — the father of value investing and Warren Buffett's mentor — that he said the hosts return to often: "In the short run, the market is a voting machine, but in the long run, it's a weighing machine." Their bottom line is that Uber's demand-aggregating network, not any single robotaxi maker, is still the best way to convert a spiky, seasonal ride-and-delivery market into steady profit — and that the market, once it weighs rather than votes on Uber, will eventually price that in.
Products, Companies & Tools Mentioned
Uber (The episode's subject — a ride-hailing and delivery marketplace the hosts argue the market is mispricing against the autonomous-vehicle threat)
Waymo (Alphabet's self-driving subsidiary, raised $16B at a $126B valuation and is ending exclusive Uber partnerships in Austin and Atlanta)
Delivery Hero (Berlin-based food-delivery company Uber is acquiring for about $14.8B, its largest deal ever)
Zipline (Drone-delivery company partnering with Uber to bring drone delivery to millions of Americans by 2029)
Rivian (Building up to 50,000 autonomous R2 robotaxis for Uber's network; Uber is investing up to $1.25B in the company)
Nvidia (Launching its self-driving software stack on Uber in Los Angeles and San Francisco, targeting 28 cities by 2028)
Nuro and Lucid (Partnered on 35,000 Lucid Gravity SUVs running Nuro's self-driving system for Uber's network)
Hertz (Its new Oro Mobility affiliate will run charging, cleaning and depot operations for Uber's AV fleets)
WeRide, Baidu's Apollo Go, and Pony.ai (International autonomous-driving partners bringing driverless service to Uber in Dubai, Abu Dhabi and beyond)
Tesla and Zoox (Amazon's Zoox runs on Uber's network in Las Vegas and Los Angeles while also building its own app; Tesla is scaling a competing robotaxi service outside Uber's coalition)
Ulta Beauty (The hosts' first-ever portfolio holding; now an Uber Eats delivery partner used as an example of Uber's speed advantage over Amazon)
Amazon (Uber's rival for on-demand convenience via Prime and Zoox; has about 260 million Prime members against Uber's 50 million Uber One members)
SpotHero (Parking app Uber acquired in February, covering more than 13,000 garages)
Kroger (Added roughly 2,700 stores to Uber Eats nationwide in January)
Expedia (Uber partnered with it to sell more than 700,000 hotel properties inside the Uber app)
Blacklane (Premium chauffeur service Uber acquired, operating in more than 500 cities)
Books & Resources Mentioned
The Intrinsic Value Podcast's original Uber episode, from 15 months before this one (The hosts replayed clips of their original pitch to test whether its thesis on hybrid ride-hailing networks still holds up)
Lewistown Capital's Ride or Die: The Self-Driving S-Curve (Cited in the show's own episode notes as further reading on the autonomous-vehicle competition discussed throughout)
If this was worth your time, send it to someone who follows the name.
Get the latest market chatter as it happens: