Bet-the-Company

The record

1. 2021

   Meta

   Held

   Whether to rebrand as Meta and pour billions into the metaverse

   • Rebrand and fund metaverse — Rename to Meta and invest many billions in Reality Labschose this
   • Stay focused on core apps — Keep the Facebook name and concentrate on the ad business

   Stated reasonZuckerberg projected the metaverse would reach a billion people and generate hundreds of billions in commerce within a decade

   What happened

   • 2023-03 layoffs: second wave of about 10,000 more in March 2023
   • 2025 Reality Labs cumulative operating loss: more than $83.6 billion cumulative losses through end of 2025
   • 2025 Family of Apps operating income: crossed $100 billion in 2025
   Read the full analysis →
2. 2023

   Moderna

   Held in part

   As COVID revenue collapsed, spend the windfall on heavy R&D to fund the pipeline rather than harvest the dying product

   • Fund the pipeline hard — Keep spending more than revenue on R&D to prove the platformchose this
   • Harvest the dying product — Cut spend and manage COVID vaccine decline for cash

   Stated reasonModerna is a platform company that used a windfall to buy a decade of R&D runway to prove mRNA beyond COVID

   Committed$4.543 billion (2024 R&D spend, more than its entire $3.236 billion total revenue)

   What happened

   • 2024 total revenue: $3.236 billion total revenue, a 53% drop
   • 2024 RSV bridge product sales: mRESVIA sold $25 million in all of 2024
   • 2024 net loss: net loss of $3.56 billion
   Read the full analysis →
3. 1966

   Pan Am

   Held

   Whether to commit the balance sheet to 25 jumbo jets on a bet of perpetual international growth

   • Order 25 jumbo jets — Commit the balance sheet to 25 Boeing 747s on continuous demand growthchose this
   • Build domestic feeder base first — Establish domestic ballast before committing to jumbo capacity

   Stated reasonThe 747 promised operating costs per seat up to 40% lower than existing subsonic jets

   Committed$525 million (For 25 Boeing 747s ordered in 1966)

   What happened

   • 1969 Annual loss: Pan Am lost $26.42 million in 1969, with the 1966 jumbo order blamed as a cause
   • 1985-04 Divestiture of core assets: Pan Am sold its entire Pacific division to United Airlines for $750 million
   • 1991-01-08 Bankruptcy: Pan Am filed for Chapter 11
   Read the full analysis →
4. 2000

   Emirates

   Held

   Dominate a niche nobody else can profit from, or diversify into what everyone else is chasing

   • Dominate the A380 niche — Absorb the supply of a plane only Emirates' hub geometry could profitably flychose this
   • Diversify into efficient twins — Chase the point-to-point twin-engine market everyone else was pursuing

   Stated reasonEmirates ran the purest hub-and-spoke machine, so more premium seats per slot at a slot-constrained megahub extracted more revenue than any rival could, and no competitor could copy it

   What happened

   • 2020-12 share of profit from A380: 80% of profits came from the A380 before the pandemic (85% cited in a separate interview), an unaudited operator claim
   • 2019-02-14 order book pivot: Emirates cut its A380 order from 162 to 123, canceled 39 planes, and swapped them for 40 A330neos and 30 A350s
   • 2021-12-16 fleet delivered: Emirates took delivery of its 123rd and final A380, the last A380 ever built
   Read the full analysis →
5. 2022

   Ford

   Reversed

   How large a factory to build for the electric F-150 Lightning?

   • Build for 150,000 units — Size the plant to the optimistic adoption curve and reservations.chose this
   • Size to actual orders — Pour capacity only against demand that cost customers something.

   Stated reasonNearly 200,000 reservations looked like multi-year demand at a $40,000 price Ford believed it could deliver.

   What happened

   • 2023 Lightning annual deliveries: 24,165 units in 2023, against a planned run-rate of 150,000
   • 2023 electric-pickup segment share: held 72% of the electric-pickup segment in Q4 2023
   • 2024 production plan: Ford halved its 2024 production plan from roughly 3,200 to 1,600 trucks a week
   Read the full analysis →
6. 2000

   Airbus

   Reversed

   Whether to build a 500+ seat four-engine superjumbo betting on a hub-and-spoke future

   • Build the A380 superjumbo — Bet on hub-and-spoke with a 500+ seat four-engine jetchose this
   • Back smaller efficient twins — Point-to-point flying with smaller twin-engine widebodies

   Stated reasonAirbus read the future as a hub-and-spoke world where congested mega-airports meant the way to win was carrying more people per slot

   Committed€9.5 billion (Launch cost at the December 2000 board vote; the budget later grew past €15 billion.)

   What happened

   • 2006-10 break-even point revised: break-even revised from 270 to 420 aircraft after €4.8 billion in cost overruns
   • 2019-02-14 program ended: Airbus announced the end of A380 production after Emirates cut 39 orders
   • 2021-12 total deliveries: 251 passenger A380s delivered to 14 customers, all delivered as of December 2021
   Read the full analysis →
7. 2004

   Apple

   Held

   Whether to bet Apple on a keyboard-less touchscreen pocket computer

   • Bet on touchscreen architecture — Ship a keyboard-less pocket computer and rename the company around itchose this
   • Wait for a polished version — Hold the product until price, apps, and carrier terms were fixed

   Stated reasonThe core architecture couldn't be retrofitted, so Apple committed to it before any market evidence

   What happened

   • 2007 share of Apple revenue: iPhone represented only ~0.5% of Apple's total revenue that year, with 1,389,000 units sold
   • 2007-09-06 price cut and apology: Apple cut the 8 GB iPhone from $599 to $399 and Jobs published an open letter of apology
   • 2008-07-11 App Store launch: The App Store launched July 11, 2008, more than a year after the phone shipped
   Read the full analysis →
8. 2021

   Beyond Meat

   Held

   Whether to scale capacity and brand spend against the 2021 peak demand

   • Scale into peak demand — Pour capital into capacity and brand at the 2021 high-water markchose this
   • Scale to confirmed repeat demand — Build capacity only against defensible repeat purchase, not the spike

   Stated reasonThe company believed the growth curve was durable demand and scaled to meet it

   What happened

   • 2022 Net loss: $366.1 million net loss in 2022 on falling revenue
   • 2025 Net revenue: FY2025 revenue was $275.5 million, more than 40% below peak
   • 2025 Stock price: The stock had fallen over 99% from its IPO-era peak
   Read the full analysis →
9. 2021

   Meta

   Held

   Whether to bet the company on building the metaverse as the next computing platform

   • Bet the company on metaverse — Spend many billions building an owned next-generation platformchose this
   • Stay a platform tenant — Remain dependent on Apple's and Google's platforms

   Stated reasonOwning the next computing platform would end Meta's dependence on Apple's and Google's platforms

   What happened

   • 2026 Reality Labs quarterly operating loss: lost $4.03 billion against $402 million of revenue
   • 2026 cumulative Reality Labs losses: lost a total of $83.5 billion over 21 quarterly reports
   • 2021 operating profit reduction: reduced 2021 overall operating profit by approximately $10 billion
   Read the full analysis →
10. 2022

    Eli Lilly

    Held

    Whether to pursue an obesity label for tirzepatide alongside its diabetes approval

    • File tirzepatide for obesity — Pursue a weight-management label for the same molecule via rolling NDA.chose this
    • Stay diabetes-only — Keep tirzepatide marketed solely as a type 2 diabetes drug.

    Stated reasonLilly wagered that hitting two gut hormones at once — GIP and GLP-1 — would drive deeper weight loss than single-agonist rivals.

    What happened

    • 2024 2024 revenue vs guidance: Full-year 2024 revenue ~$45 billion, roughly $400 million (about 3%) below the guidance range
    • 2025 Best-selling medicine status: Tirzepatide passed Merck's Keytruda to become the world's best-selling medicine
    • 2025 Q3 2025 tirzepatide revenue: Mounjaro $6.515B (up 109% YoY) and Zepbound $3.588B (up 185% YoY)
    Read the full analysis →
11. 2023

    Microsoft

    Held in part

    How to structure and disclose its multibillion-dollar OpenAI investment

    • Structure as compute lock-in — Route funding as Azure credits and IP rights rather than pure equitychose this
    • Buy majority equity control — Take a larger ownership stake and capacity exclusivity in OpenAI

    Stated reasonThe bet was never about owning OpenAI but about being the infrastructure OpenAI runs on

    Committed$13 billion total commitment, $11.6 billion funded (partly in Azure credits)

    What happened

    • 2024-10-30 SEC disclosure of the commitment: The $13 billion total commitment figure first appeared in a 10-Q filed October 30, 2024
    • 2025-10 net income hit from equity-method loss: In Q1 FY2026 the line item cut Microsoft's net income by $3.1 billion
    • 2025-10 revised partnership terms: Stake fell to ~27%, IP rights extend to 2032, OpenAI committed to $250 billion more Azure
    Read the full analysis →
12. 2023

    Apple Vision Pro

    Reversed

    Price Vision Pro as a developer-platform filter or as a consumer device?

    • Price as developer filter — Set a high price to admit committed developers and studios.chose this
    • Price as consumer device — Sell at scale to millions of everyday buyers.

    Stated reasonA new platform is worthless empty; a high price kept out casual buyers and let in committed developers.

    What happened

    • 2024-02 returns from loyal buyers: within two weeks of launch some of Apple's most dedicated customers were returning the device
    • 2024 US unit sales: fewer than 100,000 units sold in the US by mid-2024, with demand stalling
    • 2026-04 program status: M5 failed to revive interest, Vision Pro team redistributed, and Vision Air canceled
    Read the full analysis →