Fortune 500 · 2026 edition · published Jun 3 · already inside
Every firm on the new 2026 list, the one where Amazon just ended Walmart's 13-year run at #1. How they make money. How a day in ops runs. How their moat holds against AI.
Clearing layer for ~$15T in annual payments. Banks can't replace it without each other.
FIELD NOTE ▸ COSTCO02
The moat
Members pay $65/yr just to shop. The membership is the margin.
FIELD NOTE ▸ APPLE01
The moat
Owns silicon, OS, store, and clearing. Taste sits on top of all of it. Each layer reinforces the next.
What you can't copy
Twenty years of compounding vertical integration.
the abstract
This is the whole product, shown on four firms. The other 496 get called the exact same way.
Each firm in the dataset resolves to a two-line structural verdict: what holds against the AI wave, and where the margin moves. If these read like value to you, the four questions behind every call are further down. If they don't, you have your answer in 30 seconds.
Walmart
retail / physical infrastructure
Money flow
10,500 stores · $0.50 unit margin · 2.1M employees
Daily ops
1.7M items/hour through cross-dock DCs; HQ Saturday meeting drives the week
Moat
Physical infra · distribution · vertical
AI risk
Middle-office, merchandising
▸ Structural floor high. Margin moves to logistics agents.
▸ Claims data is the moat. Prior-auth denials are the political risk.
firms
500 firms. You know all of them.
Top 20 by market cap · 500 in dataset
NVDA
NVIDIA
Info Tech · Semiconductors · $5.46T
Capability · Switching
GOOGL
Alphabet
Comm Services · Interactive Media · $4.81T
Data · Network
AAPL
Apple Inc.
Info Tech · Tech Hardware · $4.41T
Vertical · Taste
MSFT
Microsoft
Info Tech · Software · $3.13T
Distribution · Switching
AMZN
Amazon
Cons. Disc. · Broadline Retail · $2.84T
Infra · Vertical
AVGO
Broadcom
Info Tech · Semiconductors · $2.01T
Capability · Switching
TSLA
Tesla
Cons. Disc. · Automobiles · $1.59T
Vertical · Brand
META
Meta Platforms
Comm Services · Interactive Media · $1.56T
Data · Network
WMT
Walmart
Cons. Staples · Staples Retail · $1.05T
Physical · Scale
BRK.B
Berkshire Hathaway
Financials · Financial Services · $1.04T
Float · Trust
LLY
Eli Lilly
Health Care · Pharmaceuticals · $896B
Patents · Regulatory
JPM
JPMorgan Chase
Financials · Banks · $798B
Regulatory · Trust
AMD
AMD
Info Tech · Semiconductors · $692B
Capability · Switching
XOM
ExxonMobil
Energy · Oil & Gas · $655B
Reserves · Capital
V
Visa
Financials · Financial Services · $620B
Network · Regulatory
ORCL
Oracle
Info Tech · Software · $555B
Distribution · Switching
INTC
Intel
Info Tech · Semiconductors · $547B
Manufacturing · Capital
JNJ
Johnson & Johnson
Health Care · Pharmaceuticals · $546B
Patents · Regulatory
CSCO
Cisco Systems
Info Tech · Comms Equipment · $467B
Network · Switching
COST
Costco
Cons. Staples · Staples Retail · $465B
Membership · Trust
approach
I ask every firm the same four questions. Ask them 500 times and the answers start talking to each other.
01 // MONEY
How it makes money
Where revenue is. Where margin lives. Unit economics.
02 // DAY
What a day in ops looks like
Hourly to quarterly rhythm. Triggers, queues, breaks.
03 // MOAT
Why it compounds
License, distribution, vertical, taste, data, or infra. Named.
04 // RISK
Where AI hits
Which roles compress. Which moats erode.
browse
Six firms. Tap to load.
Costco
retail · membership
MoatDistribution + trust
MarginMembership fees
AI riskLow
Money
$65/yr to walk in. Retail margin is near zero. Membership fee is the profit. ~4,000 SKUs (Walmart has 140k) = brutal supplier leverage.
Day
Buyers cut SKUs daily. Receiving runs pallet-in, pallet-out (no shelving). 93% renewals roll the flywheel. Treasure-hunt curation is the in-store ritual.
Apple
consumer tech · vertical
MoatVertical + taste
MarginServices · ~70% GM
AI riskLow
Money
Six layers: silicon, OS, hardware, App Store, payments, identity. Hardware funds the lock-in. Services ride on top at ~70% GM.
Day
Cupertino runs ops; Foxconn runs the line. Weekly demand pulls cascade to Asia. Launch quarters bend the supply chain around one product window.
Visa
payments · network
MoatNetwork + regulatory
Margin~$0.13 / swipe
AI riskMedium
Money
Rails between 14k banks · 80M merchants. ~$15T/yr flows through. ~13¢ per swipe. ~50% drops to operating income.
Day
Mutual hostage: no one defects because everyone else is on it. Settlement runs 24/7. Fraud-graph updated by the second. Compliance file open in 200+ jurisdictions.
Berkshire
diversified · float
MoatTrust + float
MarginFloat invested
AI riskLow
Money
Insurance generates ~$170B float: premiums before claims. Invested at ~10%/yr. Negative cost to hold thanks to underwriting profit.
Day
Omaha doesn't run ops, subs do. GEICO writes policies; BNSF moves trains; Dairy Queen sells cones. Capital allocation is the only daily decision at HQ.
UnitedHealth
healthcare · payer · data
MoatRegulatory + data
MarginPremiums + PBM
AI riskHigh (political)
Money
Three stacked: UHC (~$300B premiums), Optum Rx (PBM), Optum Health (~90k physicians). Each sees data the others can't.
Day
Claims processed continuously. Prior-auth queues are the rate-limiter. Quarterly MLR refile sets the next-year premium. Politics is the daily weather.
Boeing
aerospace · regulatory
MoatRegulatory + capital
MarginDefense + services
AI riskLow
Money
Commercial razor-thin. Defense + services carry OI. Order books extend 8+ years. Duopoly with Airbus. FAA cert is the third moat.
Day
Rate-of-build governs everything. Each program gated by FAA cert milestones. 100k engineers trained on one program; tooling costs $10-15B per program to replicate.
sectors
11 GICS sectors. Ordered by total market cap.
Information Technology
$19.0Ttotal mcap
NVIDIA · Apple · Microsoft · Broadcom · AMD · Oracle · Intel · Cisco
Communication Services
$7.4Ttotal mcap
Alphabet · Meta · Netflix · T-Mobile · Verizon · Disney · AT&T · Comcast
Consumer Discretionary
$6.4Ttotal mcap
Amazon · Tesla · Home Depot · McDonald's · Booking · TJX · Starbucks · Lowe's
Financials
$6.3Ttotal mcap
Berkshire · JPMorgan · Visa · Mastercard · BofA · Wells Fargo · Morgan Stanley · Goldman
American Tower · Welltower · Prologis · Equinix · Public Storage · Simon Property · Realty Income · Digital Realty
by
VR
VITALII RATUSHNYI
Ex-quant. Currently AI research.
Trained to read companies as systems, seeking alpha and finding edge. Making money without understanding the underlying reality is gambling. Learn the game and build the judgement.
I hate subscriptions. Buy once, forget the paid part exists.
$89. Yours forever.
97% OFF · LAUNCH PRICE
Fortune 500 Business Models
500 Fortune 500 firms, on the 2026 list published June 3, 2026. Three lenses on each: how the business runs, what the numbers say, and whether AI breaks it.
Fresh: 2026 ranks + FY2025 revenue, one week after Fortune published
Turnover tracked: Amazon's new #1, the exits (Walgreens, Hess, Foot Locker, Nordstrom...) and the debutants (Galaxy Digital, Venture Global, Arista...)
Process: queues, handoffs, rhythm
Financials: revenue, margin, FCF, debt, R&D
AI durability: what holds, what compresses, what to watch
Peer benchmark vs same-sector cohort
Fund-grade data, cited at every line, same primary feeds institutional desks rely on
Source-of-truth artifacts (web · html · markdown · agent skill · cli) for both humans and agents. Verifiable, traceable.
faq
Why not just do this myself with Claude Code / OpenClaw / Hermes / Codex?+
What's important and what's not, that's the actual question. Even on the smartest tier (Opus 4.7 max thinking, GPT-5.5 xhigh) you burn ~$15K in API tokens to model 500 firms at quality, and still need ~80 hours to verify the output. Fortune 500 Business Models is what you'd build with that time and budget.
vs reading the 10-K?+
10-K = 200 pages of unfiltered disclosure, written for lawyers. Fortune 500 Business Models = how the firm actually makes money, in plain terms.
Who's it for?+
Analysts, researchers, quants. Founders. Investors. And agents. Anyone (or anything) building judgment about how businesses actually run.
What's “interactive”?+
For you: a dashboard to learn from. For your agent: a well-maintained context layer to plug into, ready for agentic RAG.
How do I use it?+
Pick a firm. 5 min on money, 5 on the day. Patterns surface across sectors.
Updates?+
Free, forever. And fast: Fortune published the 2026 list on June 3; this dataset was re-based to it within the week. Every rank and revenue is FY2025, the M&A and take-private exits are flagged, the 2026 debutants are in the queue.