Visa

A global toll on electronic payments with near-zero incremental cost per swipe.

FortressForeverFullTech fluentBerkshire overlap

Longhold

Earnings

Moat

Hold

Ten-year thesis

Visa is a network, not a bank. It does not take credit risk; it taxes the rails. Volume grows with commerce, e-commerce, and the slow death of cash — especially outside the US. A ten-year hold is a bet that account-to-account rails and real-time payment systems nibble take rates but do not replace brand, acceptance, and fraud tooling. This is one of the great capital-light compounders, tech included.

The moat

Two-sided network (cardholders and merchants), brand at the point of sale, and regulatory/compliance burden that freezes the oligopoly with Mastercard. Switching costs for issuers are high; for consumers they are a wallet habit.

Switching costs82

Pain, risk, or retraining required to leave.

Network effects96

The product gets better as more people use it.

Intangibles & IP84

Patents, data, licenses, process knowledge, regulation.

Scale advantage93

Fixed costs spread over a volume rivals cannot match.

Cost advantage80

Structural ability to be the low-cost producer.

Brand90

Pricing power from trust, habit, or identity.

Tech overlay

The network is software, tokenization, and trust at planetary scale. Fintechs (Stripe, Square, PayPal) rent the rails. That is the tell: disruptors became distributors.

What can break it

Government-pushed A2A rails compressing take rates
Merchant litigation on fees
A large-market digital currency that bypasses cards

Earnings path

USD billions, illustrative 2016–2025 series

Owner metrics

P/E

30.0×

P/FCF

28.0×

ROIC

38%

ROE

48%

FCF conversion

102%

Debt / equity

0.50

Gross margin

80%

Op. margin

66%

10y rev CAGR

10%

10y EPS CAGR

13%

Owner earnings

$19.0B

Mkt cap

$620B

Owner earnings yield 3.1%. Yield is judged against quality, not against a single hurdle — a fortress can be fairly priced at a lower yield than a fragile name.

PaymentsNetworkCapital-light