Listen to the deep-dive discussion –  Why Visa is Safer Than Treasury Bonds (19:05 min)

 

I. The Architecture

Visa does not lend money. Visa does not hold deposits. Visa does not take credit risk. Visa does not own inventory. Visa does not build factories. Visa does one thing: it operates a network through which a buyer and a seller trust each other enough to complete a transaction. That trust is the product. Everything else — the plastic card, the chip, the tap, the digital wallet — is a vehicle. The vehicles change every decade. The trust does not.

Four billion cards in circulation. 130 million merchants. Over $17 trillion in annual payment volume. Every transaction that crosses the network reinforces it — one more merchant makes the card more useful for the consumer, one more consumer makes the card more necessary for the merchant. The moat is not built by Visa. The moat is built by the ocean of transactions that flows over it. Every wave that passes deposits another layer of coral.

This is what distinguishes a coral moat from a sand moat. A sand moat — regulatory, contractual, based on a piece of paper — can be dissolved by another piece of paper. A coral moat is built by usage and strengthened by usage. No regulator created Visa’s network. And no regulator can uncreate four billion cards, 130 million merchants, and the daily habits of billions of people with a single rule change.

*     *     *

II. The Wave That Has Barely Begun

Eighty-five percent of global transactions are still conducted in cash. Not in developed markets alone — globally. In India, in Southeast Asia, in Africa, in Latin America, in the Middle East, the vast majority of commerce still happens with physical currency.

Every one of those cash transactions is a future digital transaction. Every future digital transaction will need a trusted intermediary — a network that the buyer and the seller both accept. The addressable market for Visa is not the $17 trillion it currently processes. It is the entire global economy. And the portion that has not yet been digitalized is vastly larger than the portion that has.

Every business has two components: the toll and the growth. The toll is the mechanism that collects revenue today. The growth is the force that makes tomorrow’s revenue larger than today’s. In most businesses, growth costs money — every dollar spent building the future is a dollar the toll cannot return to shareholders. The toll and the growth compete for the same dollar. But Visa’s growth does not cost a dollar. Visa does not pay for the digitalization of global commerce. Visa does not build the smartphones that enable mobile payments. Visa does not install the point-of-sale terminals. Visa does not create the e-commerce platforms. Governments, banks, technology companies, and entrepreneurs build the infrastructure. Visa collects a fraction of a percent on every transaction that flows through it. The wave arrives uninvited. The toll collects automatically. The cost of the growth is zero. Because the growth is free, the toll and the growth do not compete — they multiply. The toll retains 100% of its cash while the wave expands the volume passing through it. This is why Visa can simultaneously return massive cash to shareholders through buybacks AND grow revenue at double digits. It does not have to choose. The growth dissolves into the toll.

This is not a projection. It is not a narrative. It is the structural reality of a world that is irreversibly migrating from physical to digital commerce. The Second Law of Thermodynamics applies — the complexity of the payment system increases; it does not reverse. Cash does not replace cards. Cards do not replace digital wallets. The direction is one-way. And every step forward adds volume to Visa’s network.

*     *     *

III. The Three Headwinds

No honest analysis of Visa ignores the headwinds. They are real. They deserve examination. And they deserve context.

The first headwind is regulatory pressure on interchange. The Credit Card Competition Act would require multi-network routing for credit cards. A proposed settlement would cap standard consumer credit card interchange at 1.25% for eight years. The direction is toward lower interchange fees for merchants. But interchange is not Visa’s revenue — interchange flows to the issuing banks. Visa collects network assessment fees on every transaction regardless of the interchange rate. When the Durbin Amendment capped debit interchange in 2011, debit transaction volume increased. The highway carried more traffic at a lower toll per car. Visa’s assessment — the fee on every transaction that crosses the network — grew because the volume grew. The regulation compressed the banks’ economics. Not Visa’s.

The second headwind is stablecoins. Transfer volume hit $33 trillion in 2025, growing 72% year over year. The promise is compelling: settle transactions on blockchain at fractions of a cent, bypassing the 2–3% card rails entirely. But the rail is not the product. The product is the trust layer above the rail. Visa offers fraud protection, chargebacks, dispute resolution, credit, rewards, compliance, and KYC. A consumer who taps Visa at a store has 60 days to contest a fraudulent transaction. A consumer who pays in USDC has nothing. Until stablecoins replicate the trust infrastructure that has been built over seven decades, they remain a settlement tool for institutions and traders — not a replacement for consumer payments at the point of sale.

And Visa is not defending against stablecoins. Visa is absorbing them. $4.5 billion in annualized stablecoin settlement already flows through VisaNet. Visa is positioning as the bridge between the blockchain rail and the real world — the on-ramp and off-ramp. If stablecoins succeed, Visa becomes the trust layer on top of the new rail. The coral does not resist the new current. It absorbs it and grows.

The third headwind is sovereign payment rails. UPI in India. PIX in Brazil. SEPA Instant in Europe. Government-built systems that bypass card networks for domestic transactions. These are real — Visa has lost domestic market share in India. But sovereign rails are domestic by definition. The consumer who travels from Paris to Tokyo, the business that sells across borders, the remittance that crosses continents — all require a global network. No sovereign rail provides one. And no sovereign rail offers credit. They are debit systems. The credit infrastructure — risk assessment, underwriting, distribution — remains the domain of the global card networks.

Each headwind attacks a different layer. Regulation attacks the price of the toll. Stablecoins attack the rail. Sovereign rails attack domestic volume. No headwind attacks all three layers simultaneously. And the coral reef does not depend on a single current. It is nourished by every current that passes over it.

*     *     *

IV. Seven Decades of Waves

Visa has survived every technological disruption that was supposed to kill it. The magnetic stripe was supposed to be replaced. It was — by the chip. The chip was supposed to be replaced. It was — by contactless. Contactless was supposed to be replaced. It is being — by mobile wallets. Each disruption was a new vehicle for the same trust. Each vehicle change increased the volume flowing through the network.

The pattern is structural, not coincidental. Visa’s moat is not attached to a specific technology. It is attached to the concept of trusted intermediation between a buyer and a seller. The technology underneath changes every decade. The need for trusted intermediation does not. Stablecoins are the latest in a seventy-year sequence of new vehicles — and Visa is already integrating them into the network, as it integrated every previous vehicle.

The businesses that die from technological disruption are those whose moat is the technology itself. When the technology changes, the moat disappears. Visa’s moat is not the plastic card. It is the trust embedded in the network. And trust does not become obsolete when the card becomes a phone becomes a token becomes a stablecoin.

*     *     *

V. The Gratuity Premium and the Negative Risk Premium

The market sees Visa clearly. The toll is transparent. The margins are public. The growth rate is published. Every analyst models the revenue, the take rate, the volume. Visa is perhaps the most visible business on Earth.

And yet Visa is structurally undervalued. Not because the market is irrational. Because the model is incomplete.

The standard valuation model discounts future cash flows at a rate that includes a risk premium — typically 3–5% above the risk-free rate. This risk premium compensates the investor for the possibility that the cash flows will not materialize. The business might fail. The growth might stall. The moat might erode. The risk premium assumes that the future is uncertain and that the uncertainty has a cost.

But Visa’s future cash flows are less uncertain than a Treasury bond’s. A Treasury bond depends on the solvency of the U.S. government. Visa depends on the structural reality that humans will continue to buy things and that the digitalization of commerce will continue. The probability that humans stop transacting is lower than the probability that a government defaults. The risk of holding Visa is, in a meaningful sense, negative.

If the risk premium should be negative — if the discount rate should be below the risk-free rate — then the present value of Visa’s cash flows is higher than any standard model computes. At a 5% discount rate, Visa’s FCF of $19 per share is worth approximately $380. At a 3% rate — below risk-free, reflecting the negative risk premium — the same cash flow is worth approximately $640. The current price is roughly $300. The gap between the price and the structural value is the Gratuity Premium — the portion of Visa’s worth that is permanently absent from the market price because the spreadsheet cannot model it.

The Gratuity Premium is not created by a market crash. It is not temporary. It exists at every price — at PE 35, at PE 28, at PE 22 — because the cause is permanent. The spreadsheet will never model a negative risk premium. The cell cannot accept it. The model breaks at exactly the point where the truth begins.

This is why Visa appears expensive and is actually cheap. The market overvalues what it can see and undervalues what it cannot see. The stablecoin threat is visible — $33 trillion in volume, headlines, analyst reports. Visa has fallen 19% from its peak. The visible risk is in the price. The invisible defense — the trust layer, the fraud infrastructure, the compliance network, the credit system, the 130 million merchant relationships, the four billion cards, the seven decades of institutional trust — is not in the price. It is never in the price. The threats are visible, loud, and priced. The moat is invisible, silent, and free. The PE of 28 contains the toll and a modest growth expectation. It does not contain the structural undervaluation created by a risk profile that is lower than the risk-free rate, nor the free growth that the wave delivers, nor the self-reinforcing moat that strengthens with every transaction. All of that is absent from the price. All of that is the Gratuity Premium.

*     *     *

VI. Where Visa Sits

Not all moats are created equal. A moat built by regulation can be dissolved by regulation — a single rule change can erase decades of dominance overnight. A moat built by technology must be rebuilt every cycle — each generation of product requires new investment, new decisions, new execution. A moat built by network usage grows stronger with every transaction, every user, every merchant — it is self-reinforcing and free.

Visa’s moat is the third kind. Every transaction strengthens the network. Every merchant added makes the card more useful. Every consumer added makes the merchant more dependent. The moat costs nothing to maintain because the users maintain it by using it. And the growth costs nothing because the wave — the digitalization of global commerce — is driven by forces external to the business. The moat is natural. The growth is natural. Both are free. This is the rarest architecture in investing — what the framework calls Double Naturality.

The businesses that appear similar to Visa but are structurally different are those whose moats depend on regulation rather than usage. Their tolls look identical — a fraction of a percent on transactions they do not create. But the source of the moat determines its permanence. A toll protected by a rule can be repriced or eliminated by a new rule. A toll protected by four billion cards and 130 million merchants can only be eliminated if four billion people simultaneously change their habits. The first is fragile. The second is antifragile.

*     *     *

VII. The Reef

The take rate will compress. Slowly. Over a decade. From 0.20% to perhaps 0.15%. The cross-border premium will narrow as stablecoins capture some institutional flows. Domestic share in some emerging markets will erode where sovereign rails succeed. These are real headwinds. They will moderate Visa’s growth from 15% to perhaps 10–12% annually over the next decade.

And 10–12% annual growth on a business with 50% operating margins, near-zero capital expenditure, a moat that strengthens with every transaction, a wave that has digitalized only 15% of global commerce, and a risk profile that is structurally lower than a government bond — compounded over twenty years — is one of the most powerful wealth creation engines available at any price.

The headwinds are real. The reef is more real. The headwinds are visible, temporary, and priced. The reef is invisible, permanent, and free.

The coral grows in silence. The ocean does the work. The toll collects. And the wave — 85% of the world still in cash — has barely begun.

*     *     *

Watch the video

averagingup.com