What structural forces are reshaping global payments, and who wins — Visa/Mastercard, fintechs, or central banks
Who Controls How Money Moves — and Why That's Changing
Based on analysis of a 120-node, 497-edge knowledge graph mapping the structural forces reshaping global payment systems.
First, what is a “payment network” anyway?
When you buy something with a credit card, the money doesn’t just jump from your bank to the store’s bank. It travels through a system — a set of rules, pipes, and middlemen — and everyone in that system takes a small cut. Visa and Mastercard don’t actually hold your money. They run the roads that money drives on, and they charge a toll.
This analysis maps out who controls those roads today, who is trying to build new ones, and what happens to everyone involved.
The thing at the center of everything
The most connected concept in the entire map is something called the Four-Party Network Model — which is just a formal name for how Visa and Mastercard work. The four parties are: you, your bank, the store, and the store’s bank. Visa and Mastercard sit in the middle setting the rules.
This model is simultaneously the most defended and the most attacked structure in the whole map. More than a dozen forces are actively trying to undermine it. And more than a dozen other forces are actively reinforcing it. Nothing else in the map has this many connections pulling in both directions at once.
Think of it like a toll bridge that everyone uses. Half the towns nearby are trying to build their own bridges to avoid the toll. The other half keep using the toll bridge because it has ambulances on standby — and the new bridges don’t.
How the toll bridge defends itself
The clearest loop in the map is what happens with rewards. The toll bridge uses its toll revenue to give travelers free coffee and airline miles. Travelers prefer the toll bridge because of the free coffee. More travelers means more toll revenue. More toll revenue means more free coffee. The loop closes on itself and gets stronger every turn.
This is the single strongest positive feedback loop in the entire map. The individual edge connecting rewards back to the card network has the highest weight of any connection in the data — a 10 out of 10.
But here’s the thing that makes it structurally fragile: the rewards only work because merchants pay for them through the tolls. Merchants have been trying to break this for years. Every time a government caps the toll (as Europe did), the rewards shrink. Every time a merchant is big enough to threaten to leave the road entirely, Visa and Mastercard have to negotiate.
The one event that changed everything else
In 2022, Western governments cut Russia off from the international banking system (specifically something called SWIFT, which is the messaging system banks use to talk to each other across borders). This was meant as punishment. It worked — but it also sent a signal to every country in the world: your access to the global payment system can be turned off by someone else.
That single event branches outward in the map to almost every other major trend. China accelerated its own payment system. Brazil, India, and others sped up their national payment infrastructure. The European Union started treating payment infrastructure as a strategic priority rather than a commercial convenience. And stablecoins — digital currencies that run on software rather than banks — started looking attractive to anyone who wanted to move money without asking permission.
One node in the map captures a dark irony: the policy that created the stablecoin boom by weaponizing the banking system also created the tool that future sanctions targets might use to evade the next round of sanctions. The map labels this connection an “ironic echo” — the only edge in the entire dataset with that label.
Brazil figured something out that the US hasn’t
Brazil built a payment system called PIX, and it now processes more transactions per year than credit cards in Brazil. India built UPI. Both are national instant payment systems where money moves directly between bank accounts in seconds, for free.
The United States built something similar called FedNow. The difference: PIX is mandatory for banks above a certain size. FedNow is optional.
The map encodes this as the strongest inverse correlation in the entire dataset — a 9.5 out of 10 negative relationship. The US system’s voluntary nature is structurally associated with its limited adoption. The gap this leaves isn’t just inconvenient. It actively creates space for alternatives to fill in — including stablecoins and private payment apps that wouldn’t have as much room in Brazil or India.
The paradox of volume without profit
Here’s something the map highlights that isn’t obvious: having lots of transactions on your network doesn’t automatically mean you make a lot of money.
Brazil’s PIX processes enormous transaction volumes. The banks make almost nothing per transaction. Visa and Mastercard process smaller volumes in many markets but make substantial profits per transaction. The map has a specific node for this — the “Public Rail Volume vs. Private Network Profit Paradox” — and it’s connected to a lot of things.
This matters because it explains why governments passing laws to force more competition in payments doesn’t automatically hurt Visa and Mastercard as much as you’d expect. If you force the toll lower, the toll bridge might lose some revenue, but the alternative roads are still free — and free roads don’t pay for the ambulances. Merchants, consumers, and governments are all navigating this tension.
The fraud problem that keeps the old system alive
Real-time payment systems have a problem: when money moves instantly, it’s very hard to get it back if something goes wrong. If you get tricked into sending money to a scammer, the money is gone.
Credit cards have chargebacks. If you’re defrauded, you call your bank and they reverse the charge. The card networks run this entire dispute infrastructure.
The map shows that fraud on faster payment systems is not just a problem for consumers — it is a structural advantage for the card networks. Every news story about someone losing money to a bank transfer scam reinforces why people keep their Visa card. The map specifically encodes fraud characteristics as a mechanism that “strengthens the moat” around card network protection guarantees.
This also shows up as a constraint on European payment alternatives like Wero. The technical infrastructure exists. The map shows it is being held back partly by consumer expectation of chargeback rights — a legal and cultural infrastructure that took decades to build.
The AI twist that surprises people
Many analysts assume that AI will route around card networks — that AI agents making purchases autonomously will find cheaper paths to move money. The map encodes the opposite relationship.
The way AI agents identify themselves and authorize payments relies on something called tokenization — a system where your card number is replaced by a secure digital token. Visa and Mastercard built and control this tokenization infrastructure. The map shows a dependency edge from AI agent payment infrastructure back to Visa and Mastercard’s tokenization layer, weighted at 9.6 out of 10 — one of the strongest dependency relationships in the dataset.
The structural implication is that if AI commerce scales through tokenized credentials, the card networks may end up functioning as the identity and authorization layer for autonomous transactions. Not bypassed — embedded deeper.
The stablecoin acceleration
Stablecoins are digital currencies pegged to a real currency (usually dollars) that run on software networks instead of banks. Moving money via stablecoin can be faster and cheaper than sending it through international banking, especially for cross-border payments where traditional banking charges high fees and takes days.
The map shows high-cost international banking infrastructure as the single strongest enabler of stablecoin adoption — weighted at 10, the highest enabling edge in the data. As the US has moved toward regulating stablecoins (through legislation called the GENIUS Act), this has further accelerated the structural shift.
But it also creates a loop that constrains government-issued digital currencies (CBDCs). If governments try to issue their own digital money, they run into a structural problem: if it’s easy to move money in and out instantly, people might pull their money out of banks during a crisis — a digital bank run. This design constraint makes government digital currencies harder to build safely. The map encodes the failure mode of government digital money as an indirect advantage for private stablecoins.
Who benefits from the roads getting more complicated
When every country or region has its own payment system, moving money between them gets complicated. Someone has to figure out which roads connect, translate between formats, and route transactions efficiently. That aggregation layer — companies like Stripe and Adyen — benefits structurally from every new national payment rail that gets built.
The map shows this explicitly: geopolitical fragmentation of payment infrastructure, which appears as a destabilizing force everywhere else in the map, is a direct benefit to payment orchestration platforms. More roads means more need for navigation systems.
The tensions the map doesn’t resolve
Several major questions appear in the map as open, meaning the data encodes the competing forces but not the outcome:
Visa and Mastercard are actively building systems to operate on other payment rails, not just their own. The map shows this strategy both cannibalizing their existing business and potentially creating a new moat through tokenization. Which effect dominates is not resolved.
Dollar stablecoins and bank-issued digital money are competing for the same role. Regulatory sequencing — which framework gets finalized first — appears to be the deciding factor, and the map encodes this as open.
Trade policy accelerates both dollar alternatives and dollar infrastructure simultaneously. The net direction is not encoded.
Bottom line
The map shows a payment system under simultaneous pressure from multiple directions, but with more structural resilience at its center than the disruption narrative usually acknowledges.
The four-party card network model is the most attacked node in the map and also the most defended. Its primary defense mechanisms — rewards lock-in, fraud protection, and tokenization infrastructure for AI commerce — are structurally different from its primary attack mechanisms, which are government rails, regulatory intervention, and stablecoin bypass. These don’t cancel each other out simply.
The single clearest structural lesson in the data is the mandatory-versus-voluntary distinction in government payment infrastructure. Brazil mandated PIX. India mandated UPI. The US made FedNow optional. The outcomes differ accordingly. The graph treats this design choice — not technology — as the primary explanatory variable for why public payment rails succeed or fail.
And the one structural beneficiary that appears across nearly every scenario — incumbent victory, incumbent erosion, government intervention, geopolitical fragmentation, stablecoin rise — is the aggregation layer above the rails. When roads multiply, navigation becomes valuable.