The Dollar Is Our Local Currency. It Is Still Somebody Else's Money.

Dollarization removes currency risk and replaces it with something harder to see.

Share
The Dollar Is Our Local Currency. It Is Still Somebody Else's Money.

Dollarization removes currency risk and replaces it with something harder to see.

Most payments analysis treats currency as a variable. In a dollarized economy it is a constant, and that changes what every other variable does.

I run payments technology in a market that uses the U.S. dollar as its everyday money. No conversion on a domestic card purchase. No FX line on a merchant statement. No devaluation risk sitting between authorization and settlement. From the outside this reads as the easy version of Latin American payments — all the volume, none of the currency drama.

It is not the easy version. It is a different version, and the differences are worth understanding even if you will never operate here, because a dollarized market is where several questions the rest of the industry is still arguing about have already been answered.

What dollarization actually removes

Start with the genuine advantages, because they are real and large.

There is no conversion on a domestic transaction. The cardholder is billed in the same unit the merchant is paid in, which is the same unit the issuer settles in. An entire category of complexity disappears: no conversion spread to disclose, no dynamic currency conversion to govern, no exchange-rate timestamp to argue about when a cardholder disputes the amount on their statement.

There is no FX position between authorization and settlement. In a market with its own currency, the days between those two events are a window in which the value of a transaction can move. Treasury hedges it, and the hedge costs money and attention. Here the window exists but the value does not move, so nobody staffs for it.

There is no devaluation scenario in the product roadmap. Pricing, interchange, fee schedules and merchant contracts are all denominated in a unit whose purchasing power is set elsewhere and does not jump. A processor in a market with a volatile local currency carries a constant repricing problem that simply is not on our agenda.

And inflation imported from the anchor economy is, for most of the last decades, lower and steadier than what the region has lived through. That is a real benefit to the people in the market, and it is also a benefit to anyone whose revenue is a percentage of transaction value.

All of which is why, from outside, a dollarized market reads as the simplified case. It has the regional volume and none of the currency drama. That reading is correct about what was removed and silent about what moved in.

What moves in to take its place

A currency is not only a unit of account. It is also an institution with a balance sheet behind it. Adopt someone else's currency and you keep the unit and give up the institution.

There is no lender of last resort. A central bank that issues its own currency can create liquidity in a crisis. It is the reason a solvent bank facing a run does not become an insolvent bank by Friday. A fully dollarized system has no such facility, because nobody in the country can print the currency it uses. Liquidity in the banking system is a function of actual reserves and actual credit lines, and when it is gone, it is gone.

For a processor that sounds like somebody else's problem until you look at what you do all day. Settlement is a credit event. Between the moment a transaction is authorized and the moment funds move, somebody is carrying exposure. In a system with a central bank, the worst case is cushioned by an institution whose job is to cushion it. Here the worst case is whatever your counterparty actually has. Your settlement risk is not backstopped; it is just risk.

Correspondent banking becomes a single point of failure. This is the part outsiders consistently underestimate. Our everyday money is issued by another country, which means any movement of that money beyond our borders ultimately clears through banks inside the issuing country's system. The local currency requires international correspondent relationships to function internationally. That is the structural oddity of dollarization, and it is not a detail.

Correspondent relationships are commercial decisions made by foreign institutions on their own risk appetite. Over the past fifteen years, global banks have withdrawn from entire categories of counterparty and jurisdiction, and the decision rarely has anything to do with the specific institution being cut. A bank can be well run, well capitalized and well supervised and still lose a correspondent because someone in another country reweighted a portfolio.

When that happens in a market with its own currency, you lose a channel. When it happens here, you lose access to the operating currency of the entire economy for anything that crosses a border. Which is why, in this market, the health of correspondent relationships is not a compliance topic filed under financial crime. It is infrastructure, in the same sense that your data center is infrastructure, and it belongs on the same risk register.

Monetary policy is an exogenous input. Interest rates here are set by decisions made in another country for that country's conditions. When those rates move, the cost of funding in this market moves with them, regardless of whether local conditions called for it. Credit card portfolios, merchant financing, the cost of carrying settlement float — all of it reprices on a signal nobody here sent and nobody here can influence.

None of this is an argument against dollarization, which has delivered decades of price stability to a region that mostly has not had it. It is an argument about where the risk went. It did not disappear. It moved from the FX desk, where it was visible, priced and hedged daily, to the correspondent and liquidity desks, where it is lumpy, sudden and much harder to hedge.

The cross-border transaction that never crosses a currency

Here is where the gap between how a market feels and how the rails classify it becomes concrete.

Card networks decide whether a transaction is cross-border by comparing the country of the issuer with the country of the acquirer. Currency is a separate dimension. So when a cardholder here buys from a U.S. merchant, the transaction is unambiguously cross-border to the network and completely unremarkable to the human being making it. No conversion appears. No foreign-transaction notice. The statement line looks exactly like a purchase made downtown.

The networks price this case specifically. Both Visa and Mastercard distinguish cross-border transactions that settle in the same currency from those that require conversion, and charge less for the first — less, but not nothing. The rates differ by region and program and are revised regularly, so the number matters less than the structure: you pay a cross-border assessment on transactions your customers experience as domestic.

Three consequences follow, and they compound.

Your cost base is heavier than your volume suggests. A meaningful share of what looks like ordinary consumer spending carries cross-border economics. In a market with its own currency, the friction of conversion is a natural brake on how much of that happens. Here there is no brake.

Your dispute exposure is written in someone else's jurisdiction. A cross-border transaction brings the acquirer's market rules into the picture. Your cardholder had an entirely domestic-feeling experience, and the dispute that follows is governed by chargeback rights, evidence standards and timelines set for a market they have never been to. Explaining that to a retail customer is one of the least rewarding conversations in this business.

Your fraud profile skews international. Card-not-present fraud follows the path of least friction, and we have removed a layer of friction that other markets still have. A dollarized market's e-commerce mix is tilted toward foreign merchants, which means your fraud models are being asked to judge transactions whose normal is set by another country's retail landscape.

None of this is exotic. It is simply that the classification the rails use and the experience the customer has point in opposite directions, permanently, and every system you build has to be honest about which of the two it is modeling.

Your customers are benchmarking you against another country

Remove the currency barrier and you remove more than a conversion fee. You remove the thing that quietly kept a local market local.

A cardholder here subscribes to foreign streaming services, foreign software, foreign storage, at foreign prices, without ever performing a mental conversion. They shop international marketplaces the way someone in Ohio does. Their expectations for how a payment should behave — how fast a refund appears, whether a card can be added to a phone wallet in thirty seconds, whether a subscription can be cancelled without a phone call — are set by the issuers and merchants of the anchor economy, not by their neighbors.

That produces a specific competitive position. You are not competing with U.S. issuers for their customers. You are being compared to U.S. issuers by yours. Feature demand arrives early, because the ecosystem that generates it reaches your market with no friction at all, and it arrives without the revenue base that funds it in the market where it originated.

Which brings up the part that rarely makes it into the optimistic version of this story.

There is no FX margin. Not on the volume that matters. For issuers in most of the world, foreign-transaction revenue is a real line: a spread on conversion, applied to every international purchase, that funds a meaningful share of the card economics. Here it survives only on the thin slice of spend that settles in a third currency — a purchase in euros, a hotel booked in sterling — while the dominant share of international volume, the part denominated in dollars, generates none of it. Your customers get the benefit, genuinely, every day, and your P&L absorbs the gap.

So the unit economics have to come from somewhere else: interchange, product, credit, fees that are visible rather than embedded. That is a harder commercial position than it looks from outside, and it is the direct cost of the simplicity everyone envies. You cannot have an economy where cross-border purchasing is frictionless and also have the revenue that friction used to generate.

The stablecoin question only a dollarized market can answer

The standard Latin American stablecoin thesis goes like this: people and businesses want dollars, local currencies are volatile, capital controls make dollars hard to hold, and a dollar-denominated token solves that. It is a good thesis. It explains most of the adoption the region has seen.

Now run it here, where everyone already has dollars. In a fully dollarized economy the entire access premise evaporates. There is no local currency to escape, no conversion to avoid, no devaluation to hedge against. If the regional thesis were the whole story, a dollarized market should be the one place in Latin America where stablecoins have nothing to offer.

That makes this a natural experiment, and the industry should pay attention to the result, because it isolates a variable nobody else can isolate: how much of the value of a stablecoin is the dollar, and how much is the rails.

What is left when you remove the dollar from the equation turns out to be specific.

Settlement that does not keep banking hours. Correspondent chains settle on business days, in windows, in the time zone of the intermediary. Value that moves continuously is worth something to a treasury operation regardless of what currency it is denominated in. This is not a dollar benefit; it is a clearing benefit, and it was always the more interesting one.

Programmability. Conditional release, automated reconciliation, settlement that carries its own instructions. Again: nothing to do with which currency.

And the one that matters most here — a second path that does not run through correspondent banking. Go back to the single point of failure. If the structural fragility of a dollarized market is its dependence on foreign correspondent relationships, then a settlement rail that reduces the number of intermediaries in that chain is not a cost optimization. It is risk diversification on the exposure that actually threatens you. A U.S. bank evaluating stablecoin settlement is weighing efficiency. An operator in a market like this is weighing concentration risk on the thing their whole economy runs on.

Two honest limits, because this is where the enthusiasm usually outruns the engineering.

It is not an escape from the banking system. Converting between tokens and spendable money happens at on- and off-ramps, which are operated by regulated institutions with their own correspondent relationships, their own compliance posture and their own ability to decide they would rather not serve you. You have shortened the chain, not left it.

And it does not create a lender of last resort. A stablecoin issuer's reserves back the token. They do not back your liquidity in a crisis. The deepest problem dollarization creates is the one stablecoins do not touch.

What is genuinely changing is who is building these rails. Through 2026 the card networks moved stablecoin settlement from pilot into commercial product — Visa took USDC pre-funding on Visa Direct into production in October, and Mastercard launched a platform with its BVNK unit the same week combining fiat and stablecoin accounts, FX, card issuing and payments behind one API. That changes the question for an institution in this market from whether to build something unfamiliar to whether to switch on something a counterparty you already have is now selling.

Meanwhile the regulatory direction is not uniform, and anyone planning around this should know it. Brazil's central bank, in a rule effective 1 October 2026, barred regulated electronic FX providers — payment institutions, e-money issuers and acquirers — from settling the offshore leg of cross-border payments in stablecoins, pushing that activity into the supervised VASP and bank channel. A capability can be available from your network and closed by your regulator in the same quarter, and in this region it has already happened once.

Why this matters if you will never operate here

A dollarized market is small. The reason to read about one is that it is running, today, several conditions the rest of the industry is moving toward.

Frictionless dollar access is becoming everyone's condition. The stablecoin thesis, if it works, ends the scarcity that made dollar access a product. When holding and spending dollars is easy everywhere, every market inherits some version of what we already live with: customers who buy internationally without thinking about it, benchmarking against foreign providers, and an issuer economics model with the conversion spread removed. We have been operating without that revenue line since long before anyone had to. The adaptations are not theoretical here.

Correspondent concentration is a small-market problem, not a dollarization problem. Any institution whose access to a major currency depends on a handful of foreign relationships has the same exposure; dollarization just makes it acute enough that you cannot ignore it. If you run payments in a smaller market anywhere, the question of what happens when a correspondent reweights its portfolio is yours too, and it is worth asking before the letter arrives rather than after.

And the structural pattern is the one the whole industry is about to meet. Everything in the third section of this piece comes down to a single mismatch: the rails classify a transaction one way, the human experiences it another way, and every system in between has to pick which reality it models. We live that mismatch on geography. The industry is about to live it on agency — a transaction the network must classify as initiated by somebody, performed by software, for a person who did not press a button.

The lesson from our side is unglamorous and transferable: when classification and experience diverge permanently, you do not resolve it. You pick, explicitly, which one each system serves, you write that choice down, and you accept that the gap will show up in disputes, in fraud models and in customer conversations for as long as the mismatch exists. Markets that decide this deliberately do better than markets that let each team decide separately.

My judgment

Dollarization does not remove risk. It relocates it, from a desk that watches it every day to a desk that discovers it all at once.

FX risk has a virtue that gets no credit: it is continuous and visible. It shows up in a daily position, it is priced by a market, and it is somebody's job. Correspondent risk and liquidity risk in a system without a lender of last resort behave in the opposite way. They sit at zero for years and then they are not at zero, and there is no gradual version in between.

That is the trade. A dollarized market buys decades of price stability and an unusually clean transaction experience, and pays for it with a concentrated dependency on institutions in another country that owe it nothing. For most of the time, the trade is excellent. The failure mode is rare and severe, which is exactly the shape of risk that institutions are worst at funding against.

So the operational conclusion is narrow and I would defend it anywhere. In a dollarized market — I write from Panamá, and the same structure holds in Ecuador and El Salvador — correspondent relationships belong on the technology and operations risk register, not only in the compliance file. They should be mapped the way you map a single-vendor dependency in your core: who the counterparties are, what the concentration looks like, what the alternative paths are, how long it takes to stand one up, and what the business does in the weeks before it is standing. If your answer to the last question is that it has never come up, that is not evidence of resilience.

And for everyone reading from a market with its own currency: the thing you are hedging daily is not the only thing that can take you down. It is just the one you can see.

Three questions

  1. If your institution lost its primary correspondent relationship this quarter, how many business days would pass before cross-border settlement resumed — and who owns that number?
  2. What share of your card volume is classified cross-border by the networks but experienced as domestic by the cardholder? If nobody reports that split, your cost and fraud models are probably treating it as one thing.
  3. For issuers outside dollarized markets: what percentage of your card P&L is foreign-transaction margin, and what does the business look like if frictionless dollar rails take it to zero?

Reply with even one line. The sharpest answers shape the next issue.

Sources

  • Tidal Commerce — cross-border fee structures, including the distinction between transactions settled in USD and those requiring conversion on Visa and Mastercard. https://tidalcommerce.com/learn/cross-border-fee
  • Digital Transactions, 15 Oct 2026 — Visa Direct launches commercial stablecoin service; Mastercard and BVNK launch Global Money Platform.
  • Banco Central do Brasil — Resolution 561, effective 1 October 2026, restricting regulated electronic FX providers from settling the offshore leg of cross-border payments in stablecoins.