In this article
Mexico, Brazil and Colombia have all moved toward faster, cheaper and more interoperable domestic payments. They did not get there the same way. Mexico built the settlement rail early. Brazil made a national payment scheme visible to almost everyone. Colombia developed large private digital-payment networks first, then built the interoperability layer that connected them.
Those differences reveal a broader point: payment modernization is not simply a technology story. It is an institutional-design problem. Who operates the infrastructure? Who owns the customer relationship? How is the recipient identified? Where does competition occur? Who carries compliance and fraud responsibilities? And what happens when money has to cross a national border?
The answers matter because the next stage of Latin American payment integration is unlikely to be another domestic app. It will be about connecting increasingly sophisticated national systems without losing the legal, compliance and settlement controls that make them trustworthy.
One region, different rails
Mexico: build the rail before the app
Banco de México launched the Sistema de Pagos Electrónicos Interbancarios, or SPEI, on 13 August 2004. SPEI was designed as a real-time settlement system able to process much higher volumes while reducing risks embedded in its predecessor, SPEUA. The two systems overlapped during a transition period; SPEUA stopped operating on 19 August 2005.
The timing matters. SPEI predates smartphones, the contemporary fintech boom and today’s global policy focus on instant retail payments. Its original significance was not as a consumer brand. It was common financial infrastructure.
Banco de México operates the settlement rail. Participating institutions own the customer interface.
That separation remains visible today. A customer opens a bank or financial-service app, instructs a transfer and the institution routes the payment through SPEI to another participating institution. The infrastructure is common; the customer relationship remains competitive.
The scale is enormous. Banco de México’s domestic electronic-transfer statistics recorded roughly 782 million SPEI operations in June 2026, totaling about MXN 21.9 trillion. SPEI has therefore become mass-market infrastructure even though most users interact with it indirectly through their bank or financial app.
Mexico also shows that settlement speed and payment addressing are different problems.
A conventional SPEI transfer can be directed using an 18-digit CLABE, as well as other supported identifiers. Later products simplified initiation without replacing the rail. CoDi introduced QR-based payment initiation. DiMo allows transfers using a mobile-phone number associated with the beneficiary’s account.
That sequence is important: Mexico could change how a payment is initiated and addressed without rebuilding the settlement layer underneath.
The economics are similarly layered. Banco de México charges SPEI participants for access and operation, while retail pricing is determined by financial institutions. In 2024, the weighted-average real commission charged to individuals by high-volume SPEI institutions fell to roughly MXN 2.6 per transfer. Several large participants charged no incremental fee for mobile-app transfers or included transfers within service packages.
The cost of the infrastructure, in other words, is not the same thing as the retail price of the payment.
Public infrastructure does not eliminate private competition. It changes the layer at which competition occurs.
That principle also appears at the border. Banco de México and the US Federal Reserve Banks created Directo a México, linking US-originated dollar payments with Mexican peso payouts through institutions connected to SPEI. The programme illustrates the additional layers that appear once a domestic payment crosses a border: foreign exchange, regulatory responsibility and settlement between separate national infrastructures. Banco de México has announced that Directo a México will conclude operations on 20 November 2026, but the architectural lesson remains.
SPEI can make the domestic leg highly efficient without solving the whole cross-border transaction.
Mexico therefore establishes the first distinction for the regional comparison:
The rail, the payment scheme and the user interface are related, but they are not the same thing.
Mexico built the rail early.
Brazil made the scheme itself visible.
Brazil: when the scheme becomes the product
Pix entered full operation on 16 November 2020 after a restricted production phase from 3 to 15 November. Banco Central do Brasil did more than provide a new settlement infrastructure. It created a common payment scheme with standardized rules, common addressing mechanisms, recognizable branding and a broadly consistent user experience delivered through competing banks, fintechs, cooperatives and payment institutions.
The architecture has three distinct pieces.
The Sistema de Pagamentos Instantâneos, or SPI, is the centralized real-time gross-settlement infrastructure operated by Banco Central do Brasil.
The Diretório de Identificadores de Contas Transacionais, or DICT, links Pix aliases — chaves — to underlying transaction accounts.
Pix sits above those components as the payment scheme: the rules and common experience through which participating institutions offer instant payments.
The central bank did not launch a retail bank competing for deposits. It created the scheme and critical infrastructure. Private institutions remained the customer-facing distributors.
That distinction helps explain why Pix became so visible.
A recipient can be identified through a Pix key linked to the underlying account: a CPF or CNPJ tax identifier, phone number, email address or random key. Payments can also be initiated through QR codes. DICT resolves the alias into the information required to route the payment.
The user identifies a person or business; the system resolves the account.
That is more than a convenience feature. It moves part of the complexity of account identification away from the user and into shared infrastructure.
Scale reinforced the standard. Banco Central do Brasil reports that more than 170 million individuals have used Pix. In 2025, the system processed almost 80 billion transactions worth more than R$35 trillion, while more than 920 million Pix keys had been registered. In the second half of 2025, Pix accounted for 54.7% of the number of payment transactions measured by the central bank.
Frequency does not mean Pix displaced every other payment instrument. Large-value transfers, cards and other methods continue to coexist with it. What Pix changed was the default experience for a very large share of routine transfers and payments.
Pricing reinforced that shift. As a general rule, individuals are not charged for sending or receiving Pix through electronic channels, subject to defined exceptions. Legal entities can be charged by their provider. “Free for the consumer” therefore does not mean costless for every actor in the ecosystem.
Standardization also changed competition. Banco Central do Brasil’s own management analysis found low concentration among Pix participants and an improving Herfindahl-Hirschman Index as new providers entered and smaller participants gained share. Pix did not cause every competitive change in Brazilian finance, but it supports a narrower proposition:
A common payment standard can reduce the importance of proprietary payment networks as a competitive advantage.
Providers can compete around the payment — on accounts, merchant services, credit, fraud controls, financial management, integrations and user experience — rather than by keeping users inside closed transfer networks.
Brazil’s lesson is therefore not that public infrastructure replaces competition. It is that standardization can move competition to a different layer.
Colombia approached the same problem from the opposite direction.
Colombia: innovation first, interoperability later
Before Bre-B, Colombia already had a large digital-finance ecosystem.
Nequi closed 2024 with more than 20 million users. DaviPlata reported 18.5 million. Millions of Colombians already used digital deposits, mobile apps and electronic transfers.
The problem was not the absence of digital adoption. It was the architecture connecting those services.
Banco de la República’s retrospective analysis is unusually clear. In 2022, digital wallets processed about 859 million transactions, but 96% of immediate wallet transfers occurred within the same institution. The central bank also notes that many users maintained multiple wallets precisely so they could transact instantly inside different closed environments.
Transfiya, launched before Bre-B, represented a meaningful step toward interoperability. But in 2022 it had 25 participating institutions and still did not provide universal reach across the market.
A collection of successful apps does not automatically become an interoperable payment environment.
Digitalization can occur without interoperability.
Bre-B was designed to solve that coordination problem. Banco de la República describes the approach as “building on what had already been built”: connecting existing industry infrastructure through common rules, technical standards and public-private governance rather than replacing the ecosystem with one new consumer app.
Bre-B entered operation on 6 October 2025.
It is not a standalone application. Customers continue using the app or website of the institution where they hold an account or deposit. Inside those channels they can register or manage Llaves and initiate immediate transfers.
The system connects multiple instant-payment systems and a wide range of banks, cooperatives, wallets and other financial entities under common interoperability rules.
Llaves solve an addressing problem similar to Pix keys. Users can link an eligible account or deposit to an identifier such as an identity number, phone number, email address or alphanumeric alias. The sender can initiate a payment using the Llave or a QR code instead of entering complete account information.
Again, the underlying account remains where the regulated money resides. The alias becomes the address through which the system locates that account.
Bre-B’s first year shows what coordination can do at scale. By October 2026, Banco de la República reported more than 36 million registered users, more than 116 million Llaves and over 1.7 billion transactions. Before Bre-B, roughly 25 entities participated in immediate-payment schemes; one year after launch, 245 entities were connected, including more than 172 cooperatives and six payment systems.
That does not prove every transaction became cheaper or every form of payment friction disappeared. It does show what interoperability changed structurally:
the number of potential connections expanded dramatically without requiring consumers to abandon the providers they already used.
Innovation did not first require a new app.
The next innovation was connection.
Three models, one regulatory problem
Mexico, Brazil and Colombia combined the same basic layers in different orders.
Settlement infrastructure determines how funds move and settle between institutions.
Addressing infrastructure determines how the system identifies where those funds should go.
Scheme rules and interoperability standards determine how different providers participate in a common payment experience.
Mexico built the settlement rail early and later simplified initiation and addressing.
Brazil combined central infrastructure, aliases and scheme rules so successfully that the scheme itself became part of everyday financial language.
Colombia developed major private digital products before introducing a common addressing and interoperability layer.
Different sequence. Similar problem.
And none of these systems eliminates the role of regulated payment providers. The customer may see only a phone number, QR code, Pix key or Llave, but underneath that interaction institutions still identify customers, maintain accounts or deposits, authenticate instructions, monitor suspicious activity, manage fraud, keep records and respond when transactions fail or are disputed.
The simpler the payment looks to the user, the more infrastructure must work correctly underneath it.
That becomes more difficult when the transaction crosses a border.
Domestic systems can define participation and responsibility inside one legal and supervisory environment. Cross-border payments connect institutions operating under different licensing, consumer-protection, AML/CFT, sanctions and data rules.
The Financial Stability Board has identified those regulatory and supervisory inconsistencies — particularly differences in the treatment of bank and non-bank payment service providers — as a source of cross-border friction.
Regional payment integration is therefore not only an engineering problem.
It is a regulatory-coordination problem.
Does Latin America need its own IBAN?
At first glance, the regional solution appears obvious.
Europe has IBAN.
Mexico has CLABE.
Brazil has Pix keys.
Colombia has Bre-B Llaves.
Why not create a standardized Latin American account identifier?
The question is useful because it exposes a common confusion: addressing is not the same thing as interoperability.
IBAN is an international standard for identifying an account. It does not itself create a payment rail, perform foreign exchange, determine settlement finality or harmonize regulatory obligations.
Brazil is the best illustration.
Banco Central do Brasil regulated IBAN as a valid standard for identifying Brazilian bank accounts in international transfers in 2013. Since July of that year, Brazilian financial institutions have been required to provide an IBAN on request and to accept incoming transfers that correctly identify a domestic account using it.
Yet Brazil still needed Pix, SPI and DICT to solve a different set of payment problems.
IBAN answers:
Which account is this?
Pix answers:
How can the payment be initiated, addressed and executed immediately across participating institutions?
Those functions interact. They are not interchangeable.
A regional account standard could still be useful. It might reduce formatting errors, simplify account identification and provide a common convention across borders.
But it would only be one layer.
The more important question is:
What common architecture would allow different national payment systems to discover, communicate with and settle against one another?
The regional stack
A genuinely interoperable regional environment would need at least five layers.
First, addressing.
A sender should not need to understand the banking architecture of the recipient’s country. A common account identifier could help, but so could a regional alias-resolution layer capable of connecting different domestic addressing systems.
Think of it, narrowly, as a directory that resolves different national payment addresses — closer to a DNS for payment addresses than to a new account number.
The analogy stops there. Address discovery cannot solve FX, settlement, compliance or liability.
Second, messaging.
The systems must exchange information in formats they interpret consistently. ISO 20022 matters because it provides a framework for structured financial messages. But adoption alone is not enough. Two systems can both use ISO 20022 while implementing fields or business rules differently.
The Committee on Payments and Market Infrastructures has therefore developed harmonized ISO 20022 data requirements for cross-border payments and encourages alignment by the end of 2027.
A common language produces interoperability only when participants agree sufficiently on how to use it.
Third, scheme rules.
Systems need compatible answers to procedural questions: when is a payment final, what happens when an account is invalid, how are errors and fraud handled, which institution responds to the customer, and how quickly must funds become available?
Technical connectivity without compatible operating rules can still produce a poor payment experience.
Fourth, compliance and data.
In June 2025, FATF revised Recommendation 16 on payment transparency, strengthening expectations around the information accompanying cross-border payments and fraud/error controls. FATF expects most or all new requirements to be in force by the end of 2030.
Any regional architecture being designed now should therefore anticipate the data standards it will have to support in the next decade rather than retrofit them later.
Fifth, foreign exchange and settlement.
This is the layer an alias cannot solve.
A Mexican sender may hold pesos. A Colombian recipient may want Colombian pesos. A Brazilian merchant may price in reais. Somewhere in the chain, institutions must provide FX conversion, liquidity, pricing and settlement across separate monetary systems.
Today, much cross-border banking still depends on correspondent-banking relationships. An IMF working paper on Latin American and Caribbean payment integration describes wholesale international payments as typically relying on a relatively concentrated network of correspondent banks. It also shows that several regional arrangements designed to conserve hard currency have ultimately settled net obligations in US dollars. The Central American SIP, for example, connects national real-time settlement systems but uses dollar-denominated settlement accounts.
That does not mean every regional payment is dollar-based.
It does mean Latin American integration cannot be designed as though dollar liquidity, correspondent banking and the US financial system do not exist.
Connect the systems, not rebuild them
Project Nexus provides a useful international model.
The BIS Innovation Hub designed Nexus around a simple proposition: countries should not have to replace their domestic instant-payment systems to make them interoperable.
Instead of each system building a bespoke bilateral link to every other system, a domestic instant-payment system connects once to a standardized multilateral architecture and reaches the other connected systems through that common layer.
The BIS says the model can enable cross-border payments to reach recipients within 60 seconds in most cases.
In 2025, the central banks of India, Indonesia, Malaysia, the Philippines, Singapore and Thailand incorporated Nexus Global Payments to take the framework toward live implementation.
Latin America does not need to copy Nexus mechanically. But the architecture illustrates the principle that matters:
regional integration does not require countries to abandon the domestic infrastructure they have already built.
Mexico can keep SPEI.
Brazil can keep Pix.
Colombia can keep Bre-B.
The harder task is building the layer between them.
An IBAN could help. It would not be enough.
The border is the next payment problem
Domestic instant payments show how much can be improved without changing the nature of money itself.
Cross the border and the problem becomes different.
Currencies must be exchanged. Institutions in different jurisdictions must communicate. Compliance information must travel with the transaction. Liquidity must exist in the right place. Settlement may depend on correspondent accounts and intermediaries the sender and recipient never see.
The next question is no longer simply:
How can domestic payments become instant?
It is:
How can value cross a monetary and regulatory border as easily as it moves inside one?
That is where stablecoins become relevant.
Not because blockchain makes SPEI, Pix or Bre-B obsolete.
But because stablecoins approach one difficult part of the cross-border stack through a different architecture.
Moving the token is not completing the payment
In correspondent banking, an international payment can require several messages, account relationships, FX transactions and balance-sheet updates before the recipient obtains usable funds.
A stablecoin can change the middle of that chain.
If a sender holds a dollar-denominated token and transfers it to a compatible wallet, ownership can move across a common blockchain ledger without establishing a new bilateral correspondent relationship for that individual transfer.
That may remove some reconciliation, messaging and operating-hour frictions.
But moving the token is not necessarily the same thing as completing the payment.
If the recipient ultimately wants Colombian pesos in a Bre-B-connected account, reais in a Brazilian account or Mexican pesos through a domestic institution, someone still has to provide conversion, liquidity and the connection back to local money.
The cross-border middle may change.
The endpoints do not disappear.
That is why on-ramps and off-ramps are part of the payment economics. The sender has to acquire the stablecoin. The recipient may need to convert it back into domestic money. Those steps can involve KYC, exchanges or other intermediaries, bank transfers, FX spreads, platform fees, local liquidity and withdrawal costs.
The BIS’s 2026 Annual Economic Report concludes that stablecoin performance in cross-border payments is uneven once fees, spreads and on/off-ramp costs are included; total cost can in some cases equal or exceed a bank transfer.
The relevant comparison is therefore not blockchain fee versus wire fee.
It is the all-in cost of the complete journey.
A real corridor: US–Mexico
The US–Mexico corridor shows how this can work without requiring the household receiving the money to become a crypto user.
The Inter-American Development Bank cites Bitso’s report that it processed more than US$6.5 billion in US–Mexico remittances in 2024, more than 10% of total corridor volume. Bitso describes infrastructure in which blockchain-based liquidity can operate behind the scenes while Mexican payouts connect to domestic rails such as SPEI.
The qualification is essential.
This does not mean Mexican households received US$6.5 billion in stablecoins.
It means a digital asset can function as an intermediate settlement or liquidity instrument while the end user still sends and receives familiar fiat money.
The consumer interface, the settlement mechanism and the monetary instrument do not have to be the same layer.
That model also exposes the remaining frictions. Exchange spreads, local cash-out infrastructure, banking access and regulatory obligations still sit at the edges.
A stablecoin does not make the domestic rail irrelevant.
It can change the connection to it.
Where digital assets actually fit
Stablecoins appear most compelling where the conventional cross-border stack itself creates friction: after banking hours, in treasury movements across jurisdictions, in corridors where correspondent relationships or prefunding are expensive, or where the recipient wants to retain dollar exposure.
A dollar-backed stablecoin can combine two functions: a transferable settlement asset and a representation of dollar value.
But scale should not be mistaken for proof that stablecoins are already a mainstream payment system.
An IMF working paper estimating 2024 international stablecoin flows found them especially large relative to GDP in Latin America and the Caribbean, at about 7.7% of regional GDP. The same research stresses that stablecoin activity spans capital flows, crypto-market settlement and other uses rather than only real-economy payments.
The BIS draws the distinction even more sharply. It estimates roughly US$28 trillion in adjusted stablecoin transaction volume in 2025, but only about US$390 billion in payment-related flows.
Stablecoin transaction volume is not the same thing as payment volume.
The relevant policy question is not how many trillions move on-chain.
It is which specific payment frictions stablecoins can economically remove.
Dollar access is both the feature and the policy problem
Stablecoins are overwhelmingly dollar-denominated. BIS estimates put the share at about 98%.
That helps explain their attraction in emerging markets.
For a user facing domestic-currency volatility, a dollar-linked token can function both as payment technology and as foreign-currency exposure.
A remittance recipient who immediately converts the token into local currency is using it primarily inside a payment chain.
A recipient who keeps the token is also making a portfolio and currency decision.
The same feature that can make stablecoins useful for cross-border settlement can therefore raise concerns about currency substitution, capital flows and monetary sovereignty.
Regulators must ask more than whether the transfer is fast.
What monetary claim does the user hold?
Who owes redemption?
Where are reserves located?
Which jurisdiction governs that claim?
What happens if redemption is suspended?
What legal rights does a holder in Latin America have against an issuer located elsewhere?
The Financial Stability Board recommends that global stablecoin arrangements provide users with a robust legal claim and timely redemption, including par redemption into fiat for single-currency arrangements.
A token intended to trade at one dollar is not automatically equivalent to a locally regulated bank deposit with the same prudential and legal protections.
Compliance does not disappear on-chain either.
A token can move directly between self-hosted wallets, but many economically useful payment journeys still touch regulated entities: issuers, exchanges, custodians, wallet providers, banks, payment companies or off-ramps.
FATF’s Travel Rule applies originator and beneficiary information requirements to transfers involving virtual-asset service providers. Pure peer-to-peer transfers between users acting through unhosted wallets are not subject to the same intermediary-based controls, but FATF expects providers interacting with those wallets to apply risk-based mitigation.
Technology can change where the intermediary sits.
It does not make financial-integrity obligations disappear.
Nor are jurisdictions regulating stablecoins uniformly. The FSB found significant gaps and inconsistencies in implementation of global crypto and stablecoin recommendations in 2025. The European Union’s MiCA regime, for example, has applied specific rules for asset-referenced and e-money tokens since June 2024, but a harmonized issuer regime in one market does not solve the differences stablecoins encounter when they circulate elsewhere.
Regional technology does not eliminate national regulation.
The competition may be between architectures
It would also be a mistake to frame the future as banks versus crypto.
The regulated monetary system is experimenting with many of the same technologies.
Project Nexus explores links between existing instant-payment systems.
Project Rialto demonstrated the technical feasibility of combining interlinked instant-payment systems with automated FX and tokenized central-bank money as the settlement asset.
Brazil’s Drex pilot tested tokenized assets, programmability and delivery-versus-payment inside a regulated financial architecture.
And Banco de México is one of the original seven central banks participating in Project Agorá. On 27 May 2026, the BIS reported that Agorá’s prototype had demonstrated atomic multi-currency settlement using tokenized commercial-bank deposits and tokenized central-bank reserves across currencies and jurisdictions.
Tokenization does not belong exclusively to cryptoasset issuers.
A future cross-border transaction could use a stablecoin, tokenized commercial-bank money, interlinked instant-payment systems, tokenized central-bank settlement assets or some combination of them.
The important competition may therefore not be crypto versus banks.
It may be:
Which architecture can deliver the best combination of speed, cost, liquidity, compliance, convertibility, legal certainty and settlement finality for a particular transaction?
The answer can differ by use case.
A domestic retail payment may remain naturally suited to Pix or Bre-B.
A wholesale treasury transfer may benefit from tokenized commercial-bank money.
A remittance corridor may use linked fast-payment systems.
Another corridor may find a regulated stablecoin useful as an intermediate settlement asset.
Payment systems have never used one instrument for every transaction.
There is little reason to expect digital finance to be different.
Building the rails between the rails
Latin America’s payment modernization is often told through national success stories.
Mexico has SPEI.
Brazil has Pix.
Colombia has Bre-B.
But together they reveal a larger institutional story.
Mexico shows the value of robust settlement infrastructure built early.
Brazil shows how a common scheme and shared addressing system can become part of everyday financial behaviour.
Colombia shows that private digital innovation can flourish before universal interoperability — and that coordination can itself become the next stage of innovation.
The logical frontier is now the border.
Latin America may benefit from common account standards or interoperable aliases. But identifiers alone cannot create a regional payment area.
The harder infrastructure lies underneath them:
messaging,
scheme compatibility,
compliance data,
foreign exchange,
liquidity,
and cross-border settlement.
That is why the region’s most consequential next innovation may not be another app, another national rail or even another form of money.
It may be infrastructure that allows different forms of money and different national payment systems to interact.
Stablecoins belong in that conversation.
So do interconnected fast-payment systems.
So do tokenized deposits and tokenized central-bank settlement assets.
None needs to be declared the inevitable winner.
The practical question is narrower:
Which architecture removes genuine payment friction without simply moving the risk somewhere less visible?
For Latin America, the first generation of instant-payment modernization was about building national rails.
The next may be about building between them.