🇲🇽 Payments in Mexico and the Andes

Mexico: SPEI, the CLABE, and the plumbing nobody sees

SPEI (Sistema de Pagos Electrónicos Interbancarios) is Mexico’s instant interbank transfer system. It has been run directly by Banco de México since 2004, six years before India’s IMPS and 16 years before Pix. That makes it one of the oldest systems of its kind in the world, and by far the largest in Latin America. The central bank never gave it a consumer brand. Each bank offers the service under its own name, inside its own app. That probably explains why Mexicans use the generic term, “te hago una transferencia” (“I’ll send you a transfer”), while a Brazilian names the system itself: “I’ll send you a Pix.”

Beneath SPEI sits an interbank settlement layer that end users never see. SIAC (Sistema de Atención a Cuentahabientes de Banco de México), live since 1995, holds the banks’ current accounts at the central bank and provides intraday liquidity. SICAM (Sistema de Cámaras), live since 1998, settles on a net basis the obligations arising from the retail clearinghouses. A single private operator, Cecoban, runs those clearinghouses. Since 1996 it has operated the CCEN (Cámara de Compensación Electrónica Nacional) for checks, non-urgent bulk transfers (TEF, at D+1), and direct debits (domiciliación de recibos). TEF is still running. But it no longer makes economic sense next to SPEI, which settles immediately while TEF still settles at D+1, so there is no reason to make it the default in a new integration.

The CLABE (Clave Bancaria Estandarizada) is the identifier that routes a transfer to an account in Mexico. It has 18 digits, which encode the institution code, the city, the account number, and a modulo-10 check digit. Under a standard set by the Asociación de Bancos de México, it has been mandatory for every domestic electronic transfer since June 1, 2004. The check digit lets a server validate the key before anything is sent, which makes it a robust identifier. For users, though, the 18 digits have to be read out, copied, or retyped for every new payee, which makes entry slow and error-prone. CoDi and then DiMo were both designed to spare users that step.

A switch is the infrastructure that routes authorization requests between acquirer and issuer. Mexico has only two, and this switch duopoly is the second defining feature of its market. Every card or ATM transaction goes through one or the other. Prosa, founded in 1968, is owned by Banorte, Santander México, Scotiabank México, HSBC México, Invex, and Banjército. E-Global is owned by BBVA México and Citibanamex. Prosa also owns and runs Carnet, Mexico’s domestic card scheme and one of the few national schemes in Latin America to survive against Visa and Mastercard. Carnet owes its survival to its place inside the switch rather than to its own brand, and consumers barely recognize it. North America has nothing like this architecture, and Cofece, Mexico’s competition authority, has reviewed it on antitrust grounds. An issuer or acquirer entering Mexico has to connect to one of the two switches. There is no third option.

CoDi and DiMo: Latin America’s best-documented failure

CoDi (Cobro Digital) is the QR and NFC payment initiation service Banco de México launched in 2019. Its codes are interoperable across all institutions and run on SPEI, so settlement happens in real time. Payments are completely free for merchant and customer alike, and the central bank operates the service itself. A year later, Pix launched in Brazil with the same features and took the market. CoDi never took off. In 2024 its average volume was still counted in thousands of transactions a day, in a country of 130 million people.

CoDi and Pix are both free and interoperable, yet their adoption curves diverged. That shows being free and interoperable does not, on its own, drive adoption. Three conditions that came together in Brazil were missing in Mexico. The first is mandated distribution in banking apps. In Brazil, large institutions were required to join Pix, and they gave the Pix screen pride of place in their apps. The second is a merchant incentive, which Mexico lacked: a merchant who accepts CoDi gives up undeclared cash revenue with nothing visible in return. The third is public awareness. In November 2025, Banxico and the Asociación de Bancos de México publicly acknowledged that the public knows little about CoDi and DiMo.

DiMo (Dinero Móvil), launched in 2023 by Banco de México with the ABM, is a phone-number alias layer on top of SPEI. It adds no new settlement rail. It works like Pix keys or Zelle: money goes to a mobile number, and that number is resolved to the matching CLABE in the background. It is doing far better than CoDi: the registered base more than doubled in 2024 alone. In 2026, Banxico issued new rules requiring banks to simplify their mobile apps to speed up adoption. The regulator, in other words, sees the obstacle in the app journey, not in the settlement rail.

• Virtual CLABE per order: the standard pattern for A2A pay-ins in Mexico. The PSP assigns a unique CLABE per transaction or per customer, which makes reconciliation automatic. You still need to find out who holds the underlying deposit account: a bank or an IFPE.
• No direct debit mandate on SPEI: recurring payments go through Cecoban’s domiciliación de recibos, a separate regime, or through cards. This is the costliest gap to discover late in a subscription project.
• SPEI does not guarantee the right account: payee name checking is less mature than on the newest rails. For outbound bulk payments, you still need a check step before execution.
• DiMo does not replace the CLABE: it hides it. A pay-in system still has to store and reconcile a CLABE.

OXXO and cash at the counter: the rail merchants forget to plug in

Cash-at-the-counter payment is the method in which the customer buys online and pays for the order in cash at a neighborhood store. The flow has four steps. The customer places the order and receives a barcode reference, takes it to a convenience store checkout, pays in cash, and the order is released. In Mexico, this rail is called OXXO Pay (run by FEMSA since 2015) and Paynet, a multi-retailer network covering pharmacies, supermarkets, and rival convenience stores. Nothing of comparable scale exists in the US or Canada. A merchant selling in Mexico without this option cuts itself off from a significant share of its addressable market.

The same store network then becomes the base for an offer that competes with the banks. Spin by OXXO, launched in 2021, is a wallet built on that network, where the store acts as a cash-in and cash-out point. It gives FEMSA a reach into the unbanked that few Mexican institutions can match, and that reach comes from the number of stores rather than from the technology. Uruguay followed a similar path. Abitab was born in 1993 when a network of lottery agencies became a payment network. Today it has about 240 locations in Montevideo and another 240 spread across more than 100 towns in the rest of the country.

Colombia: Bre-B, or building on what already exists

Bre-B is Colombia’s interoperable instant payment system, set up by the Banco de la República and live since October 6, 2025. Colombia built it later than its neighbors, on an architecture no one had tried before. The central bank’s 2022 diagnosis put a number on the lag. The country recorded 36 account-to-account transactions per person per year, against 72 in Peru, 89 in Argentina, 90 in Chile, and 199 in Brazil. Cash was the preferred payment method for 78.6% of adults, transfers for only 14%. Instant payments existed, but almost entirely within each institution: in 2022, 96% of transactions through low-value deposit accounts were intra-institution. Transfiya, launched in 2020, carried only 6.8% of interbank transfers, with 37.8 million transactions for 15 million registered users.

Rather than build a single public rail on the Brazilian model, the Banco de la República chose to make the existing private systems interoperate. The project’s motto, written into its own technical document, is “construir sobre lo construido” (“build on what’s already built”). The central bank runs no retail payment system. It provides two centralized components that the private systems connect to. DICE (Directorio Centralizado de Llaves) stores, validates, and resolves llaves (keys), whether a mobile number, an ID number, an email address, an alphanumeric code, or a merchant code. It guarantees their uniqueness, so each llave points to a single deposit product. MOL (Mecanismo Operativo para la Liquidación) settles each order individually, in real time, and in central bank money, on accounts in the CUD (Cuentas de Depósito) system.

The first economic effects are already measurable. Bre-B P2P is free for end users, whereas traditional deferred transfers were still charged for. Some EASPBVIs (the system operators) have dropped their access fees and cut the prices they charge participants. The arrival of new SPBVIs has brought pricing tiered by merchant size. Half of all payments originate from low-value deposit accounts, mostly e-wallets such as Nequi or Daviplata, and the other half from savings accounts. These two account populations used to transact only within their own institution. Bre-B now moves funds from one to the other.

Peru: Yape, Plin, and interoperability imposed from above

Yape is the mobile payment wallet Banco de Crédito del Perú (Credicorp group) launched in 2017. It has grown into a super app combining payments, microcredit, top-ups, and a marketplace, with some 19 million users at the end of 2025. Plin, launched in 2020, is a P2P network linking the banking apps of BBVA, Interbank, Scotiabank, and BanBif. It has no app of its own. BCP’s competitors mounted a joint response without having to create a shared brand. Peru took the opposite path from Colombia. The private sector built the market, and the central bank stepped in only afterward.

In 2023, the Banco Central de Reserva del Perú mandated interoperability between these wallets, on the Transferencias Inmediatas infrastructure run by the Cámara de Compensación Electrónica (CCE). Peruvian banks have owned this clearinghouse since 2000. Few central banks anywhere have forced wallets that were already dominant to open up to each other without building a competing public rail. Since then, a Yape user can pay a Plin user with just a mobile number, and the clearinghouse settles the payment.

• To accept online payments in Peru: the main acquirers are Niubiz, Izipay, and Culqi; PagoEfectivo (CIP codes) remains essential for unbanked customers.
• Settlement always ends up in the BCRP’s RTGS system (LBTR): the CCE settles its balances there, and the whole retail chain, Yape, Plin, and Bim included, flows into it. It is the country’s systemic risk point.
• A BCRP × NPCI International agreement, signed in June 2024, provides for a public retail payment platform modeled on India’s UPI. Status: announced, not deployed. It would be the first UPI deployment in Latin America, worth watching as a sign that the Indian model is being exported as an alternative to the Pix model.
• Yape is not a rail: it is a product of a private banking group. Its commercial terms, exposure, and reversibility are not those of public infrastructure.

Argentina: Transferencias 3.0 and interoperability by mandate

The Banco Central de la República Argentina set up the Transferencias 3.0 framework starting in 2020, and it has been live as Pago con Transferencia (PCT) since 2021. It requires that any QR code displayed in Argentina be readable by any banking app or wallet. Argentina solved the same problem as Colombia and Peru by a third route: pure regulatory mandate. The resulting payment is a credit transfer, cleared by COELSA, the country’s single electronic clearinghouse since 1997. The payment therefore runs on the interbank transfer rail. No card network is involved.

Timing is what sets Argentina apart from the other two cases. Elsewhere, interoperability was built before private players became dominant, or negotiated with them. In Argentina it was imposed on a player that already dominated: Mercado Pago, whose proprietary QR codes were in a large share of neighborhood stores. The regulator did not build a competing public rail. It banned any in-store QR code that only its issuer’s own app could read.

Argentina’s ecosystem is the most fragmented in the region. Three wallets shape the market. Mercado Pago (MercadoLibre) serves as the de facto payment infrastructure for informal merchants. MODO, a joint venture of more than 30 banks launched in 2020, is a rare case of a bank-consortium wallet that actually caught on. Ualá (2017) is a neobank valued at $3.2 billion after a $195 million raise in March 2026. Then there is Cuenta DNI, the wallet of Banco de la Provincia de Buenos Aires. This case is unique to the region: the driver of adoption is a provincial public bank, not a fintech.

Bolivia, Ecuador, Venezuela, Chile: four models that don’t generalize

Bolivia, Ecuador, Venezuela, and Chile run on four mutually incompatible payment architectures, even though the four countries border one another. Bolivia made the interoperable QR code its dominant rail. Ecuador, a dollarized economy, has no public instant rail. Venezuela made phone-number transfers universal under hyperinflation. Chile has a mature card market that still lacks an instant retail rail. These four markets do not form a uniform “Andean market,” and an acceptance architecture designed for one of them cannot be carried over to any of the other three without rework.

QR Simple interoperable is the QR code payment system of the Banco Central de Bolivia, launched in 2019 with ASFI, the financial supervisor. By design, it links banks, cooperatives, mutual societies, and development institutions. In Bolivia, the dominant rail became the QR code, not the alias-based transfer. The figures are far from marginal: 891 million transactions worth $51,293 million in 2025, with volume growth of 131% year over year. The pace reaches 1,697 payments per minute, about 28 per second, with 88.4% of transactions under 500 bolivianos and 49.7% under 50 bolivianos. It is the region’s main counter-model to Pix, reaching mass use in micropayments through a QR architecture rather than an alias directory.

Ecuador is a dollarized economy: its currency is the US dollar. Its national payment system is run directly by the central bank and split into clearly defined components. SPI handles interbank payments, SCI collections, SSP the public sector, SPL online payments, the clearinghouses checks, and SOI international payments. There is no documented public instant retail rail. Acceptance shifted through private initiative alone. DeUna!, Banco Pichincha’s wallet, is open to customers of other institutions and claims more than 314,000 accepting merchants in 2026.

Pago Móvil Interbancario is the phone-number transfer service of the Banco Central de Venezuela, launched in 2017. This A2A rail works without a modern smartphone or advanced infrastructure, which is how Venezuela adopted it. It handled 376 million transactions in April 2025 alone, or 41% of the country’s banking transactions, against 47% for card terminals. The merchant variant, C2P (Pago Móvil Comercio a Persona), is fee-based and runs the other way: the merchant initiates the request and the customer confirms it with a code. Its fee cap is 1.5% of the amount, against 0.3% for P2P, with a minimum of 2 bolívares. Confusing the two variants therefore underestimates the cost of acceptance fivefold.

Chile is the most card-heavy market in the region and the one where A2A modernization has come latest. Transbank (1989) was long a near-monopoly acquirer; the Tribunal de Defensa de la Libre Competencia forced it to move to the four-party model. Its Webpay Plus gateway is still the benchmark in Chilean e-commerce, but Getnet Chile, backed by Santander, has built a competing position. Interchange has been capped since October 2023 by the Comité de Tasas de Intercambio at 0.50% for debit and 1.14% for credit (0.94% for prepaid); a further cut to 0.35% and 0.80%, planned for October 2024, was suspended. Financial inclusion in Chile rests on a single product, BancoEstado’s CuentaRUT (2006), opened with nothing more than a national ID number. It claims more than 16 million users in 2025, about 90% of the population over age 12.

Remesas: the region’s largest cross-border payment flow

Remesas (remittances) are the money that emigrant workers send home. For several countries in the region, they are the largest source of foreign currency after exports, and the main point of contact with a formal financial service for part of the population. Their weight puts them among the region’s main payment flows. The US → Mexico corridor is the world’s largest by value. It has just seen its first reversal in more than a decade.

These figures show two opposing trends. Mexico is declining, with fewer transfers and a barely higher average amount, while Colombia and Peru are setting records. Migration demographics, the exchange rate regime, and the sending country’s immigration policy each act separately on every corridor. A single regional assumption therefore cannot account for all corridors, and the Mexican corridor no longer foreshadows the others.

Operating in the region: what to connect, what breaks, and who to know

Three pay-in rails coexist in every country in the region: cards, instant transfers or QR codes, and cash at the counter. A payment stack designed for Europe or North America covers only the first. Deploying it produces poor conversion, which teams often blame on the market being too small. These markets pay differently, and some customers have neither a card nor an account. A missing rail shows up immediately in the conversion funnel.