There is a wire transfer sitting in a queue somewhere in the world right now that will take five business days and cost forty dollars to move two hundred dollars from a migrant worker in London to his mother in Lagos. The wire will pass through at least two correspondent banks, each charging a fee for the privilege of holding the funds for a portion of its journey. The recipient will collect the equivalent of roughly a hundred and seventy dollars after fees and exchange rate spread. This is not an exceptional case. It is the median experience of international remittance from a developed country to sub-Saharan Africa.
The correspondent banking system that enables this transaction is among the least visible and most consequential pieces of financial infrastructure in existence. It is also, by most measures, failing — not dramatically, but persistently and in ways that disproportionately harm the world's poorest economies. And it is failing at exactly the moment that stablecoins, dollar-denominated tokens running on public blockchains, are demonstrating that international value transfer can be done in minutes for fractions of a cent.
What Correspondent Banking Actually Is
When a bank in Lagos wants to send dollars to a bank in Singapore, it almost certainly does not have a direct relationship with that Singaporean institution. Instead, it uses a correspondent bank — typically a large global institution like Citibank, JPMorgan, or Deutsche Bank — that maintains nostro and vostro accounts on behalf of both parties. The Lagos bank holds a dollar account at Citibank. Citibank holds a relationship with a local Singaporean institution. The payment hops through that chain, each participant debiting and crediting its internal ledgers.
This system emerged in the nineteenth century and was formalised through the twentieth. It works, in a mechanical sense, but it is slow, expensive, and structurally fragile. Each hop in the correspondent chain introduces fees. Each hop adds settlement risk — the possibility that one party fails before the payment completes. The chain is also opaque: the originating bank frequently cannot see exactly where in the chain its funds are sitting, and tracking a delayed payment requires manual investigation across multiple institutions that have no obligation to respond quickly.
The system has been under stress for a decade because of a phenomenon called de-risking. Following a series of multi-billion-dollar anti-money-laundering fines levied by US regulators against global banks — HSBC, Standard Chartered, BNP Paribas, Wachovia — correspondent banks began systematically terminating relationships with counterparties they deemed too risky to maintain. The calculation was straightforward: the revenue from correspondent banking relationships with banks in Haiti, Somalia, or Afghanistan was small, and the compliance cost and regulatory exposure of maintaining those relationships was large. The rational response was to exit.
The World Bank has tracked the resulting decline for more than a decade. The number of active correspondent banking relationships globally has fallen by more than a quarter since 2011. The decline has been concentrated in small island nations, sub-Saharan African countries, and conflict-affected states. In some jurisdictions, only one or two correspondent banking relationships remain, creating single points of failure for the entire national payments system.
The De-Risking Trap
The perverse consequence of de-risking is that it pushes legitimate transactions into informal channels that are genuinely less transparent and more vulnerable to abuse. When a business in Port-au-Prince cannot send a wire transfer to pay a supplier in Miami, it does not stop doing business with that supplier. It finds an alternative — cash carried across borders, value transfer through commodity trades, informal money transmitters that operate outside the banking system entirely. These alternatives are harder to monitor, easier to abuse, and more expensive for the legitimate user.
The compliance logic that drove de-risking — reduce exposure to money laundering risk by cutting off high-risk jurisdictions — has therefore produced the opposite of its intended effect in many cases. The transactions that used to flow through a monitored, documented correspondent banking channel now flow through channels that generate no paper trail at all. Regulators have acknowledged this dynamic without resolving it: the incentive structure for large correspondent banks has not changed, because the upside of maintaining a small-country relationship is low and the downside of an enforcement action is enormous.
Stablecoins do not require a correspondent bank. A business in Lagos with a crypto wallet can send USDC to a counterparty in Singapore in under two minutes for a transaction fee measured in cents. No nostro account. No SWIFT message. No compliance department at Citibank making a risk-adjusted decision about whether to process the payment. The transfer settles on a public ledger that both parties can verify independently.
Stablecoins as Correspondent Banking Replacement
The practical use of stablecoins for cross-border payments has grown far faster than most observers anticipated. Tether's USDT and Circle's USDC together represent more than $150 billion in circulation as of mid-2026. A significant portion of that float is held not by crypto traders but by businesses and individuals in emerging markets who use stablecoins for trade settlement, remittance, and as a dollar hedge against local currency depreciation.
The use cases are specific. In Venezuela and Argentina, where local currency has depreciated dramatically, stablecoins serve as a dollar savings vehicle for ordinary people who cannot access US bank accounts. In Nigeria, where the naira has lost more than half its value against the dollar since 2020 and official dollar access is rationed, crypto exchanges and peer-to-peer stablecoin markets have become the primary mechanism by which small importers access the dollar liquidity they need to pay overseas suppliers. In the Philippines, one of the world's largest remittance-receiving countries, stablecoin-based remittance services are processing a meaningful and growing share of flows that previously went through Western Union and MoneyGram.
The competitive threat to correspondent banking is not yet existential, but it is directionally clear. The margin structure of international payments has depended on the absence of a cheaper alternative. That alternative now exists. The remaining question is how quickly users migrate to it, and how regulators respond.
The Regulatory Response
Stablecoin regulation in the United States crystallised in 2025 and early 2026. The GENIUS Act, passed in the Senate and signed into law, established a federal licensing framework for payment stablecoin issuers. The law requires stablecoin issuers to maintain dollar-for-dollar reserves in cash or short-term Treasuries, to undergo regular attestations, and to obtain a federal or state licence. Circle and Tether both received federal licences under the new framework.
The European Union's MiCA regulation, which came into full effect in late 2024, similarly requires e-money token issuers to hold full reserves and maintain licensed status. The framework has pushed several smaller stablecoin issuers out of the European market, but has given regulatory clarity to the major players. Circle has expanded its euro-denominated EURC token significantly under MiCA's licensing framework.
The global regulatory trend is toward licensing and reserve requirements rather than prohibition. This is a significant shift from the regulatory environment of 2022, when the collapse of TerraUSD — an algorithmic stablecoin with no real reserves — provoked calls for outright bans. The distinction between algorithmic stablecoins and fully reserved payment stablecoins has been firmly established in the regulatory discourse.
What This Means for the Banks
The correspondent banking business is not a trivial revenue line for the institutions that dominate it. JPMorgan processes more than six trillion dollars in payments daily through its correspondent banking network. SWIFT, the messaging standard that underpins most correspondent banking, carries more than fifty million financial messages per day. The infrastructure is enormous, the revenue is substantial, and the institutional commitment to maintaining it is deep.
The response from the large correspondent banks has been two-track. On one track, they have invested in upgrading the existing system: JPMorgan's Onyx network represents an attempt to create a faster, cheaper correspondent banking infrastructure using distributed ledger technology, without abandoning the bank-to-bank architecture that the existing system depends on. On the other track, they are building or acquiring capabilities in stablecoin custody and payments, acknowledging that the distribution channel for cross-border value transfer may shift even if the issuer of the underlying asset remains a regulated financial institution.
The most likely near-term outcome is not the elimination of correspondent banking but its compression. The fee structure that has persisted for decades is not sustainable in a world where a freely available alternative processes the same transaction for a fraction of a cent. The major banks understand this. Their competitive response will be to reduce the number of hops, improve transparency, and lower fees, while simultaneously investing in the stablecoin infrastructure that will eventually carry a larger share of the volume. The intermediaries who cannot make that transition — the smaller correspondent banks that lack the capital to invest in new infrastructure — will find their business eroding faster than they can replace it.