The United Nations marks today as the International Day of Family Remittances, recognising the contribution made by millions of migrant workers who send money to relatives in their countries of origin.
Few people have studied those flows as closely as Dilip Ratha. During more than three decades at the World Bank, he produced the first comprehensive global remittance dataset, created its Remittance Prices Worldwide database and led international work on migration and development finance.
Now founder and CEO of Ratha Global and a board member at Encryptus, Ratha considers why the correspondent banking system continues to impose such high costs on migrant families and whether regulated stablecoin infrastructure could provide an alternative.
When a construction worker in Dubai sends $200 home to his family in Lagos, the transaction passes through a chain of intermediary banks before it arrives. Each one takes a cut. By the time the money reaches his mother, she may receive $185, or less.
The $15 that vanished did not pay for innovation, speed or security. It paid for an infrastructure designed in the 1970s that the global banking industry has had no commercial incentive to fix.
The problem lies not in the technology, but in the business model that maximizes profits while taking advantage of onerous regulations.
Correspondent banking, the system through which most cross-border payments still move, was built for a world of telexes and paper ledgers. A sending bank does not have a direct relationship with a receiving bank in another country, so it routes the payment through one, two, sometimes three intermediary institutions, each of which charges a fee, applies an exchange rate margin and adds processing time. A payment that could settle in seconds instead takes days. A transaction that should cost pennies instead costs dollars.
The people absorbing these costs are overwhelmingly migrant workers, refugees and low-income households in developing countries. In 2025, $685 billion was remitted to low- and middle-income countries, up 5.8% on a year earlier, a figure that dwarfs foreign direct investment and official development assistance combined. Yet the average cost of sending money internationally remains above 6%, more than double the United Nations Sustainable Development Goal target of 3% by 2030. In some corridors serving sub-Saharan Africa, costs regularly exceed 8%.
At a 6% average fee, the global remittance system extracted over $42 billion from migrant workers and their families in a single year. That is not a rounding error. That is the GDP of a small nation, taken from people earning hourly wages and sending home what they can spare.
The architecture of extraction
The correspondent banking model persists not because it works well, but because it works well enough for the institutions that profit from it. Large global banks sit at the centre of a hub-and-spoke network that smaller banks in developing countries must access to move money across borders. This creates structural dependency. A regional bank in West Africa cannot settle a dollar-denominated payment without routing it through a correspondent in New York or London.
Because settlement is slow and unpredictable, payment service providers must pre-fund accounts in destination corridors, locking up capital that could otherwise be deployed productively. Because fees are layered and opaque, end users rarely know how much they are paying until the money arrives. Because compliance requirements are duplicated at every node in the chain, smaller institutions in developing markets face disproportionate regulatory burden and some lose correspondent banking relationships entirely – a phenomenon known as “de-risking” that has cut off entire regions from the global financial system.
When a global bank decides that maintaining a correspondent relationship with a bank in Somalia or the Pacific Islands is not worth the compliance cost, the impact is not to eliminate the demand for cross-border payments. The real impact is to push those transactions into informal channels – hawala networks, cash couriers, unregulated transfer services – where costs are higher, protections are weaker and financial intelligence disappears entirely. This does not reduce risk but shifts it to the people least equipped to bear it.
Who benefits from inertia?
The standard defence from incumbent institutions is that cross-border payments are inherently complex. Multiple jurisdictions, currencies, regulatory regimes and time zones create genuine operational challenges. This is true. But the banking industry has had decades to address these challenges and has instead optimised for margin preservation.
Consider the contrast with domestic payments. In India, the Unified Payments Interface processes billions of transactions monthly at near-zero cost. In Brazil, Pix settles payments in seconds, around the clock. In Kenya, M-Pesa transformed financial access for millions. These systems demonstrate that fast, cheap, reliable payments are technically achievable, albeit within national borders. The reason cross-border payments remain slow and expensive is not that the problem is unsolvable. The institutions controlling the infrastructure have insufficient incentive to solve it.
For too long, the remittance industry has been discussed as a technology challenge or a regulatory puzzle. The current system serves intermediaries. A system designed for the people who actually send and receive money would look fundamentally different.
The wrong escape route
Frustration with the incumbent system has driven some migrant workers toward unregulated alternatives – peer-to-peer stablecoin transfers that promise better rates by operating outside formal banking channels. When a worker discovers they can bypass correspondent banking fees entirely and deliver more money to their family, the immediate savings are real and impactful.
But the long-term costs are substantial. Unregulated stablecoin channels bypass know-your-customer and anti-money laundering checks, creating compliance risks for both sender and recipient. They underreport foreign exchange flows, weakening the visibility that central banks need to manage monetary policy. They drain deposits from the formal banking system, reducing capital available for lending in developing economies. And they concentrate settlement risk in unregulated intermediaries who may move funds across borders with no oversight, creating systemic vulnerabilities that recipient countries are least equipped to manage.
At scale, informal stablecoin channels do not fix the remittance problem. They replace one set of harms with another, trading high fees for regulatory arbitrage, weak governance and erosion of the monetary sovereignty that developing countries need to manage their own economic stability. The solution is not to abandon regulation. It is to build regulated infrastructure that works better than what currently exists.
What comes next
The tools to build something better already exist. Real-time settlement technology, digital currency infrastructure and modern compliance frameworks can reduce costs dramatically while increasing speed and transparency. What has been missing is the institutional will to deploy them at scale within a properly regulated framework, and the honest acknowledgment that the current system is not failing; for the banks that profit from it, it is working exactly as designed.
Properly regulated stablecoin-based payment networks, operating through licensed financial institutions with transparent compliance and custody arrangements, demonstrate that cross-border settlement can be fast, affordable and visible to regulators. The infrastructure exists. The question is whether policymakers will create the conditions for it to compete on equal terms with incumbent systems, or whether they will continue to protect a correspondent banking model that extracts billions annually from the world’s poorest households.
Every year that the correspondent banking model persists in its current form, billions of dollars are siphoned from the world’s most vulnerable households. Roughly one billion people globally send or receive remittances.
For them, these are lived realities, not abstract policy concerns. The data has been clear for more than two decades. The question is how long we continue to tolerate an infrastructure that treats the world’s poorest as a revenue stream and whether we are willing to allow properly regulated alternatives to replace it.
The post The Invisible Toll: How Correspondent Banking Quietly Strips Wealth from Those Who Can Least Afford It appeared first on The Fintech Times.
