DISCOVER
Eraj Alisherov
Correspondent Accounts
Global Finance's Softest Target
September 2026
— I am writing this to answer a question compliance officers ask each other in hallways but rarely put in writing: how does a bank end up moving another institution's money for years without ever truly knowing whose money it is? I have spent a considerable stretch of my career examining the plumbing beneath international finance — not the glass towers or the trading floors, but the quiet, unglamorous machinery of the correspondent account, the arrangement that lets one bank act on behalf of another across a border it has no branch in. What I have found, case after case, is that the largest failures in this system rarely announce themselves. They arrive dressed as ordinary business, wearing the paperwork of a routine relationship, for a decade or more, before anyone in a position of authority finally asks the one question that actually matters. This is not an indictment of any single institution. It is an account of a structural blind spot that has already cost the financial system hundreds of billions of dollars, and shows no sign of closing on its own.
SERVICES
Risk Advisor, Financial Crime Analyst, Strategic Forecaster
Origins of the System
Correspondent banking is one of the oldest working arrangements in finance, older than the wire transfer, older than the telegraph that first made international wires possible. Its logic is almost embarrassingly simple: a bank that has no physical presence in a foreign country opens an account with a bank that does, and the two institutions agree to act for each other's customers as if they were extensions of the same branch network. The account the domestic bank holds abroad is called a nostro account, Latin for “ours”; the same account, viewed from the foreign correspondent's side of the ledger, is called a vostro account, “yours.”
Two names, one balance, one shared exposure. For most of banking history this was simply how the world moved money — through the SWIFT messaging network, banks that had never met, in countries that had never traded directly, could still settle a payment by routing it through a correspondent that knew both sides. What the system was never built to handle gracefully is depth.
A correspondent bank can itself be a respondent to another correspondent further up the chain, a structure the industry calls nested, or downstream, correspondent banking. Every additional layer adds a bank that is one step further removed from the customer at the very start of the chain — and one step less likely to ever ask who that customer really is.
Current Reality
Distance Costs Money
Two cases, a decade apart, illustrate exactly how expensive that distance can become. Between 2007 and roughly 2015, the Estonian branch of Danske Bank, Denmark's largest financial institution, processed suspicious transactions now estimated at over $230 billion — a sum flowing overwhelmingly from Russia, Azerbaijan, and Moldova through a portfolio of non-resident clients with no discernible reason to be banking in Tallinn at all. The branch's American correspondent banks, including JPMorgan, Bank of America, and Deutsche Bank, were not passive bystanders. By 2013, JPMorgan had grown uneasy enough about the pattern of activity to raise it directly with Danske's Estonian leadership, in a meeting alongside Bank of America and Deutsche Bank, and press for the non-resident portfolio to be reduced. Danske refused. It took a whistleblower, an external report running to 87 pages, the resignation of the bank's chief executive, and eventually a $2.06 billion settlement with U.S. and Danish authorities before the branch was finally shut down in 2019. The most unsettling detail in the entire affair is not the size of the sum. It is that the bank's own correspondents saw the warning signs years before the public did, said so out loud, and were told no.
The second case moves the same mechanism into a different hemisphere and a different substance. Between 2004 and 2007, Wachovia Bank — now part of Wells Fargo — allowed an estimated $378.4 billion to pass through its correspondent relationships with Mexican currency exchange houses, or casas de cambio, tied to the Sinaloa cartel. Drug proceeds moved through Wachovia's correspondent accounts to purchase aircraft; one of those planes was later seized carrying 20,000 kilograms of cocaine. The bank's total penalty for what prosecutors called the largest violation of the Bank Secrecy Act in U.S. history came to $160 million — a figure that represented less than two percent of Wells Fargo's profit that same year. No individual at the bank was indicted. The deferred prosecution agreement simply required better controls going forward, and Wachovia's stock rose the week the settlement was announced. Insiders in the compliance world have a blunter name for what connects these two stories than “failure of oversight.” They call it correspondent pruning — or its absence. A bank that notices a relationship is generating outsized profit relative to its size, or activity that doesn't match its stated business, is supposed to prune the relationship: reduce it, question it, or end it outright. What both Danske and Wachovia demonstrate is that pruning is a discipline most institutions only practice after the damage is already public.
The Future
Regulators responded to a decade of cases like these the only way regulators know how: by tightening enforcement, raising the cost of every cross-border relationship, and making the paperwork behind due diligence considerably heavier. Banks responded the only way large, risk-averse institutions know how: by leaving. The industry calls this de-risking, and it is now the dominant trend reshaping correspondent banking worldwide. Rather than absorb the cost of monitoring a client, a country, or an entire region more closely, a growing number of global banks simply exit the relationship altogether — no engagement, no remediation plan, just termination.
The trouble with de-risking is that it solves the bank's exposure without solving anything else. The underlying demand for cross-border payment does not disappear the moment a correspondent account closes; it is displaced into channels that are considerably harder to see — informal value transfer networks, nested arrangements several layers removed from any single bank's due diligence, corridors with only one remaining correspondent relationship propping up an entire country's connection to the global financial system.
Cut the wrong thread and, according to World Bank and industry research, remittance costs into the affected corridor climb, pushing still more of that flow into channels with even less visibility than the one that was just closed. The people who pay for de-risking are rarely the institutions doing the risking. They are the small exporter who can no longer receive payment, the family relying on a remittance corridor that just lost its last bank, and the regional lender whose international access now runs through a single point of failure.
What Question Should We Really Be Asking?
This is where the honest version of the industry conversation gets uncomfortable, because the question everyone in this field eventually has to sit with is not whether de-risking reduces a bank's own legal exposure — it clearly does — but whether it reduces financial crime at all, or simply relocates it somewhere with weaker lighting. A relationship that generates thin margins and heavy compliance cost is, from a purely commercial standpoint, an easy one to close regardless of the risk profile behind it; the decision to exit often looks identical whether the underlying concern is genuine criminal exposure or simply that the account was never especially profitable to begin with. Which raises the harder version of the same question: is the global banking system currently organized to detect financial crime, or merely to ensure that whichever institution is holding the account when the music stops isn't the one left standing? I have never found a fully satisfying answer. I suspect there isn't one — only a shifting balance between the cost of looking closely and the cost of getting caught not looking at all.
''A senior anti-money-laundering investigator once put the entire problem to me more bluntly than any regulatory report I have read since: “A correspondent account doesn't launder money. It just agrees, quietly, not to look too closely at who did.” I had no risk model or forecast to offer back. I simply wrote it down. Every bank that has ever opened a nostro account has made the same quiet agreement, whether it meant to or not — and someone, eventually, ends up paying for how closely that agreement was honored.'' — Eraj Alisherov, The Patriot
The Land of Ajam
''Compliance desks have an old maxim, older than the term “anti-money-laundering” itself: know your customer's customer, or don't bother knowing your customer at all. It's a warning about depth, about the layers hiding behind a name on a wire instruction. In Danske's case, and in Wachovia's, the maxim didn't stay a maxim. It became a court filing, a deferred prosecution agreement, and a multi-billion-dollar settlement — proof that the principle was always more than a saying.'' — Eraj Alisherov, The Patriot
The End
''And so the wires keep moving, the nostro and vostro ledgers keep balancing every evening across dozens of currencies, and for every correspondent relationship a bank quietly prunes in the name of caution, another opens somewhere considerably less examined. The corridors change. The underlying temptation — to look away when the paperwork is thick, the client is profitable, and the questions are inconvenient — does not.''
— Eraj Alisherov, The Patriot
The unease documented in cases like Danske and Wachovia is not an isolated pattern; it is a measurable trend across the entire correspondent banking system. According to Bank for International Settlements data, global correspondent banking relationships declined by roughly 25 percent between 2011 and 2024, as bank after bank chose exit over remediation. And in a 2015 World Bank survey of the world's largest international banks, 75 percent reported actively reducing their correspondent banking relationships in direct response to the mounting cost and risk of compliance.
These numbers confirm what compliance officers already sense at their own desks: closing an account is easy. Making the money it used to carry stop moving is not — it simply finds a door that nobody is watching yet.
GLOBAL CORRESPONDENT BANKING DECLINE (2011–2024)
LARGE BANKS REPORTING CBR WITHDRAWAL (2015 SURVEY)
