In an essay published by Foresight, Bitget Wallet researcher Lacie Zhang frames a long-running question at the center of offshore dollar finance: who actually owes you that one U.S. dollar? Her answer runs from London bank ledgers to stablecoins and self-custody wallets, arguing that stablecoins are not an entirely new invention but the latest form of a much older offshore dollar story.
Zhang writes that some people see the true global reserve currency not as the dollar itself, but as the eurodollar. The term began as a bank telex address and later came to describe all U.S. dollars held outside the United States.
How the eurodollar market began
The article says that 70 years ago, the Soviet Union and Eastern European countries moved dollars into Banque Commerciale pour l’Europe du Nord in Paris and Moscow Narodny Bank in London to avoid the risk that dollar accounts inside the U.S. could be frozen. Banque Commerciale pour l’Europe du Nord used the telex address “Eurobank,” which is where the name eurodollar came from.
Britain, though, was what turned those offshore dollars into a full credit market. After the 1956 Suez Crisis, the U.K. tightened foreign-exchange controls, and bankers in London began lending against those offshore dollar deposits. That was the start of eurodollar credit. In 1957, after the Bank of England loosened policy further, London became the center of the market.
Zhang argues that the later surge in eurodollars was driven mainly by the U.S. itself. Deposit rates at home were too low to pull dollars back onshore. During the oil crises of the 1970s, oil exporters also left much of their dollar income in London and other offshore banks rather than returning it to the U.S. The market grew from millions of dollars into the trillions, and from that point the eurodollar market was no longer just a European story.
She presents that history along two tracks. On one side, the institutions carrying dollar credit kept changing, from bank ledgers to fintech databases to the reserve structures of stablecoin issuers. On the other, the relationship between users and accounts also changed, from turning money over entirely to institutions to holding direct control over assets today.
Across both lines, Zhang says, three things never changed: the dollar can keep expanding outside the U.S.; final settlement still depends on the U.S.; and the party managing your account is not always the same one that ultimately promises redemption.
When dollars left the U.S., the debtor changed with them
Once a dollar deposit moved to London, something easy to miss had already happened. The money unit stayed the same, but the debtor changed. A New York bank no longer owed that dollar. A London bank did.
London banks then found that they could do more than receive dollar deposits. They could create more dollar balances around them. When a bank extends a dollar loan to a company, it records a claim on the borrower as an asset and, at the same time, creates a dollar deposit as a liability. That deposit can be used right away to pay suppliers, buy equipment, or settle other debts.
Zhang cites Milton Friedman’s line that the source of eurodollars was not a printing press but “the pen of a clerk.” The point, in her telling, is that banks were not creating wealth from nothing. They were using a long-standing rule of credit: if a payment promise is accepted by the market, a ledger-entry dollar can function like a real one. This was the first proof that dollar credit did not need a bank inside the United States to exist.
What fed that market was a wall built by regulation. U.S. Regulation Q capped the rates banks could pay depositors. London banks faced no such ceiling, so they could offer higher returns to attract dollar funding. Economic historian Catherine Schenk, Zhang writes, found in U.K. archives that in June 1955 Midland Bank in London drew about $49 million in 30-day dollar deposits in a single month because it could offer rates above those available from U.S. peers.
After the 1957 sterling crisis, Britain also barred domestic banks from using pounds to finance third-country trade. London banks shifted decisively toward dollar business. Companies and governments seeking funding began going straight to London instead of New York.
The gap widened as global appetite for dollars kept rising while U.S. banks remained constrained. Around 1960, the eurodollar market was roughly $1 billion. Ten years later, it was close to $50 billion. After the 1973 oil crisis, petrodollars flowed back into the banking system through London. By 2007, offshore dollar deposits had reached about $8.9 trillion, more than 150% of U.S. domestic bank deposits. Today, according to the Bank for International Settlements figures cited in the article, dollar credit to non-bank borrowers outside the U.S. stands above $14.3 trillion.
Yet the market carried a built-in paradox from the start. London banks could create dollar deposits, but they could not create Federal Reserve reserves. They could write “I owe you one dollar” on a balance sheet, but if that promise had to be turned into cash, or if market funding suddenly tightened, they still had to rely on U.S. correspondent banks and U.S. clearing rails. Dollar credit had moved outside the U.S. banking system. It had not moved outside the U.S. settlement system.
Three crises exposed three layers of power
Herstatt risk: clearing power decides whether money actually arrives
Zhang starts with June 26, 1974. In New York that morning, traders were waiting for dollars that would never arrive. Hours earlier, they had completed a foreign-exchange transaction with Germany’s Herstatt Bank. Deutsche marks had already been delivered in Frankfurt, and the counterpart dollar leg was supposed to settle in New York. Because of the time difference, Germany’s afternoon was New York’s morning. In that German afternoon, regulators ordered Herstatt to shut down.
The New York side was not waiting on money anymore. It was waiting on a bank that no longer existed. The event showed that what counterparties often hold is only a promise, not final money itself. Between promise and receipt stand the counterparty, the correspondent bank, time zones, and the settlement system. If any one of those links fails, the right shown on a balance sheet does not turn automatically into spendable dollars.
The article says the episode later helped give rise to the Basel Committee on Banking Supervision and left behind the term “Herstatt risk.” For Zhang, this was the first open demonstration of a basic truth inside the eurodollar system: the power to issue a dollar payment promise is not the same as the power to ensure final payment.
The 2008 funding squeeze and the 2020 replay: only the Fed could backstop dollars
The second layer of power surfaced in 2008. European banks, Zhang writes, were holding large volumes of dollar assets, including U.S. mortgage securities, corporate bonds, and other dollar paper. But they did not have a stable base of dollar deposits behind those positions. They were funding themselves through short-term channels such as money market funds, commercial paper, and interbank borrowing.
That structure was cheap in normal times because it assumed those funding channels would stay open. After Lehman Brothers failed, markets began doubting the value of banks’ assets, and short-term lenders stopped rolling their funding. European banks were left with trillions in assets but not enough cash to repay liabilities coming due. A global dollar shortage followed.
That raised an awkward question. These banks were not in the United States and were not supervised by the Federal Reserve. Who would supply dollars? Zhang’s answer is direct: the Fed did. Through central bank swap lines, the Federal Reserve lent dollars to foreign central banks, which then passed those funds on to local banks. In December 2008, swap balances climbed to about $583 billion, roughly one-quarter of the Fed’s total assets at the time. When the pandemic hit in 2020, the same mechanism was used again, and balances came close to $450 billion.
The article treats this as the moment the system’s second layer of authority became plain. Offshore banks can create dollar deposits through lending, but they cannot create the hard dollars needed for settlement and debt repayment. When everyone wants to convert a bank promise into the highest-quality money at once, only the Federal Reserve can absorb the shock.
LIBOR’s fall: pricing power did not stay in London either
The third layer was pricing. The funding costs quoted by London banks evolved into the London Interbank Offered Rate, or LIBOR, which became the benchmark for global loans, bonds, and derivatives. At its peak, financial contracts tied to LIBOR ran into the quadrillions of dollars. London banks were not only creating dollar credit outside the U.S.; for a time, they also held the power to price dollar funding around the world.
But the benchmark had a fatal weakness. It depended on what banks said their funding costs were, not on executed transactions. Once the scandal broke, the flaw was impossible to ignore. Zhang notes that Barclays alone paid $450 million in penalties to U.S. and U.K. regulators for manipulating submissions.
After confidence collapsed, LIBOR was replaced by SOFR, a benchmark built on actual repo transactions. In June 2023, the U.S. dollar LIBOR panel shut down for good. The eurodollar market did not disappear, Zhang writes, but the era when London banks could set the terms was over.
Herstatt, the dollar shortage, and the end of LIBOR each exposed a different hidden layer of control: clearing, lender-of-last-liquidity power, and pricing. Offshore banks gained the ability to expand dollar credit, but they never secured final authority over the system itself.
Fintech moved the dollar account onto the phone
Over the past decade and a half, Zhang says, fintech’s biggest success has been to compress an entire set of banking procedures into a mobile app. Opening an account, exchanging currencies, and sending cross-border transfers no longer require branch visits, paper forms, and waiting several days. The access point for dollar accounts moved from the counter to the software interface.
She uses Revolut and Wise to show that products which look similar to users can rest on very different legal foundations.
Revolut, in this account, chose the route of becoming a real bank. In 2018, it obtained a Lithuanian banking license, allowing it to provide banking services across multiple European countries. Eligible customer balances could become actual bank deposits protected by local deposit insurance. In 2026, it also began moving U.K. customers into a banking entity there. The result is that the same app can contain balances that are legally very different depending on the region and the entity involved: insured bank deposits in some cases, e-money in others, and client funds safeguarded through partner institutions elsewhere.
Wise took a different route. Zhang describes it as closer to an “electronic money machine,” one that does not generally transform customer funds into bank deposits on its own balance sheet. The balance a user sees inside Wise is an e-money claim that Wise promises to honor. Customer funds backing that promise must be kept separate from Wise’s own money. If Wise were to fail, users would in principle have priority over those segregated assets, though whether they recover all of it, and how quickly, would depend on the clarity of records and the local insolvency process.
One model leans toward traditional banking, with heavier regulation and deposit insurance. The other avoids bank-style credit expansion and relies on segregation of funds. The difference comes down to who stands behind the claim, how the money is stored, and what route users would take to recover it if the platform shut down.
Still, both systems share one feature. The account record lives inside the institution’s own database. Users can tap a button in an app, but account opening, freezing, transfers, and withdrawals are still governed by the institution’s systems. What the user owns is a contractual claim, not direct control over the underlying funds.
That is why Zhang says fintech redesigned the entry point for dollar accounts without changing custody itself. The phone replaced the counter. Control did not move.
Stablecoins moved dollar balances onto a public ledger
For Zhang, the real break introduced by stablecoins is not the invention of a risk-free currency. It is the change in how a dollar payment promise exists and how it travels.
Banks and e-money institutions keep their records in internal ledgers. Users can check balances, but moving money still depends on those systems. Stablecoins package a dollar redemption promise into a token that can be held and transferred directly on a public blockchain. The holder no longer has only a line in one company’s database. The holder has an asset that can move across wallets, exchanges, and onchain protocols.
Her comparison is blunt: eurodollars moved dollar credit from New York’s books to London’s books; stablecoins moved dollar balances off any single institution’s books and onto a public ledger that no one institution exclusively owns.
More precisely, she says stablecoins split “one dollar” into two functions, redemption and transfer. Redemption is not new. Issuers back their promise with reserve assets, mainly U.S. Treasuries and bank deposits, held within traditional finance, and final redemption still goes through traditional channels. What is new is transfer. For the first time, a dollar balance can change hands directly on a public ledger without sitting inside any one institution’s internal system. In that sense, stablecoins are not “bankless dollars.” They are dollars whose redemption stays in traditional finance while transfer moves onto a public ledger.
Zhang argues that the force behind eurodollars and stablecoins is the same: global demand for dollars has long exceeded the low-cost coverage that traditional banks are willing or able to provide. People in high-inflation countries want to preserve purchasing power. Businesses engaged in cross-border trade need to settle invoices. Workers abroad need to send money home. The problem is often not that they cannot touch dollars at all, but that they run into exchange controls, account-opening barriers, high fees, and slow approval processes.
Demand does not disappear when the traditional system fails to serve it. It searches for another outlet. In the 1950s, that outlet was London’s banks. Today, Zhang says, it is stablecoins and a borderless public ledger.
At the same time, she stresses that stablecoins retain the same dual character found in eurodollars. They circulate globally, but they still connect back to the U.S. financial system in the end. The largest reserve components of major stablecoins are U.S. Treasuries and bank deposits. Tokens may travel worldwide onchain, yet reserve custody, asset management, and final redemption remain tied to traditional financial infrastructure. Seen from another angle, stablecoins have not weakened the dollar system. They have built a much wider global distribution network for U.S. Treasury debt and dollar assets.
The article supports that point with several figures. As of July 2026, total stablecoin market capitalization stood at about $312 billion. Onchain settlement during 2025 reached about $33 trillion. Tether, the largest issuer, had roughly $141 billion of exposure to U.S. Treasuries. Set beside sovereign holders, Zhang says, that is near the scale of a top-20 foreign holder of U.S. government debt.
She also says stablecoins are not just a digitized copy of eurodollars. Eurodollars relied on banks taking deposits, making loans, and expanding balance sheets while bearing credit and maturity risk. Mainstream stablecoins look more like wrappers around existing dollar assets that are then redistributed, with redemption supported by reserves such as cash and short-term Treasuries. The transfer mechanism is also different. Eurodollars move through correspondent banks and settlement networks. Stablecoins can settle directly onchain. In the past, getting offshore dollars generally required a bank account. Now a blockchain address can be enough.
Regulation is catching up
In Zhang’s reading, the arrival of regulation is no surprise. The eurodollar system eventually saw each of its hidden powers constrained in some form: clearing risk helped produce the Basel framework; dollar shortages made the Federal Reserve the ultimate backstop; and LIBOR’s demise stripped London banks of benchmark-setting authority. Regulation does not stop offshore dollars from emerging, but it also does not leave them unconstrained once they become large enough to threaten the broader system.
The same pattern is now unfolding much faster for stablecoins. The European Union, Hong Kong, and the United States have moved to legislate who can issue stablecoins, what reserves must be held, and whether users can redeem at any time.
Zhang points in particular to the U.S. GENIUS Act of 2025, which she says writes insolvency treatment into law. If a compliant issuer fails, stablecoin holders rank first against reserve assets. She contrasts that with the older eurodollar world, where holders of London bank IOUs waited seven decades without ever seeing their place in the payout order clearly stated in statute. Stablecoin holders got that line within a little more than a decade. In her framing, regulation marks the coming-of-age of this newer dollar form.
Self-custody wallets change control, not the redemption promise
The final section turns to self-custody. Zhang says that banks, e-money institutions, and custodial platforms may differ in structure, but their relationship with users starts from the same premise: users hand over assets, the institution records a balance, and services are built around that balance. A self-custody wallet changes not who redeems the dollar claim, but who controls the asset.
Using Bitget Wallet as an example, she writes that a self-custody wallet does not receive user deposits, does not create a platform balance in its own books, and does not owe the user any stablecoins. The asset record sits on a blockchain, and moving it requires a private-key signature. The wallet provides address generation, key management, transaction signing, onchain connectivity, and an access point to financial services. It is not the creditor holding the user’s funds.
That changes the logic of financial service delivery. In the old model, a person first had to become a customer of an institution and place money into an institution-controlled account before using payments, trading, or wealth products. In the new model, a user can first own and control onchain assets, then connect those assets to services through a wallet. The address is not tied to one wallet company either. As long as the private key remains available, the same address can be accessed through another wallet app.
This is why Zhang separates the role of stablecoins from the role of self-custody wallets. Self-custody answers the question of who can move the money. The stablecoin issuer answers the question of who ultimately redeems it. Whether reserves are sufficient, whether redemption is honored, and whether regulators recognize the structure are still determined by the issuer and the legal framework. The wallet addresses another layer of risk: whether users must surrender control of assets to a platform in order to use financial services at all.
That, she argues, is the real dividing line between self-custody wallets and every earlier form of financial account. Financial services and asset custody are separated for the first time. Payments, trading, yield, and asset management can still be gathered into one interface, but control over transfers no longer has to be handed over. The redemption obligation behind the dollar remains with the issuer. Control over moving the account balance can now stay with the user.
Where the two migrations meet
Zhang closes by describing two intertwined migrations across the past 70 years.
One concerns who carries dollar credit. Eurodollars showed that banks outside the U.S. could create dollars. Fintech companies repackaged the dollar account as a global software product. Stablecoin issuers sealed a dollar redemption promise into a token that moves on a public ledger.
The other concerns the relationship between users and accounts. Revolut and Wise changed how ordinary people access dollars, but the account remained institution-controlled. Stablecoins broke out of the single bank-account format, yet they can still remain inside custodial platforms. Self-custody wallets, in Zhang’s view, are the first tool that lets users access a full financial stack without first giving up control of their assets.
Seventy years ago, dollars moved from New York’s ledger to London’s. Later they entered fintech databases, and then they reached a public ledger in the form of stablecoins. The books changed again and again. The institutions carrying the claims changed too. The promise behind the dollar did not disappear.
The final step is where the real break appears. Once stablecoins sit in a self-custody wallet, users still have to trust that the issuer will honor redemption, but they no longer need to place that money with another platform for safekeeping. The credit relationship remains. Control, for the first time, can stay with the holder.
Zhang ends on that distinction: the dollar never escaped its debtor. This time, though, the account may no longer belong to the debtor.
