How Digital Dollars Are Quietly Rewriting Global Finance
A company most people have never heard of now holds more US government debt than South Korea.
Tether, the issuer behind the stablecoin USDT, held roughly 141 billion dollars in direct and indirect US Treasury exposure as of its most recent quarterly attestation, filed in May 2026 for the period ending March 31. That puts it around 17th on the list of the world’s largest holders of American debt, ahead of countries like South Korea. It has no elected board, no deposit insurance, and until a few years ago was best known as a tool for crypto traders moving money between exchanges at 2am.
That single fact is a good way into a much bigger story. For centuries, money has been a government project. States decided what counted as currency. Central banks controlled its supply. Commercial banks handled the job of physically moving it from one person to another, one country to another, one company to another. That division of labor held because moving money safely required trusted institutions with regulators watching over them.
Stablecoins are quietly splitting that arrangement in two. Governments still issue the currency. They are losing their grip on the rails that currency travels through.
What a stablecoin actually does
A stablecoin is a digital token pegged to a currency, usually the dollar, backed by reserves held somewhere safe, usually short term US Treasury bills. That much is well known.
What matters more is what it replaces. A normal dollar sitting in a bank account has to pass through a chain of intermediaries to get anywhere: your bank, a clearing system, sometimes a correspondent bank if it’s crossing a border, then the receiving bank. That correspondent layer is a network of relationships where banks hold accounts with each other specifically to move money internationally, and it has depended on SWIFT, the messaging system underpinning cross border banking since the 1970s, for the better part of fifty years. Each link in that chain takes time and takes a cut. A stablecoin transfer settles directly between two wallets in minutes, at any hour, on any day, without a bank or a correspondent relationship touching it at any point.
So the actual innovation isn’t digital money. It’s a payment system that doesn’t need banking hours, doesn’t need a chain of correspondents, and doesn’t ask permission along the way. SWIFT still handles the overwhelming majority of international payments today, but for the first time in decades it’s no longer the only way to move money between two countries that don’t share a currency.
The old stack and the new one
For most of the last century, money in most countries moved along roughly the same path. A government issues currency. A central bank manages the supply. Commercial banks hold deposits and process transfers. Card networks and wire systems like SWIFT handle the actual movement between institutions.
Stablecoins run a parallel path. A government still issues the underlying currency. But instead of that currency sitting in a bank account, a private company converts it into a token backed by reserves. That token moves across a blockchain into a wallet, and from there it can be spent, saved, or sent anywhere without a bank at any step.
The token is still a dollar. Nothing about its legal status as currency has changed. What’s changed is who built the road it travels on, and increasingly, who collects the toll.
Where this is already showing up
This isn’t a forecast. In countries where the local currency has been losing value fast, dollar backed stablecoins have become a practical alternative to a local bank account. In Argentina, where the peso has repeatedly lost double digit percentages of its value in a matter of months, people use stablecoins to hold savings the same way earlier generations stuffed dollar bills under a mattress, except now it fits in an app. In Nigeria, freelancers and small exporters get paid in stablecoins because it settles in minutes instead of the days a bank wire takes, and because it sidesteps a naira that has been volatile for years. In Lebanon, where the banking system effectively froze depositors out of their own accounts starting in 2019, a token that lives outside any single bank’s balance sheet is not a convenience. It’s insurance against the bank itself.
None of this required a government decision to dollarize an economy. It happened because someone had a phone, an app, and a reason not to trust the institution that was supposed to hold their money. That is a quieter, more informal version of dollarization than anything that happened through treaties or currency boards in the twentieth century, and it is spreading without anyone having to sign anything.
The Treasury demand loop
Here is the mechanism that explains why Washington, not just crypto exchanges, has started paying close attention.
As stablecoins grow, issuers buy more Treasury bills to back them one for one. More Treasury demand makes it cheaper for the US government to borrow. Cheaper borrowing supports the dollar’s dominant position. A dominant, stable dollar gives more people around the world a reason to hold dollar stablecoins instead of their own currency, which in turn pushes stablecoin adoption higher, which sends issuers back into the Treasury market to buy more bills. That loop is real, but it’s worth being precise about its current size. A stablecoin market in the low 300 billion dollar range is still a small fraction of a Treasury market measured in the tens of trillions. The loop matters more as a direction than as a force that’s already reshaping government borrowing costs today, and its significance mostly depends on whether the market actually grows toward the scale analysts are projecting.
The total stablecoin market sits at roughly 315 to 320 billion dollars as of mid-2026. Standard Chartered has projected it could reach 2 trillion dollars by 2028, which the bank estimates would generate close to a trillion dollars in fresh Treasury bill demand over that period. If that projection holds, this stops being a modest dynamic and becomes a genuinely significant new category of buyer showing up at exactly the moment the government needs one.
That loop is a large part of why the United States passed the GENIUS Act in July 2025, the first comprehensive federal framework for payment stablecoins. It requires issuers to hold reserves close to one for one and publish monthly attestations. The law isn’t really about protecting crypto users, though it does that too. It’s about making sure the entity that’s becoming a significant buyer of US debt is stable enough to keep buying.
What it does to monetary policy
Interest rates work because banks transmit them. When a central bank raises rates, banks pass that through to savers and borrowers, and behavior changes accordingly. That transmission mechanism assumes most money sits inside regulated banks where the central bank can see it and influence it.
If a growing share of savings and payroll moves into stablecoins instead, some of that transmission gets weaker. A saver holding dollars in a stablecoin wallet in Buenos Aires or Lagos isn’t responding to their own central bank’s interest rate decisions at all. They’re effectively subject to US monetary policy instead, whether their government wants that or not. Central banks in emerging markets are only beginning to grapple with what it means to lose visibility into money that used to sit inside supervised institutions and now sits in a private company’s reserve account instead.
Who actually owns the rails
Stablecoin issuers aren’t the only companies that understand this. Visa and Mastercard, the two networks that have quietly taxed a small percentage of nearly every card transaction on earth for decades, are building their own stablecoin settlement capabilities rather than waiting to be replaced. So is SWIFT, which has started piloting blockchain based settlement rather than defending its old messaging rails as the only option. The pattern echoes what happened when Amazon Web Services became the layer other companies’ businesses ran on top of, or when Android became the operating system a huge share of the world’s phones default to. In each case, the company that ends up mattering most isn’t necessarily the one with the most visible product. It’s the one that controls the infrastructure everyone else has to build on.
Banks aren’t going away because of this. Lending, underwriting, and advice remain hard to replace outside a regulated institution. What’s exposed is the part of banking that has always been closer to plumbing than judgment: taking money from one account and reliably getting it into another one. That is exactly the function a stablecoin, a card network’s own settlement layer, or a blockchain based SWIFT pilot can now perform without a traditional bank in between.
Every company wants to become a bank
Once the infrastructure exists, the incentive to use it spreads fast. Stripe, PayPal, and Shopify have all built stablecoin settlement directly into their platforms, because a merchant that settles through a stablecoin instead of a card network keeps a larger share of every sale and gets the money days faster. Amazon and Walmart have both explored issuing their own branded stablecoins for the same reason: a retailer with tens of millions of customers already has the trust a bank spends years building, and none of the fee structure a bank depends on.
This goes beyond retail. Payroll platforms are starting to offer stablecoin payouts to contractors in countries where a bank transfer can take a week. Insurance companies are piloting stablecoin based claims payouts that settle the moment a claim is approved instead of the moment a check clears. Financial services are turning from something a company buys from a bank into something a company builds directly into its own product, the same way cloud computing turned from something a company outsourced into something every company just assumes it can build on top of.
The sovereignty fight underneath all of this
Not every government has responded the same way, and the differences are becoming a genuine geopolitical fault line.
China has taken the opposite approach from the United States. It has banned private stablecoins outright on the mainland while scaling its own central bank digital currency, the digital yuan. According to figures China itself has published, that system had reached roughly 300 million wallet holders and over 16 trillion yuan in cumulative transactions by late 2025. Beijing isn’t trying to compete with dollar stablecoins on their own terms. It’s trying to make sure no privately controlled payment network, dollar based or otherwise, gets a foothold inside its financial system at all.
The European Union has taken a third path, regulating rather than banning. Its Markets in Crypto Assets framework became fully applicable at the end of 2024, and the transitional period for crypto service providers closes in mid-2026. The practical effect has already been visible: Coinbase delisted USDT for European customers in December 2024 because Tether, incorporated in El Salvador, has shown no interest in seeking MiCA authorization. Meanwhile the European Central Bank is building its own digital euro, though Christine Lagarde has been candid that the timeline depends on political decisions well outside the central bank’s control, with first issuance not expected before 2029.
What’s forming here isn’t one global stablecoin system. It’s several competing ones, drawn along the same lines that used to define currency blocs: a dollar stablecoin zone gaining ground informally in weaker currency economies, a Chinese system that keeps private issuers out entirely, and a European system trying to build a regulated middle path before the American version becomes the default by sheer scale.
Beyond payments: the cash leg every other market needs
Every financial market, no matter what it trades, needs two things to change hands at once: the asset and the cash. A stock trade isn’t finished when shares move. It’s finished when the cash settles too, and today that settlement still takes a day or two to clear through the banking system, even for the most liquid securities on earth.
This is why stablecoins matter well beyond payments. Once a reliable, instantly settling digital dollar exists, it becomes the obvious cash leg for tokenizing everything else: Treasury bills, money market funds, corporate bonds, real estate titles, even invoices sitting on a company’s balance sheet waiting to be paid. BlackRock’s tokenized fund BUIDL is already built this way, according to the fund’s own disclosures, combining on chain liquidity with the yield from the Treasuries it actually holds, and settling in something closer to minutes than days.
A tokenized bond without a tokenized way to pay for it is just a database with extra steps. The asset side of finance has been quietly getting rebuilt for a few years now, in pilot programs and institutional trials that rarely make headlines. What was missing was a cash layer that could move as fast as the assets themselves. Stablecoins are what fill that gap, which is why so much of the tokenization experimentation happening right now in institutional finance depends on a working stablecoin market existing underneath it, whether or not the people running those pilots think of themselves as part of the same story.
When the buyer isn’t human
The most genuinely new use case here doesn’t involve people at all.
AI systems are increasingly being built to act on someone’s behalf: buying cloud compute, paying for a dataset, hiring a specialized model to complete a task, compensating another AI agent for a piece of work. None of that fits cleanly into how banking works today. A bank account assumes a human is going to log in, review a transaction, and approve a wire. It was never designed for software making thousands of small, autonomous payments a day without anyone watching each one happen.
A stablecoin, by contrast, doesn’t care who or what is holding the wallet. It settles the same way whether the sender is a person in Lagos paying rent or an AI agent purchasing an hour of GPU time from another company’s system. This isn’t hypothetical anymore. Coinbase built a protocol called x402 specifically for this, letting a server charge an AI agent for an API call the same way a webpage might once have shown an ad, except the payment clears instantly in stablecoins with no human approving it. Google has built a competing standard called AP2. Visa, Mastercard, AWS, Circle, and Cloudflare have all joined the effort to standardize how this works, and the protocol was handed over to the Linux Foundation in 2026 specifically so no single company controls it.
The honest caveat is that most of this volume is still early. Independent trackers have counted well over a hundred million of these machine initiated payments, but by their own admission roughly half of that looks like testing rather than genuine commerce. That’s worth sitting with rather than rounding up. The infrastructure is real, multiple large companies are betting on it, and the transaction counts are growing fast. Whether it becomes a meaningful share of global payment volume or stays a developer curiosity is still an open question, not a settled one.
Why this might not happen the way it looks
None of this is guaranteed to play out the way the trend lines currently suggest, and a fair account has to take the counterarguments seriously rather than waving them off.
Banks are not standing still. Several major banks are building their own tokenized deposit systems, essentially a bank branded version of the same programmable settlement idea, which would let them offer the speed of stablecoins while keeping the customer relationship and the deposit base inside the regulated banking system. If tokenized deposits win out over privately issued stablecoins, the infrastructure changes but the institutions controlling it don’t.
Regulation could also slow this down more than the current trajectory implies. The GENIUS Act explicitly bans stablecoin issuers from paying interest to holders, which limits how competitive they can be against a bank account that does pay something. The EU has already shown it’s willing to force a major issuer out of a market rather than let it operate unregulated. If more jurisdictions follow China’s approach instead of America’s, the addressable market for privately issued stablecoins shrinks considerably.
And CBDCs remain a real competitor, not a footnote. China’s digital yuan already has a real user base most Western commentary underestimates. If more countries decide monetary sovereignty is worth the slower rollout of a state run digital currency, dollar stablecoins may end up dominant in exactly the fragile currency economies they’ve already captured, and largely locked out everywhere else.
What actually failed, and what it proved
None of the argument above is worth much if it can’t survive contact with the times this has already gone wrong.
Terra’s stablecoin UST collapsed in 2022 because it was backed by an algorithm and a sister token rather than real reserves, wiping out tens of billions of dollars in days once confidence broke. USDC briefly lost its peg in March 2023 when Silicon Valley Bank failed and a portion of Circle’s reserves were temporarily stuck inside it, a reminder that even a well regulated, fully reserved stablecoin is still only as strong as the traditional banks it keeps its reserves in.
Concentration is a live risk too. Two issuers control more than 80 percent of the entire stablecoin market, which means a failure at either one wouldn’t stay contained to a niche corner of crypto. And because stablecoin transfers move quickly across borders with far less friction than a wire transfer, regulators in the US and EU have both pushed hard on monitoring and sanctions compliance, aware that the same speed that makes stablecoins useful for a freelancer in Lagos makes them useful for someone trying to move money somewhere it isn’t supposed to go.
The question this actually leaves open
The defining question in finance over the next decade probably isn’t who gets to issue money. Governments have controlled that for centuries and will likely keep controlling it. The open question is who controls the networks that money moves through once it’s issued, and whether that control stays public or drifts toward whichever private company builds the fastest, cheapest rails.
Right now the answer is still being written, country by country, wallet by wallet, and it’s being written faster than most of the institutions it affects have noticed.
