What if the United States spent years defending the dollar, then quietly made it harder for the rest of the world to hold a digital version of that same dollar? That question sat with me after a Circle executive put a blunt number on the table: as much as one trillion dollars in overseas demand that might never show up if Washington treats dollar stablecoins as a problem instead of a product people already want.
Why A Trillion-Dollar Warning Hits A Nerve
Nikhil Chandhok did not present a spreadsheet with footnotes. He offered a market instinct. People outside the United States already believe in the dollar. Give them a clean digital claim on that currency and they will use it. Resist the rails, and that demand does not vanish. It just finds another venue, another issuer, another country that is less conflicted about the opportunity.
I’ve found that big round numbers in crypto debates often get treated as slogans. This one deserves a slower look. Chandhok’s figure was an argument about potential demand, not a receipt for money already lost. He also did not pin the entire claim on a single statute. That distinction matters, because policy talk loves to collapse a messy market into one villain rule.
I was so confused. I was like, why don’t you want this, America? You can raise a trillion dollars from the world, because they believe in the dollar.
– Nikhil Chandhok
The setting was a conversation about the so-called agentic economy, where software agents pay for work without a human clicking confirm on every invoice. Circle’s executive sat with investors and builders who think the next wave of commerce will not wait for correspondent banks to finish their paperwork. That context is not decoration. It is the reason the trillion-dollar line landed with heat.
The Demand Story Starts Outside American Banks
Dollar-backed stablecoins are, at heart, a workaround for access. Someone in another country can hold a digital token tied to the dollar without waiting for a traditional international transfer to crawl through a chain of correspondent accounts. That sounds simple. It is not simple if you have ever watched a cross-border payment sit idle for days because the old pipes still run the show.
Chandhok tied the pitch to money that is already moving, just slowly. He estimated that roughly three trillion dollars sits in transit across the international banking system at any moment. Idle cash. Not earning much. Not settling fast. Stuck because, in his words, the technology and settlement protocols are very old.
Is that estimate precise to the last billion? Probably not. Does the direction of the point feel right to anyone who has waited on a wire? Yes. Speed is not a crypto talking point anymore. It is a customer complaint with a price tag.
USDC Is Circle’s Bet On Instant Settlement
Circle has spent years pushing USDC as the institutional-friendly dollar token. In midyear institutional work, the firm connected settlement tools with a custody and transfer specialist so banks and trading desks could move USDC and, in many cases, pay out in local currency through providers in more than fifty countries. Minutes instead of multiple days. That is the product promise, stripped of marketing fog.
On that partner platform, stablecoins accounted for 69 percent of digital asset transaction volume in the second quarter of 2026. Read that again. Not a niche side pot. A majority of the flow. Circle’s gateway approach lets firms manage one USDC balance across supported chains while compliance controls stay attached: approvals, screening, the unglamorous work that keeps a treasury team employed.
In my experience, this is where the public debate and the back-office debate split. The public fight is about whether tokens are “real money.” The back-office fight is about cut-off times, failed wires, and weekend settlement that still behaves like 1998.
How Stablecoins Touch Treasury Demand
Here is the part Washington actually cares about. If a payment stablecoin issuer must hold reserves equal to tokens outstanding, those reserves need a home. Eligible assets under the GENIUS framework include dollars, certain deposits, short-term US Treasury securities, and other qualifying holdings. Tokens in wallets can become bills on a balance sheet.
That is not automatic alchemy. Extra token demand does not magically become a one-for-one Treasury purchase. Still, officials have been circling the idea that overseas users who never held dollars before could lift demand for short-term government debt if they crowd into dollar tokens that must be reserved.
Committee analysis using data through September 2025 estimated that the two largest dollar issuers had increased Treasury bill holdings by about seventy billion dollars since 2022. That is a real number with a real time window. It is also not Chandhok’s trillion. Mixing the two is how a policy conversation turns sloppy.
| Claim | What It Actually Means | What It Does Not Prove |
| $1 trillion overseas demand | Potential appetite for dollar tokens | Money already lost by the US |
| $3 trillion in bank transit | Idle value in old settlement rails | A precise daily float audit |
| $70 billion extra T-bill holdings | Issuer reserves moved into bills since 2022 | Every new token buys a bill |
Perhaps the most interesting aspect is how quickly a payments product becomes a fiscal product. Once reserves sit in bills, stablecoins stop looking like a sideshow and start looking like a distribution channel for short-duration government paper. Love that or hate that, it is a different argument from “crypto is risky.”
The Rulebook Is Still Being Written In Public
Issuers are not waiting for vibes. They are waiting for reserve rules, application clocks, and effective dates. Late September brought proposed reserve standards and a separate track for insured state member banks that want to issue payment stablecoins. Short-term Treasury bills showed up among assets that could back tokens issued by supervised firms. No surprise there. Bills are liquid. Bills are familiar. Bills are politically easier to defend than a grab bag of credit risk.
The process is bureaucratic on purpose. Once an application is substantially complete, the clock can run 120 days. Interested parties get a 60-day comment window after notices hit the register. The main issuer restrictions are expected to take effect in mid-January 2027, unless final rules land earlier and pull the start date forward. If you build products on legal sand, those dates are not trivia. They are the difference between shipping and stalling.
- Reserve assets must at least match tokens in circulation.
- Short-term government securities sit near the center of the eligible list.
- Bank issuers face a formal application path, not a handshake.
- Comment periods and decision clocks will shape who launches first.
I’ve watched enough financial rulemaking to know the quiet phase is where the real product design happens. Marketing decks talk about freedom. Legal memos talk about haircuts, attestation, and who is on the hook if a redemption queue forms on a Friday night.
Agent Payments Are The Sneaky Demand Engine
The panel was not only about wires. It was about software that holds balances and pays other software. Circle launched an agent stack in the spring so programmed wallets could hold assets and fire payments without a person in the loop for every micro-task. By the second quarter, the company reported more than 900 paid services on that stack. On one agent payment protocol, USDC made up 99.3 percent of volume.
Those figures describe activity on that protocol. They do not measure the entire artificial intelligence payments market. Still, they hint at a habit: when machines need a dollar that moves at machine speed, they do not open a checking account in three countries. They grab a token.
Brokerage work cited agent payments as one possible source of future USDC demand. Adjusted stablecoin transaction volume, after stripping bots and high-frequency noise, was estimated around eleven trillion dollars in 2025 and was running at an annualized clip near seventeen trillion through July 2026. Even if Congress stalled a separate market-structure bill during the September session, the volume story did not need a standing ovation from every committee room to keep growing.
Machines do not care about nostalgia for correspondent banking. They care about finality measured in minutes.
Energy, Robots, And A Very Different Kind Of Cost
Cathie Wood took the same stage and yanked the conversation toward power. Her claim was stark: if nuclear power had not been wrapped in heavy regulation, electricity prices in the United States would be about half of what they are now. She called the current setup a travesty and labeled regulation a menace. That 50 percent figure was her judgment, not a published model dropped on the table.
Why mention power in a stablecoin piece? Because the agentic economy is not only a payments story. It is a compute story. Data centers eat watts. Watts have a price. If AI agents become heavy users of dollar tokens, the cheap-power debate and the cheap-settlement debate start sharing a roof.
Wood also argued that humanoid robots are almost a thousand times more complex than robot taxis. She put a later timeline on humanoids than the late-2028 to 2029 window often associated with one prominent industrialist, and she described that faster calendar as “Elon time.” Healthcare investors, she added, remain cautious about funding care-related AI because a medical error is not the same class of failure as a buggy consumer app. Emad Mostaque offered a cultural contrast: roughly 80 percent of Americans fear AI, while roughly 80 percent of people in China view it positively, in his telling.
None of that proves Chandhok’s trillion. It does show the room was thinking in systems: money, energy, labor, and software arriving at the same intersection. That is a more adult frame than “tokens good” versus “tokens bad.”
What “Restrictions” Might Actually Mean
People hear restrictions and picture a ban. Reality is usually duller and more damaging. A reserve rule that is too narrow. A licensing path that takes longer than a product cycle. A bank application process that scares boards. An overseas project that cannot plug into American issuance without a legal maze. Death by delay is still a policy choice.
Chandhok’s confusion is easy to share if you accept his premise. If the world wants dollars, why make the digital wrapper harder than the analog one? The counterargument is also easy to share. Fast money can hide sanctions evasion, fraud, and runs. A token that settles in minutes can also fail in minutes if reserves are sloppy.
- Write reserve rules that are tight enough to survive a redemption wave.
- Keep issuance open enough that dollar tokens do not migrate to friendlier jurisdictions.
- Treat overseas users as a demand source, not only as a risk file.
- Measure Treasury impact with actual holdings, not slogans.
- Build agent payment tools with the same screening standards used for human transfers.
That list is not a manifesto. It is the minimum adult agenda if you believe both things at once: the dollar should stay the global unit of account, and the pipes should not look like antiques.
The Quiet Contest Other Countries Are Already Playing
While American agencies draft notices, other markets are arguing about local-currency tokens and liquidity buffers. South Korea’s debate over won stablecoins is a reminder that the fight is not uniquely American. Every treasury in a mid-size economy is asking the same question with a different flag on it: if digital cash is coming, whose cash gets used?
If the United States treats dollar tokens as a nuisance, someone else will treat them as export infrastructure. That does not require a conspiracy. It only requires a faster license and a friendlier reserve menu. Money is opportunistic. It always has been.
I’ve found that sovereignty arguments get loudest right when the product is winning on convenience. That is not an accident. Convenience is how reserve currencies stay reserve currencies. Inconvenience is how they leak.
What The Trillion Figure Gets Right And Wrong
Give Chandhok this: he named the psychological core of dollar demand. People reach for dollars when local money wobbles, when invoices are priced in dollars, when savings need a harder unit. A token that moves on a weekend is a better wrapper for that instinct than a wire that closes at 3 p.m. on a Friday.
Withhold this: a trillion is not a loss statement. It is a ceiling painted on the wall. Policy can reduce that ceiling. Policy can also raise it by making issuance boring, reserved, and boring again. Markets reward boring money. They punish exciting money that cannot redeem at par.
Demand stack, in plain English: Access to dollars without a US bank account Settlement that does not sleep Reserves that can be explained to a regulator Software agents that can pay without a human click
If any layer fails, the trillion shrinks. If all four work, the number stops sounding theatrical and starts sounding like a planning assumption.
A Practical Read For Anyone Holding Dollars Digitally
You do not need to love Circle to care about the plumbing. You need to care about where reserves sit, how fast redemptions work, and whether the issuer you use will still be welcome under the next draft of the rulebook. Brand loyalty is a weak hedge. Transparency is a better one.
Watch three things over the next few months. First, the shape of final reserve definitions. Second, how many banks actually file to issue. Third, whether overseas programs keep buying short-term government paper as token supply grows. Those are the tells. Quotes on a conference stage are the color commentary.
And yes, keep an eye on agent volume. If machines become steady customers of dollar tokens, the user base stops looking like traders and starts looking like infrastructure. Infrastructure gets regulated. It also gets defended, eventually, because too many invoices depend on it.
The Uncomfortable Middle Ground
America can want two things that pull in opposite directions. It can want tight control over dollar issuance. It can also want the rest of the planet to keep choosing the dollar when a cheaper, faster alternative appears. Stablecoins force that tension into the open. Pretending the tension is not there is how you sleepwalk into someone else’s rail.
I do not buy the idea that every restriction is an own-goal. Some restrictions are how you keep a token from becoming a rumor during a panic. I also do not buy the idea that the old correspondent system is good enough because it is familiar. Familiar and expensive is still expensive.
So here is the human version of the policy memo. If people already want dollars, let them hold a reserved digital claim with rules that are strict and dates that are real. If the United States would rather lecture the market than serve it, the market will take notes. It always does. The trillion-dollar line may be too neat. The direction of travel is not.
The next chapter will not be written on a panel. It will be written in reserve reports, application files, and the quiet decision of a treasurer in another time zone who just wants a dollar that arrives before the weekend starts. That person is the real audience. Everything else is stage lighting.
