Can regulation manufacture a new source of Treasury demand?

The GENIUS Act is the first experiment.

What if some marginal Treasury demand does not have to be attracted by a higher yield? What if regulation creates a financial liability whose reserves are allowed, and in some cases economically encouraged, to sit in Treasury bills?

The GENIUS Act gives us a chance to find out.

Under the law, permitted payment stablecoin issuers must maintain identifiable reserves backing their outstanding stablecoins on at least a 1:1 basis, but the more consequential provision for the Treasury market is what Congress allows those reserves to contain. The statutory list includes U.S. currency and certain deposits, Treasury securities with 93 days or less remaining to maturity or originally issued at maturities of 93 days or less, certain Treasury-backed repo transactions, and government money-market funds invested in eligible reserve assets. The law also provides limited room for regulators to approve other similarly liquid federal government assets and permits certain eligible reserves in tokenized form.

Corporate bonds, equities and Bitcoin are absent from the list. Congress has effectively attached a private dollar liability to a restricted reserve pool in which short-duration government securities occupy a natural position.

This does not mean every new stablecoin dollar produces a new Treasury-bill purchase. An issuer can hold cash or deposits, use qualifying repo, or choose among the other eligible reserve assets. But if the quantity of regulated stablecoins outstanding grows, the required reserve pool has to grow with it. That creates a path by which demand that begins somewhere outside the Treasury market can eventually reach it.

The person supplying the money may have no interest in Treasury bills at all. They may simply want dollars in digital form.

We can already see the balance sheet

The reserve model predates GENIUS, which means we do not have to construct an imaginary stablecoin issuer to understand how it might work. Circle already provides a useful example.

As of June 30, 2026, Circle reported approximately $73.3 billion of USDC reserves, with $61.9 billion held in the Circle Reserve Fund and $11.4 billion held in cash. Roughly 84% of USDC reserves were therefore sitting in the reserve fund. Circle’s SEC filing describes that fund, which is managed by BlackRock, as holding U.S. Treasury securities with remaining maturities of three months or less, overnight U.S. Treasury repurchase agreements and cash.

This is close to the operational path I followed in my earlier series on stablecoin reserves. A customer acquires USDC because they want a digital dollar for payments, trading, settlement or simply dollar exposure on blockchain rails. Their decision does not require them to examine a Treasury auction or decide whether a three-month bill offers sufficient yield. Yet the resulting liability has to be supported by assets, and most of Circle’s USDC reserves currently sit in a fund whose portfolio is built around short-term Treasuries, Treasury repo and cash.

The Treasury security at the bottom of this structure has not changed. What has changed is the economic reason for holding it. Some of the demand for that security can originate with demand for another financial product entirely.

Tether shows what this looks like at greater scale, although its regulatory position differs from that of a U.S. permitted payment stablecoin issuer under GENIUS. At the end of 2025, Tether reported more than $122 billion of direct U.S. Treasury holdings and more than $141 billion of total direct and indirect Treasury exposure, including overnight reverse repurchase agreements. As of March 31, 2026, it reported approximately $141 billion of direct and indirect exposure to U.S. Treasury bills.

Tether’s holdings are not evidence that GENIUS itself has created Treasury demand; its reserve structure existed before the law. They do show that the connection between demand for digital dollars and ownership of U.S. government securities can become very large. Demand for USDT created liabilities on Tether’s balance sheet, and the reserve portfolio supporting those liabilities became one of the larger pools of Treasury exposure in the world.

GENIUS takes the underlying relationship between stablecoin liabilities and reserve assets and places constraints around it in federal law.

Treasury was already asking the same question

The Treasury Borrowing Advisory Committee was studying the implications before GENIUS passed. In April 2025, TBAC examined how stablecoin growth could affect Treasury demand and reached a distinction that is more useful than simply asking how many Treasuries stablecoin issuers might buy.

The Committee concluded that stablecoin growth coming from unbanked market segments would be positive for Treasury-bill demand, while growth occurring at the expense of money-market funds would likely be neutral. The reason becomes clear once the source of the stablecoin funding is considered. If money that was already sitting in a money-market fund invested in Treasury bills moves into a stablecoin whose issuer buys Treasury bills, the reported Treasury holdings of the stablecoin issuer increase even though total demand for bills may barely change.

After GENIUS became law, TBAC returned to the subject in July 2025 and said increased stablecoin issuance could create a new source of demand for short-maturity Treasury securities. It again warned that some of the apparent increase could be offset if stablecoins substitute for bank deposits, money-market funds or other cash-like instruments.

Treasury Secretary Scott Bessent has since discussed the connection more directly. Speaking at the Treasury Market Conference in November 2025, he said Treasury was monitoring the growth of money-market funds and stablecoins because both are large buyers of Treasury bills. He described a stablecoin market of roughly $300 billion that could, in his estimate, grow tenfold by the end of the decade, and argued that growth in these pools would bring additional demand for bills.

The tenfold figure is a projection and should be treated as one. The more interesting point is that the institution responsible for financing the federal government is explicitly watching stablecoin growth as a potential source of demand for the securities it issues.

There is another dimension to that relationship. Treasury controls the maturity composition of the debt it issues, and Bessent has said that if structural demand for particular products or tenors changes, Treasury will adjust how it allocates issuance. GENIUS therefore sits in an unusual position: regulation can help create a class of private dollar liabilities with a reason to hold short-term government assets, while Treasury controls the supply and maturity of the securities those balance sheets may buy.

That is separate from the Federal Reserve’s monetary-policy tools. GENIUS does not give Treasury control over interest rates, the Fed’s balance sheet or the public’s demand for stablecoins. It does, however, create a channel through which growth in private digital-dollar demand can become relevant to Treasury’s own debt-management decisions.

The potential buyer base also extends beyond the domestic financial system. Someone abroad can demand a dollar stablecoin without ever deciding to buy a Treasury security themselves. If that demand expands the stablecoin’s reserve pool, some of it can reach short-term U.S. government securities through the issuer’s balance sheet. The same architecture can therefore widen the private distribution of dollars while widening the potential buyer base for short-term U.S. government debt.

That diversification may matter for another reason. Foreign official holders sometimes need to liquidate dollar reserves for reasons unrelated to Treasury valuations, including currency intervention and domestic liquidity needs; the Federal Reserve’s FIMA Repo Facility exists in part to give those holders access to dollars without forcing Treasury sales into the open market. A larger reserve-driven buyer base tied to private dollar liabilities would not eliminate that vulnerability, but it could diversify Treasury demand away from foreign official balance sheets whose reasons for selling may have little to do with the price of the securities themselves.

Treasuries are already held for reasons that extend beyond yield. They serve as reserve assets, liquidity buffers and collateral across different parts of the financial system. Those uses do not make their holders indifferent to price, but they do mean that Treasury demand is shaped by more than valuation alone. GENIUS potentially adds another motive to that mix: short-term government securities can be held because they are eligible assets backing an outstanding private dollar liability.

The 93-day limit matters

The effect, if it develops, would not be distributed evenly across the Treasury curve. GENIUS directs the ordinary Treasury portion of the reserve pool toward securities with 93 days or less remaining to maturity, or securities originally issued with maturities of 93 days or less. A stablecoin reserve portfolio operating within those rules therefore does not provide a direct new buyer for a 10-year note or a 30-year bond.

That concentration at the front end matters because bills serve as Treasury’s financing “shock absorber”: their issuance can be adjusted relatively quickly as the government’s borrowing needs change. GENIUS has created a federal framework for private dollar liabilities whose reserve rules make those same short-duration government securities eligible reserve assets.

This brings us back to the problem in Part 1. The federal government continuously has to refinance maturing debt while raising additional money to finance deficits. Someone has to absorb that issuance, and the yield required to attract the marginal buyer affects the government’s financing cost.

Price is one way of finding that buyer. If investors are unwilling to absorb enough Treasury securities at the prevailing yield, the price can fall and the yield can rise until additional demand appears. GENIUS introduces another route by which demand can reach the same market. If stablecoin liabilities grow, the eligible reserve assets supporting them must grow as well, and some of those reserves can be held in Treasury bills.

The economic decision that begins this chain is therefore different. The person supplying the money does not have to decide that Treasury bills are cheap or that the yield is attractive. They may simply decide that they want a stablecoin. The reserve structure can make the Treasury purchase downstream of that decision.

How large could the channel become?

Consider a hypothetical regulated stablecoin market with $1 trillion outstanding. This is a scenario, not a forecast. Under a 1:1 reserve requirement, approximately $1 trillion of eligible reserve assets would have to stand behind those liabilities.

It would be wrong to translate that immediately into $1 trillion of Treasury bills. Issuers could hold portions of their reserves in cash and deposits, qualifying Treasury repo, government money-market funds and other assets permitted under the statute. But even a fraction of a reserve pool that large flowing directly or indirectly into short-duration government securities would be substantial.

There is also a persistence to the reserve requirement that ordinary discretionary investment demand does not have. If $500 billion of qualifying stablecoin liabilities remain outstanding, at least $500 billion of identifiable eligible reserve assets must remain behind them. An issuer can change the composition of those reserves within the permitted universe, and users can redeem their stablecoins, but the reserve requirement follows the outstanding liability.

This is where the structure of the demand starts to matter. The allocation can still respond to economics and move among eligible assets, but the reserve pool itself exists because the liability exists.

The harder problem is determining how much of the resulting Treasury demand is actually new.

Where did the money come from?

Suppose an investor has $1,000 in a government money-market fund that already owns Treasury bills. The investor withdraws the money and buys $1,000 of stablecoins, after which the stablecoin issuer invests the proceeds in Treasury bills. The issuer can now report an additional $1,000 of Treasury holdings, but the Treasury market may have gained almost nothing because one source of bill demand has effectively replaced another.

TBAC’s April 2025 analysis makes essentially this point by treating stablecoin growth at the expense of money-market funds as likely neutral for T-bill demand.

The calculation changes when the stablecoin attracts money from somewhere that previously generated little or no Treasury demand. TBAC specifically identified growth from unbanked market segments as positive for T-bill demand. In that case, the stablecoin reserve structure can introduce Treasury exposure where little existed before.

This is why simply adding up the Treasury holdings reported by Circle, Tether and future stablecoin issuers would give us the wrong answer. Those holdings tell us what sits inside the stablecoin reserve system. They do not tell us what the owners of those dollars held before entering it.

A more useful framework is:

Net-new Treasury demand from stablecoins = Treasury demand created through stablecoin reserves − Treasury demand displaced at the source of the funds.

The first part is becoming increasingly visible through issuer disclosures, reserve reports and regulatory filings. The second requires us to understand where stablecoin funding originates and what assets that money would otherwise have financed.

That is also why I am reluctant to describe the GENIUS Act as having “created” a new Treasury buyer yet. Congress has created the regulatory channel. Whether the channel increases aggregate Treasury demand depends on what flows into it.

Now we get to watch it happen

On August 17, 2026, Treasury issued its proposed rule implementing Section 3 of the GENIUS Act, covering the issuance, offering and sale of payment stablecoins in the United States. Treasury currently describes January 18, 2027 as the expected effective date, while the statute provides that the Act becomes effective on the earlier of 18 months after enactment or 120 days after federal regulators issue final implementing regulations.

That timing makes this less of a theoretical exercise than it was a year ago. Congress has defined the reserve architecture, existing issuers have shown that stablecoin reserve portfolios can accumulate tens of billions of dollars of short-term government assets, and TBAC has already identified the condition that determines whether the resulting demand is additive or merely displaced.

Part 1 asked who absorbs the next dollar of Treasury issuance. GENIUS presents one possible answer: some Treasury demand may originate with people who never intended to buy a Treasury security. Their decision was to hold a dollar in digital form, while the regulated balance sheet behind that dollar determined where at least some of the corresponding assets could go.

We can now observe whether that channel grows.

The more difficult question is whether it brings Treasury a new buyer or simply gives an old buyer a new wrapper.

Thank you for reading,

Kirandeep Kaur

Up Next : Part 3: A Stablecoin Dollar Isn’t Necessarily a New Treasury Buyer.

Part 1: America needs someone to finance the next Treasury dollar.