A stablecoin issuer can become a larger holder of government debt without creating the same amount of new demand for that debt. The missing information is what the token buyer held before making the purchase. That accounting question belongs at the centre of the latest discussion about digital dollars and Treasury markets.
Cointelegraph's September 15 report describes Carolyn Wilkins's argument that stablecoins could support the dollar's international role and increase Treasury demand, while introducing a reverse channel through redemptions. Her Bank of England speech is conditional about scale: it does not describe the sector as a current major threat to the Treasury market. The useful investment implication is a framework for tracing flows, not a prediction that stablecoins will reliably lower government borrowing costs.
An issuer balance sheet is not a net-flow measure
Consider two hypothetical buyers. One sells an investment already backed by Treasury bills to obtain a stablecoin. The other reallocates money from an asset outside that market. If the stablecoin issuer buys bills in both cases, its reported holdings increase. But the first transaction may largely move the same exposure between financial wrappers, whereas the second can represent a more substantial shift in demand.
This is an accounting example, not an estimate of the current customer mix. It shows why issuer reserve growth and net market demand should not be treated as interchangeable. The seller, the buyer, the reserve manager and any intermediate fund all have to be considered before adding up the effect. A token supply series cannot, on its own, reveal the complete chain.
A similar distinction applies to payment activity. A stablecoin can change hands repeatedly while the outstanding supply remains unchanged. Faster turnover may indicate a useful payment service without requiring a matching increase in reserves for every transaction. Conversely, a growing stock of tokens can reflect balances being held rather than spent. Payment volume and reserve assets answer different commercial questions.
Wilkins explicitly highlights funding origins, including the difference between movement from Treasury money-market funds and movement from other currencies or assets. This is a reason to demand a funding map alongside a growth forecast. It also means that the interests of an issuer, a bank losing balances and a government issuing bills need not be identical.
The evidence belongs to the short end of the curve
The BIS working paper on stablecoins and safe-asset prices studies the effect of stablecoin flows on three-month Treasury bill yields. Its revised analysis uses daily data running through March 2026 and an identification strategy intended to separate the flow effect from other influences. The researchers find downward pressure associated with inflows. That is evidence about a particular market segment and historical setting, not a rule for every government bond.
The distinction is financially important. A short-dated bill returns principal soon, while a long bond embeds exposure to many future periods. An additional buyer concentrated at the short end cannot simply be assumed to lower yields by the same amount across the curve. Expectations for policy, inflation, issuance and compensation for longer commitments remain relevant to different instruments.
An empirical estimate also cannot be multiplied indefinitely by an optimistic token-growth scenario. Market depth, competing buyers, the supply of bills and the origins of new balances may change as the sector expands. The right use of the research is to identify a transmission mechanism worth measuring. Treating it as a guaranteed saving on future public debt would claim more than the evidence establishes.
Redemption quality depends on the route back to cash
Circle's reserve disclosure gives a concrete example of the assets behind this discussion. It describes cash, short-dated US Treasuries and overnight Treasury repurchase agreements, including holdings within its reserve fund arrangements. This is issuer-specific information, not a description of every stablecoin. It also demonstrates why a reserve total alone is insufficient: the form and availability of assets matter.
A redemption request needs spendable cash at the appropriate place and time. Cash already available, securities reaching maturity, overnight arrangements unwinding and outright asset sales can provide different routes. It would be inaccurate to assume that every redemption automatically produces an immediate Treasury sale. The relevant stress question is how those routes perform together when requests are unusually concentrated.
There is a credible stabilising case. Short-duration assets reduce some price exposure, and recurring payment balances could be less flighty than purely speculative balances. But these are conditions to examine, not universal protections. An asset can be high quality while the process of converting it into a customer payment encounters timing or operational friction.
For bond-market analysis, the most useful new evidence would combine reserve composition with issuance and redemption flows, the sources of customer funds and liquidity arrangements. For an issuer, sustained useful balances and dependable redemption are different achievements from rapid gross transaction growth. Digital dollars may broaden access to dollar assets; whether they add durable financing demand depends on the financial movements surrounding the token.
