The note arrived without a press conference. No podium, no red banner, no three-word headline engineered for a screenshot. Just two staff economists at the Board of Governors, names typeset at the top of a FEDS Note, proposing across a handful of unexcited pages that the Federal Reserve begin counting stablecoins inside M1 and M2. I read it twice on a wet afternoon in Hong Kong and noticed how little it wants to be read. No urgency. No rhetorical escalation. No claim that anything is about to change. That restraint is the tell. Hype announces itself; framework does not. The documents that end up mattering in this industry arrive quietly and then never leave.
To see why a note about a statistical table deserves that attention, you have to look at the table. H.6 is the Fed's money stock release — the running account of how much money exists in the United States and in what form. M1, the narrow measure, sits near $19.9 trillion: currency in circulation plus checkable deposits. M2, the broad measure, sits near $23.2 trillion, adding savings deposits and retail government money market fund shares. By the Fed's own description it is one of the most-downloaded datasets on FRED — a primary pulse reading of the American economy, watched by rate desks and pension allocators who rarely agree on anything.
The Payne–Styczynski proposal sits on top of that. It does not amend H.6. It commits the Federal Reserve to nothing. It offers a method — and methods, unlike policies, are hard to unwind once published.
What strikes me is the shift in the question being asked. For most of the past six years, the debate around stablecoins was about legitimacy: whether an issuer could operate, whether reserves were sufficient, whether the instrument was a security. That is a compliance cliff, and the industry spent enormous energy climbing it. This note does not engage with that argument at all. It assumes the answer and moves to circulation — how much of this is actually moving, and through whose accounts. Regulation asks permission. Statistics ask arithmetic. The second question is much harder to refuse.
None of it is happening alone, either. The Office of the Comptroller of the Currency has committed publicly to a rulemaking timeline; the GENIUS Act carries an effective date; researchers at the New York Fed are working the deposit-outflow angle in parallel. Read together, the three lines form a stack — issuance legitimacy from Congress, operating rules from the OCC, and now measurement from the Board.
The method has three moving parts. The second is where the elegance breaks.
First: classify by function, not by wrapper. A stablecoin used as a daily transaction medium belongs in M1; one held as a value store or as crypto-market collateral belongs in the broader M2 bucket, outside M1. This is not a new idea dressed up. It is the logic the Fed applied in 2020, when it stopped distinguishing savings deposits from transaction deposits and reclassified by what the balance is actually used for. The taxonomy is inherited. Only the instrument is new.
Second: the double-counting problem. This is the hard knot, and the staff say so themselves, which is rarer and more interesting than the headline. Stablecoin reserves sit overwhelmingly in bank deposits, U.S. Treasuries, and government money market fund shares — every one of which is already counted inside M1 or M2. Add stablecoin face value on top and you have counted the same dollar twice. Reserve composition varies by issuer and is disclosed only in aggregate, which is why the staff cannot simply subtract. A stablecoin is not new money arriving from outside the system; it is an existing dollar wearing a different costume. Accounting for the costume as though it were the person is how a money stock series loses its credibility.
Third: the data gap. There is currently no separate tracking line for tokenized deposits — bank liabilities placed on a chain that behave like deposits because they are deposits. Without that line, de-duplication cannot be done cleanly. The Fed is disclosing the boundary of its own instrumentation in public.
Then comes the piece that turns method into structure. The GENIUS Act, Section 4(a)(11), prohibits paying interest directly on stablecoins. Strip the yield away and a stablecoin stops behaving like a savings instrument and starts behaving like a checking account — which lands it in M1. Not because the Fed decided so, but because Congress removed the attribute that would have pushed it elsewhere. Two institutions, working from opposite ends, arrive at the same classification without appearing to coordinate. That convergence is the actual finding, and it sits in the footnote of a statistics note.
Now, scale. If every stablecoin were folded in tomorrow, the disturbance to a $19.9 trillion M1 would be a rounding artifact. M2 is larger still, and the same arithmetic holds: the aggregate barely notices. The signal is not in the size. It is in the fact that the question is being asked at all. Echoes of early hype in the quiet of current data — except here the noise never came, and the quiet is the news.
Which brings me to the reading I keep returning to, and it is not the optimistic one.
Everyone treats this as institutional embrace. I read it as a tide gauge. You do not install a tide gauge because you love the water; you install it once the water is already in the street. The New York Fed's parallel work is explicitly about deposit outflow — the possibility that stablecoin growth drains bank deposits and thins the credit base banks fund from them. That is a stability question, not an adoption question. Measurement is often the first act of containment. The taxonomist arrives before the supervisor.
There is a second unease. A yardstick gets cut to the length of the ruler you own. Without tokenized deposit tracking, the functional categories risk becoming aesthetic — a clean taxonomy drawn over an admitted hole. And “daily transaction medium” is a judgment, not an observable. Whether a given stablecoin qualifies for M1 will be decided by staff discretion, and that discretion is exactly where the next argument lives. I have spent enough time staring at Aave's rate curves to know how much authority hides inside a parameter someone chose by hand.
A caution about the reading itself, too. FEDS Notes are staff work. They carry no vote, no dot plot, no obligation. Anyone who reprices a portfolio because two economists published a method has confused a research library with a policy committee.
Watching from Hong Kong, the pattern looks familiar for different reasons. The licensing regime here was assembled with one eye on the regional scoreboard, and measurement frameworks tend to follow the same instinct: define the category, then own the definition.
So watch one item and ignore the speeches. Watch for a tokenized deposit line item appearing inside H.6. Not a hearing, not a commissioner's remark, not a consultative paper. A line item. The calendar around it is already fixed: OCC rules expected before November 2026, GENIUS Act effective January 18, 2027. Repricing here will not be reflexive. It will be bureaucratic, gradual, almost invisible.
And in the meantime the quiet continues. The quiet is where the actual information is.