The Fed Counted Stablecoins for the First Time. The Ledger Doesn't Add Up.
Hook
Nothing happened to the peg. That's the story.
Over the past seven days, no major stablecoin drifted more than a few basis points from a dollar. No depegs. No emergency governance votes. No white-knuckle weekend on Crypto Twitter. And yet the single most consequential document to cross my desk this quarter was a piece of Federal Reserve staff research that, on its surface, does absolutely nothing — it explicitly refuses to change how the Fed publishes money data.
Don't buy the chart. Buy the chaos. The chaos here is bureaucratic, and it's louder than any candlestick. Because for the first time, the Federal Reserve's own researchers have put stablecoins on the same page as M1 and M2 — the numbers that decide how much money exists in the United States. They didn't announce a rule. They did something quieter and more permanent.
They started counting.
Context
Let me set the table, because the source material is dry enough that most people read the headline and moved on. The document is a FEDS Note — a staff research note from the Board of Governors, authored by Kristen Payne and Mary-Frances Styczynski. To be precise about what that is: a FEDS Note is not a policy decision. It is a note. The Fed said so itself, more than once, and I want to hold that caveat for the entire length of this piece. If you walk away thinking the FOMC just blessed stablecoins, you misread it.
Here's why it still matters. The note is wrestling with H.6 — the Fed's weekly money stock report. H.6 is not a niche document. It is the second most downloaded dataset on FRED and one of the primary pulse indicators of the US economy. When H.6 speaks, macro models move. M1 currently sits at roughly $19.9 trillion. M2 at roughly $23.2 trillion. These are the yardsticks against which every dollar in the system is measured.
Now drop a couple hundred billion dollars of stablecoins into that frame and ask a deceptively simple question: is a stablecoin money? And if it is, which kind?
The staff's answer is a framework, not a verdict. And the framework has a hole in it so obvious that the authors admit it in the text — which, for a central bank, is roughly the equivalent of a fire alarm going off in a library.
There's precedent worth knowing. In 2020, the Fed reclassified savings deposits, redrawing the boundary between M1 and M2. That was a functional decision — where does this balance actually live in the economy? The stablecoin note borrows that exact logic. Which is why the details matter far more than the headline, and why I spent my week reading the methodology instead of the press coverage.
There's also a broader backdrop. Stablecoins have been dragged off the margin of finance and into the middle of it. A major US bank regulator, the OCC, has committed to a rules timeline. A stablecoin law, the GENIUS Act, carries an explicit effective date of January 18, 2027. And according to the note's own framing, the market is no longer an experiment on the edge of the system — it has moved inside the walls. That single sentence is the whole reason I'm writing this at all.
Core
Let me get into the machinery, because this is where the narrative actually lives.
The note proposes a functional classification method: don't ask what a stablecoin is; ask what it does. A stablecoin used as a daily transaction medium belongs in M1 — the narrow, spendable layer. A stablecoin held as a store of value, or parked inside crypto trading desks, belongs in non-M1 M2. Same token, two different statistical homes, depending entirely on behavior. It's an elegant idea. It's also exactly the kind of elegant idea that dies the moment it touches real data.
Which brings us to the hard wall. The double-counting paradox. Here it is in one sentence: a stablecoin is typically backed by bank deposits, US Treasuries, and government money market fund shares — and every one of those backing assets is already counted inside M1 or M2. If you simply add stablecoin face value on top, you inflate the money supply without a single new dollar existing anywhere in the world.
This is not a rounding error. This is the whole ballgame. And the staff admits it in black and white.
Based on my audit experience mapping reserve structures for token funds, this is the trap that eats naive stablecoin models alive. The reason is structural, not fixable with clever code. A stablecoin is not new money — it is existing money wearing a new costume. The dollar backing it came from a bank account or a T-bill somewhere upstream. When it circulates, ownership changes; the quantity of base money does not. Every time a stablecoin changes hands, a ledger entry moves. No new dollar is born. So the Fed's most honest feature here is also its most damning: it is publicly disclosing the edge of its own capability.
And when a central bank tells you it can't track something, that is rarely a dead end. It's usually the first line of a budget request.
The specific thing they can't track is tokenized deposits — the on-chain form of a bank's own liability. Same species as a savings account, just wearing a distributed ledger. Without separate tracking of tokenized deposits, de-duplication becomes guesswork. I read that omission as a verifiable future signal: expect H.6 to sprout a new line item — something like "tokenized deposits" — before this framework becomes policy. That's the checkpoint I'd actually watch, and it's the one thing a serious allocator can put on a calendar.
There's a second soft spot the note waves at but doesn't resolve. If you classify by function — daily transaction medium versus store of value — then who decides which stablecoin is which? A token used for remittances in one country is a treasury reserve in another. Same asset, opposite buckets. That discretionary space is not a footnote. It's the seed of every future argument about whether a given stablecoin counts toward M1. And I've watched enough regulatory language turn into interpretive warfare to know where this goes.
Now zoom out and do the arithmetic everyone skipped. $19.9 trillion of M1. $23.2 trillion of M2. Even if every stablecoin on earth were folded in tomorrow, the aggregate distortion is statistically close to noise. Which tells you this was never about the size. It's about the signal. The Fed is minting a category, not measuring a number — and categories outlive numbers in this business.
Then there's the loop almost nobody drew. The GENIUS Act — Section 4(a)(11), to be exact — prohibits paying interest directly on stablecoins. No yield. No savings instrument. And because the non-M1 portion of M2 is partly defined by yield-bearing characteristics, a no-yield stablecoin drifts naturally toward M1, the transaction layer. Two independent institutions — a Congress that wrote the ban and a Fed that wrote the classification — arrive at the same destination from opposite directions. That unintended convergence is, to me, the most important structural fact in the entire document. Not a coordinated plan. A collision that happens to look like one.
There's a regulatory-stack reading here too. Issuance lives with the legislature. Operation lives with the OCC. Measurement lives with the Fed. Three agencies, one asset class, a shared timeline stretching into 2026 and 2027. That is not a scramble. That is a system being assembled in public, one layer at a time.
Contrarian
Here's the angle that will make me unpopular on both sides.
Everyone is reading this as a Fed story. It isn't. It's a GENIUS Act story wearing a Fed costume. The interest ban is the invisible hand here — it does more to determine where stablecoins land in the money aggregates than any framework the Fed can publish. Kill the yield, and you have already answered the classification question. The Fed isn't choosing; it's following legislative gravity. Read the law first, the Fed note second, and half the confusion evaporates.
The second blind spot: the market keeps treating this as a bullish catalyst. It isn't. There is no price link. Your stablecoin will not move a basis point because a staff note got published on a Thursday. Cycle after cycle, I've watched traders buy the wrong narrative at the exact moment the right one appeared. Code breaks. Stories don't. The story that broke this week is that stablecoins have been pulled off the margin of the financial system and booked into its ledger. That's a slow, institutional, multi-year arc — not a Tuesday trade.
The third thing almost nobody mentions: this framework quietly puts stablecoins and tokenized deposits in a cage match. Both want to be the on-chain form of money. Both want the regulatory seat. And banks are already circling, because if stablecoins take the transaction layer, the deposit franchise bleeds. That deposit outflow is the risk the Fed is actually worried about — it's why parallel research is already probing what happens when money walks out of bank accounts and into tokens. The Fed is building the scoreboard before the fight even starts.
Takeaway
Watch H.6, not the price feed. The confirmation event isn't a rally — it's a new line item appearing in a Fed table that nobody reads until it suddenly matters. If "tokenized deposits" shows up there, the framework graduated from concept to infrastructure, and the institutionalization of stablecoins stops being a narrative and becomes a schedule.
The question is no longer whether stablecoins are money. The Fed is asking a harder one: what kind of money — and who gets to decide when the numbers don't add up?