The Fed’s Stablecoin M1/M2 Note Is Not About Legality. It’s About Counting the Same Dollar Twice.

On September 4, a Federal Reserve staff paper appeared with none of the theatrics this industry usually demands. There was no token launch, no GitHub commit, no testnet. The note asked an apparently simple question: if dollar-pegged stablecoins become widely used, how should the Federal Reserve count them inside its official M1 and M2 money supply statistics? For most crypto natives, that is about as exciting as a footnote in a bank examiner’s manual. It should not be. The most consequential code being written this cycle is not a consensus engine or a layer-2 bridge. It is the statistical framework that decides whether a dollar token is money or merely a crypto product that looks like money. And the Fed’s own analysis exposes a deeper problem: the same dollar may already appear twice. I have watched this market long enough to distrust clean narratives. In 2017, projects with no smart contract and no prototype raised nine-figure ICOs. In 2020, liquidity pools that could not survive a governance vote were called decentralized banks. In 2022, Terra’s algorithmic miracle turned out to be a system where the stabilization mechanism was also the failure vector. The current stablecoin moment feels different because the balance sheets are real, the attestations are monthly, and the issuers are politically connected. Yet the Fed note is a reminder that regulation does not end with licensing. Regulators will eventually have to build the accounting infrastructure to see these tokens. That is where the real architecture gets exposed. Let me begin with the basic context, because much of the commentary around this story promptly got lost in monetary jargon. M1 is the narrow measure of money in the United States. It includes currency in circulation and highly liquid deposits that can be used for payments almost immediately. M2 is the broader measure. It includes everything in M1 plus savings deposits, small time deposits, and retail money market fund shares. These measures are not trivial statistics. They shape interest rate policy, liquidity forecasts, and the Federal Reserve’s understanding of how fast money is moving through the economy. If a new liability enters M1 or M2, the Federal Reserve is making a judgment about whether that liability is a credible substitute for a dollar in a checking account or a savings deposit. The paper under discussion is a staff note, not a policy decision. That distinction matters. The Federal Reserve system is filled with research economists who publish exploratory work that does not necessarily reflect Board policy. The note does not say that stablecoins will be added to M1 next quarter. It explores the statistical methods that would need to be developed if they were to be added. But the timing is powerful. The United States is now operating under the GENIUS Act framework, which requires stablecoin issuers to maintain one-to-one reserves and publish monthly reports. The act left open a key question: who decides whether a stablecoin is money for statistical purposes? The answer implied by the Fed note is that the Fed itself will make that decision. That is not a purely technical matter. It is a grant of structural power. The first insight that most market commentary missed is that this is not about scalability, throughput, or transaction fees. There is no TPS metric in the Fed note. There is no comparison of validator sets. There is no audit of smart-contract vulnerabilities. Instead, the relevant variables are reserve asset classification, holder jurisdiction, and economic function. That should be the first red flag for anyone who still believes the stablecoin fight will be won by the fastest chain. It will be won by the cleanest balance sheet. A stablecoin is a promise wearing an accounting costume. If the accounting cannot be reconciled with the Federal Reserve’s existing definitions, no amount of technical elegance will make it legal money. Consider the accounting problem at the heart of the Fed note. When a user converts one hundred dollars into a stablecoin, that one hundred dollars does not vanish from the banking system. It moves from the user’s bank account into an issuer-controlled account, where it is then reinvested in reserve assets. Those reserve assets are frequently bank deposits, U.S. Treasury bills, and money market instruments. Some of those assets are already captured, in one form or another, by existing monetary aggregates. A bank deposit held by the issuer may already sit inside a measure like M1 if the deposit is held by a domestic nonbank entity. Money market fund shares, depending on the holder and the fund, can already appear in M2. Treasury bills sit outside M1 and M2, but they are still extraordinarily liquid claims on the federal government. The problem is double counting. If the Federal Reserve adds stablecoin liabilities to M1 without simultaneously removing the reserve assets that already sit inside M1, the same underlying dollar is counted twice. It appears once as the issuer’s bank deposit or reserve asset, and again as the user’s stablecoin token. The aggregate money supply then increases by an amount that reflects packaging rather than actual new purchasing power. This is not a double-spend in the blockchain sense. No one is cheating. It is a methodological failure that would make the official money supply less honest. The Fed note correctly identifies this risk. The question is whether the market understands how difficult the solution will be. The fix is not simple. One approach would exclude any reserve assets that are already inside M1 or M2 from the issuer’s reserve calculation whenever the corresponding stablecoins are added to those same aggregates. But that would punish stablecoin issuers for holding bank deposits and favor those holding assets outside the aggregate, such as longer-dated Treasury securities. Another approach would require issuers to report their reserve composition asset by asset, day by day, so that the Federal Reserve can adjust the broad aggregates by the precise amount of overlap. That reporting burden already exists in rough form under the GENIUS Act, but monthly reports are not enough for monetary statistics that are often compiled weekly or even daily. The deeper problem is that no public blockchain was designed to answer the question: who owns that token, and where do they live? A transfer event from address A to address B tells you almost nothing about the nationality or residency of the economic owner. The blockchain event log is useful for proving that a token moved. It is not useful for determining whether that token belongs to a U.S. person or a person in Singapore, London, or Tokyo. That distinction is essential for constructing a U.S. money supply figure. The Federal Reserve does not count every dollar in the world. It counts dollars held by the U.S. public in specific types of accounts. A stablecoin circulating across the globe has no natural mechanism for assigning holders to jurisdictions. This geographic separation issue is a far larger barrier than most people realize. I spent time inside the central bank digital currency research world, building prototype systems that tried to reconcile monetary privacy with monetary reporting. One of the hardest problems was proving that a user belonged to a specific jurisdiction without exposing their entire transaction history. We could process enormous transaction volumes, simulate stress tests, and keep the system alive under significant pressure. But the accounting identity that regulators actually needed was not throughput. It was proof of location. The Federal Reserve needs to know that a token is held by a resident of the United States, and not merely that it exists on a chain used heavily by U.S. market participants. No zero-knowledge proof, by itself, answers that question. The cryptographic proof must be combined with a legal identity layer and a reporting standard. That is a bureaucratic architecture, not merely a cryptographic one. The Fed note also raises a functional distinction that the market has not priced carefully: the difference between money used for transactions and money used as a store of value. M1 instruments are generally transferable on demand. They can be used to settle payments quickly. M2 instruments, by contrast, include claims that are more savings-like. A stablecoin used primarily as collateral inside DeFi lending protocols may be closer to a savings-like instrument than to a checking account balance. A stablecoin used to buy coffee or to settle a cross-border invoice may function more like M1. The token alone cannot tell the statistician which use case dominates. The electronic wrappers around the token, the counterparties, and the settlement history must be examined. This is where my forensic instincts kick in. I have spent years looking at token flows and asking whether a protocol’s metrics reflect real economic use or self-referential activity. DeFi lending creates enormous volumes of stablecoin transfers that are not, in any meaningful sense, payments for goods and services. A user deposits a stablecoin into a lending pool, borrows against it, and moves the borrowed stablecoin into another pool. That activity produces blockchain events, but it does not necessarily indicate that the stablecoin is acting as M1-style money. It may simply be a collateral token moving between counterparties that are both part of the same financial plumbing. If the Federal Reserve were to count every stablecoin transfer as evidence of transactional money, it would seriously misunderstand the economy. The note implicitly understands this by emphasizing economic use. That phrase sounds like a policy term, but in practice it forces the Federal Reserve to look through the token to the underlying real-world transaction. That is extraordinarily difficult. The data needed to make that determination is not public. It lives inside exchanges, payment processors, stablecoin issuers, and banks. The blockchain shows the transfer of the token. It does not show the invoice, the bill of lading, or the payroll record that motivated the transfer. Someone will need to build a reporting chain that links off-chain economic context to on-chain events. That is not a minor engineering project. It is the beginning of a new category of financial surveillance and reporting infrastructure. Let me state the core insight plainly: the stablecoin industry is about to discover that statistical classification is a form of regulation. The GENIUS Act tells issuers what reserves they must hold. The Federal Reserve’s statistical framework will decide whether those reserves are compatible with money supply definitions. An issuer can hold perfect reserves, publish monthly attestations, and still find its token excluded from M1 because the reserve assets overlap with existing monetary aggregates. The very transparency that makes USDC attractive as a regulated product becomes a constraint. Every stablecoin issuer wants regulatory legitimacy, but legitimacy on the monetary side means being forced into a narrow, bank-like box. The market currently reads the Fed note as a positive sign. I understand why. A path toward inclusion in official monetary aggregates would make stablecoins look less like shadow assets and more like settlement infrastructure. Institutional investors could justify holding them in ways that current regulation does not permit. Payment companies could integrate them with fewer legal doubts. The demand for stablecoin exposure would almost certainly rise. That optimistic reading is not wrong, but it is incomplete. Inclusion in M1 or M2 is not a trophy. It is a leash. Once the Federal Reserve defines the statistical conditions under which a stablecoin counts as money, it will also define the conditions under which that stablecoin stops counting as money. The issuance process will become a regulatory event, not merely a software action. The contrarian position is that the most powerful stablecoin issuers may not actually want to be included in M1 or M2. The benefit of stablecoin legitimacy today comes from being treated as a money-like product without being fully integrated into monetary policy measurement. Inclusion would expose issuers to new scrutiny around reserve composition, holder jurisdiction, and user behavior. The Federal Reserve would have an incentive to demand highly granular data on every wallet holding more than a trivial balance. That data would need to be standardized across issuers, verified by auditors, and reconciled with bank records. For an asset engineered to flow around borders at the speed of the internet, that is a massive operational burden. There is a possible decoupling thesis buried in all of this. If stablecoins are eventually treated as official money, their value proposition changes. They stop being a bet on crypto adoption and become a bet on regulated digital payment infrastructure. The price of a stablecoin is, by definition, close to one dollar. The real variable is not the token price. It is the spread between the yield on the reserve portfolio and the operational cost of regulatory compliance. That spread will be determined more by Federal Reserve policy and Treasury market conditions than by crypto market sentiment. Stablecoin issuers will become, in effect, regulated money market intermediaries with a blockchain distribution layer. The successful ones will behave less like protocols and more like banks. The market that still values them on user growth alone may be measuring the wrong thing. Let me go back to the lessons I learned in 2022, when Terra’s UST collapsed. At that time, I was not interested in assigning blame to one founder or one token. I was interested in the missing regulatory framework that allowed a supposedly stable asset to unravel without a single honest public report. The obvious lesson was that reserves matter. The less obvious lesson was that reporting infrastructure matters just as much. Terra did not fail only because the algorithm was weak. It failed because no one could see the precise composition of the backing until it was already gone. If the Federal Reserve begins counting stablecoins in official aggregates, it will impose a reporting burden that makes Terra’s collapse impossible for a well-regulated issuer. But it will also impose that burden on every issuer, including the ones that do not want it. The GENIUS Act may be the legal foundation, but the Fed note is the accounting blueprint. The act says that reserves must be safe and liquid. The Federal Reserve must decide which reserve assets can be mapped back to M1 and M2 without double counting. If the reserve asset is an insured bank deposit, the overlap with existing money supply measures is high. If the reserve asset is a Treasury bill, the overlap is lower, but the issuer then depends on the Treasury market and must manage duration risk. If the reserve asset is a money market fund, the classification depends on whether the fund is retail or institutional, and whether its shares are redeemable daily. The reserve composition tables that issuers publish today are marketing documents compared with the granularity that monetary statisticians will eventually require. There is also a compliance problem that no one has solved: issuer-level reserves do not reveal wallet-level jurisdiction. A stablecoin issuer can know the total face value of its outstanding tokens. It does not necessarily know whether a particular token holder is a U.S. resident, a foreign exchange trader, a Colombian merchant, or a Nigerian software developer. The issuer can implement geographic blockers at the exchange level. It cannot police self-hosted wallets. If the Federal Reserve only counts stablecoins held by U.S. residents, then the entire stock of stablecoins held in self-hosted wallets outside the United States is a statistical unknown. The current industry standard, which relies on blockchain analytics, is not sufficient. Analytics firms can segment wallets by behavioral patterns, but they cannot determine legal residency with the certainty required for an official monetary statistic. This is precisely where the industry’s obsession with Layer 2 fragmentation becomes relevant. I have made the same complaint about Layer 2s for years: dozens of new chains are not scaling liquidity, they are slicing already-scarce liquidity into fragments. The same problem applies to monetary reporting. Stablecoin balances are now spread across Ethereum, Trons, Solana, Base, Arbitrum, Optimism, and a hundred other environments. Every chain has its own event log. Every bridge introduces another layer of custody and another potential gap in reporting. If the Federal Reserve wants to count stablecoins as money, it will need to aggregate across all of these environments, including the ones that are not yet in existence. The statistical challenge grows with every new chain. In that sense, the industry’s refusal to consolidate is making its own regulation more difficult. There is another layer to this that deserves attention. The Fed note is not just about what statisticians count. It is also about what they choose not to count. If stablecoins are excluded from M1 and M2, they remain part of the shadow financial system. They will still be used for payments, lending, and settlement, but they will not appear in the official narrative about the money supply. That creates a strange double life. The assets will be regulated enough to satisfy consumer protection laws, but invisible enough to avoid monetary policy scrutiny. That is a comfortable position for the largest issuers because it gives them regulatory cover without the full burden of bank-like reporting. I suspect some issuers privately prefer this outcome. The bullish public narrative around M1 and M2 inclusion may obscure an institutional preference to remain statistically invisible. What would change if stablecoins were actually included in M1? The immediate effect would be positive for adoption. Banks would have clearer guidance on how to treat stablecoin balances. Institutional funds would have a basis for classification. The label money would remove much of the uncertainty that currently suppresses enterprise use. But the longer-term effect would be restrictive. The Federal Reserve could respond to stablecoin growth by tightening reserve requirements, demanding location data, or imposing reporting thresholds based on wallet size. The infrastructure that makes stablecoins convenient and composable on-chain would be wrapped in off-chain compliance layers. The very features that make DeFi attractive, permissionlessness, interoperability, algorithmic composability, are the features that make monetary classification difficult. I am not saying that this is a bearish outcome. I am saying that the market is asking the wrong question. The relevant question is not whether the Fed will bless stablecoins. The relevant question is whether stablecoins can survive the statistical transparency that a blessing will require. A stablecoin issuer might have a perfect reserve portfolio today. It might publish monthly attestations with impeccable accounting. But the Federal Reserve’s M1 and M2 definitions do not operate on monthly cycles. They operate at the level of the entire system, and they require identifying not only the value of the liability but also the location and behavior of the holder. That is not something a blockchain can provide by itself. It requires a parallel reporting stack. Let me add one more layer that is missing from almost every public reaction: the classification of a stablecoin as money versus as a security will likely affect the Howey analysis. The Federal Reserve’s statistical treatment is not the same as the SEC’s securities classification, but the two are not unrelated. If a stablecoin is treated by the Fed as a monetary instrument with functional economic use, the argument that it is merely an investment contract becomes weaker. If the stablecoin is excluded from monetary aggregates because it functions more like a savings vehicle than a medium of exchange, the securities argument becomes stronger. Stablecoin issuers should not assume that their legal future will be decided exclusively by Congress or the SEC. It may be decided by the obscure statistical definitions that the Federal Reserve adopts over the next few years. This is why I find the Fed note so compelling. It is not a forward-looking policy essay full of generic praise for innovation. It is a warning that stablecoin entry into official monetary statistics will require new data sets, new reporting standards, and new forms of cooperation between private issuers and the state. The market should not wait for the Federal Reserve to decide the question in a memo. It should look at the memo as the first concrete signal of the reporting architecture that is coming. Every issuer should be asking which of its reserve assets are already counted in M1 or M2. Every issuer should be asking whether its blockchain network provides the information needed to prove where its token holders reside. Every issuer should be asking whether its monthly attestation is fast enough for the Federal Reserve’s statistical clock. The great irony is that the crypto industry is finally getting what it always wanted. Stablecoins are no longer being dismissed as speculative toys. They are being taken seriously enough to be counted in the Federal Reserve’s official ledgers. But counting is not an act of celebration. It is an act of control. The moment the Federal Reserve determines how to count stablecoins, it will also determine what kinds of stablecoins are allowed to exist. That, rather than TPS numbers or Total Value Locked, will become the real liquidity constraint of the next cycle. 2017’s dream is today’s regulation. Back then, the dream was that permissionless money could bypass the official financial system. Today, the challenge is not bypassing the system. It is convincing the system’s statisticians that a token can fit into the same definitions that were built for checking accounts and savings deposits. The winners will not be the founders who give the best keynote speeches. They will be the teams that can deliver clean, standardized, jurisdiction-aware, reserve-transparent data to the Federal Reserve on demand. That is a harder engineering problem than any bridge or rollup I have seen. It is also the problem that matters. The Federal Reserve has opened the door. What lies behind it is not a simple corridor to legitimacy. It is a forensic accounting operation. The next bull case for stablecoins will not be written in a tokenomics deck. It will be written in a regulatory reporting standard. The teams that understand this will position themselves inside the new architecture before the market catches up. The teams that do not will continue to advertise their reserves while avoiding the real question: if the Fed counted every dollar behind the token, would the number stay honest?