Markets Went Continuous. Central-Bank Money Didn’t. Stablecoins Settle the Difference.

Start with where the bill actually is, because most of the commentary is one procedural step behind.

The Digital Asset Market Clarity Act passed the House on July 17, 2025 by 294–134. The Senate Agriculture Committee approved its companion in January 2026. Senate Banking reported its version 15–9 on May 14, 2026, after Senators Thom Tillis and Angela Alsobrooks cut a deal on the stablecoin-yield question that had blocked the markup for months. Every substantive committee hurdle has been cleared.

It has sat on the floor ever since. On July 27 Majority Leader John Thune deprioritized it in favor of a Russia sanctions bill and a backlog of nominations; his office said leadership would have to "see where the votes are." Through the first week of August the bill was expected to get a procedural vote before recess. As of the morning of August 7, no cloture motion had been filed, which eliminated any possibility of a weekend vote. Senator Cynthia Lummis, the chief sponsor, was trying to hold the chamber in Washington past its scheduled Friday departure.

The blocker is not tokens. It is the government-ethics provision — restrictions on senior officials, including the President and his family, profiting from crypto ventures. The White House has been reviewing Tillis–Gallego language. Senators Murphy, Van Hollen and Merkley came off the bill when a merged draft thinned the ethics text. Senator Warren is blocking outright, on corruption, consumer protection and national security grounds. Senator Hawley opposes on community-bank grounds. Republicans hold 53 seats; cloture takes 60.

So the base case is now that the most consequential US crypto market-structure bill in history does not become law this year. The interesting question is what that actually costs — and the answer is much less than the sector's own rhetoric implies.

CLARITY answers a specific question: is a given token a security or a digital commodity, which agency supervises the intermediary that trades it, and what does registration look like. That question matters enormously for spot crypto venues, for DeFi, and for the several hundred tokens sitting in definitional limbo.

It matters remarkably little for the institutional stack that got built in 2026.

Stablecoins already have their own statute. The GENIUS Act was signed July 18, 2025. Whatever CLARITY does or doesn't do, payment stablecoins are governed by a separate federal regime with its own charters, its own reserve rules, and its own effective date.

Tokenized securities were never in definitional doubt. Commissioner Hester Peirce settled it on July 9, 2025 in a statement titled Enchanting, but Not Magical: "tokenized securities are still securities." A joint staff statement from Corporation Finance, Investment Management, and Trading and Markets on January 28, 2026 formalized the taxonomy — issuer-sponsored versus third-party-sponsored — and confirmed that putting a share on a ledger changes nothing about the securities laws that apply to it. There was no classification fight to win, so none was fought.

The plumbing was approved administratively, in sequence, over eight months:

Not one line in that table required CLARITY. Every one is a rule approval, a staff letter, a no-action position, a chartering decision, a pilot program, or an operational change.

Bernstein's Gautam Chhugani made the same call from the other direction on August 3: if CLARITY fails, the SEC and CFTC accelerate rather than retreat — interpretive releases on token taxonomy, DeFi and self-custody, plus the long-delayed innovation exemption. CFTC Chairman Michael Selig had already said the quiet part: absent Congress, regulators "end up writing all the rules."

What CLARITY would actually change

Four things, and they are narrower than the headlines suggest.

1. Spot-venue certainty. Bitnomial's leveraged spot listing works by treating every spot contract as if it were a futures contract under CEA §2(c)(2)(D). That is a workaround, and it does not scale to a national spot market.
2. DeFi and self-custody. Genuinely unaddressed by any current rule. This is the largest real gap.
3. The stablecoin yield loophole. Section 404 is where the bank lobby and the exchanges are actually fighting. More on this in §4.
4. Durability. This is the one that matters. Everything in the table above is reversible — by a new administration, a new Comptroller, a new Commission, or a federal judge. A statute is not. The sector built a 24/7 collateral system on administrative discretion, and the instrument that would have made it permanent is the one sitting on the Senate floor.

The standard stablecoin chart is market capitalization, and it looks boring right now. Total supply is $300.3B, against an all-time high of $321B set on April 21, 2026 and roughly $270B a year ago. Eleven percent growth. On that chart the sector is stalling.

Now put the flow next to it. Visa Onchain Analytics, which strips out bot traffic, exchange rebalancing and repeated smart-contract cycles, reported $1.79 trillion of adjusted stablecoin transaction volume in June 2026 — a record, up 63% month-over-month and 125% year-over-year from roughly $795B in June 2025.

Float grew 11%. Flow grew 125%. The ratio between them — turns per month — roughly doubled.

Circle's own second-quarter 2026 results say the same thing inside a single P&L. On-chain transaction volume of $14.8 trillion, up 151%. Reserve income of $668 million, up 5%. Total revenue and reserve income up 7%. The company's product got used half again as much as the year before and its revenue barely moved.

The sequential figures are sharper still, and they cut against the bullish reading of the float: USDC circulation of $73.3B at quarter end was up 19% year over year but down 4.8% from $77.0B the previous quarter, with management noting that the primary outflow came at the quarter-end period, and market share down 66 basis points year over year. Circle raised guidance on other revenue to $310–330 million from $150–170 million. That is a company telling you, in the only language a public filer has, that it is trying to earn money from something other than the float.

That is not a failure of execution. It is what happens when a business model priced on balances meets a product whose value is movement. Stablecoins are converting from a store — idle exchange balances that quietly print float income for whoever holds the reserve — into a rail, where the same dollar is used and released many times a month and the issuer earns on the average balance, not the traffic.

Follow the float income and the whole regulatory fight makes sense

If the float stops growing, the pool of reserve income stops growing with it — and the fight shifts from expanding the pool to dividing it. That is precisely what the last six months of stablecoin policy has been about.

With that number in view, three otherwise-unrelated stories turn out to be one story.

The yield war is existential, not technical. GENIUS §4(a)(11) bars a permitted issuer from paying interest or yield for merely holding the stablecoin. On its face it binds only the issuer — so the value routes around it, to affiliates and exchanges. The OCC tried to close that by rule: its proposed 12 CFR §15.10(c)(4)(i) would create a rebuttable presumption that an issuer is violating the prohibition if it has an arrangement with an affiliate or related third party to pay yield. Coinbase's USDC rewards program would trigger it. The rule is proposed, not final. Meanwhile CLARITY §404 would extend the prohibition to service providers and their affiliates — but permits activity-based rewards that may factor in "balance, duration, tenure, or any combination of the foregoing." The banking lobby's joint statement of May 4, 2026 called that clause the surviving loophole, arguing that "overtly incentivizing the idle holding of payment stablecoins for extended periods of time … would negate the goals of the upfront prohibition." Both sides are correct about the stakes. This is a fight over who books the pool in Figure 2.

Open USD is an attack on float capture, not on Tether. Announced June 30, 2026 with 140-plus launch partners — Visa, Mastercard, American Express, Stripe, Coinbase, BlackRock, BNY, Standard Chartered, Google, Shopify — OUSD's distinguishing feature is that it distributes reserve income to participating businesses after a small management fee, under shared governance rather than a single issuer. Read against Figure 2, that is not a product feature. It is a proposal to redistribute the entire pool away from the incumbent issuers, assembled by the distributors who currently hand it to them. It is expected to launch later in 2026 and is not yet live.

Ethena is what happens when yield is the product. USDe peaked around $14B and now stands at $3.9B, having fallen below $6B by April 2026. The mechanism paid yield out of perpetual-futures funding rates; when funding compressed, the yield went and the supply went with it. A synthetic dollar whose demand is entirely a function of a carry trade is a carry trade, not a dollar. It is also the clearest evidence that stablecoin float is not sticky when the reason for holding it changes.

Given where CLARITY sits, the practical question for anyone building or allocating is which regulatory pathway a dollar token can actually travel in the next eighteen months. There are three under GENIUS, plus a fourth that is not a stablecoin at all.

Path A — the statutory backstop. This is the base case.

GENIUS §20 sets the effective date as the earlier of eighteen months after enactment — January 18, 2027 — or 120 days after the primary federal payment stablecoin regulators issue final rules. The second branch is now dead. No agency has issued a single final rule. All six missed the July 18, 2026 statutory deadline, in several cases because their own comment windows were still open past it.

The mechanic that matters: only rules from the OCC, Federal Reserve, FDIC and NCUA start the 120-day clock. Treasury's and FinCEN's rulemakings — which are the most advanced — do not. For acceleration to have beaten the backstop, a primary-regulator final rule would have needed to land by roughly September 20, 2026. It will not. Plan around January 18, 2027.

The inventory as it stands: Treasury has an NPRM on state-regime equivalence (April 1) and a joint FinCEN/OFAC AML NPRM (April 8), both with comments closed. The OCC's main GENIUS proposal creating 12 CFR Part 15 was published March 2, comments closed May 1; a BSA proposal followed in June, plus proposed Forms PS-01 and PS-02 for weekly reserve and quarterly condition reporting. The FDIC has two proposals out. The NCUA has two. The Federal Reserve has no standalone prudential proposal at all — its only visible action is the seven-agency joint Customer Identification Program NPRM approved June 18. The Stablecoin Certification Review Committee has not issued its ownership-limitation rules, and the expedited state-review process that was statutorily due January 14, 2026 does not exist.

Then the real cliff: July 18, 2028, after which digital asset service providers may not offer or sell non-permitted payment stablecoins in the United States. That date, not the effective date, is what forces the offshore/onshore decision for every issuer serving US venues.

Path B — the charter route, which did not wait

The OCC simply moved. Conditional national trust charters went to Circle, Ripple, BitGo, Fidelity and Paxos on December 12, 2025, and to Coinbase on April 2, 2026. Circle received final approval on July 10, 2026 — the first stablecoin issuer all the way through the federal door, chartered to custody digital assets for itself and affiliates, positioning it to eventually manage the USDC reserve under federal supervision. Paxos went the other way on purpose, converting out of NYDFS supervision into the OCC to put PYUSD, PAXG, custody and reserve management under one regulator.

The most underrated structural fact in the sector: Anchorage Digital has become the federal issuance utility. It is the issuer of record for Tether's US product USAT, plus USDGO and USDtb, with Western Union's USDPT built on it, and it has disclosed a pipeline of up to twenty large firms looking to issue through it. A single OCC-chartered bank is quietly becoming the on-ramp through which most branded US dollar tokens reach the market.

This path has real opposition. The Bank Policy Institute retained outside counsel in March 2026 to weigh suing the OCC over these charters and, as of late July, had not filed. The ICBA has asked the OCC to rescind Coinbase's conditional charter and to halt Kraken's and Payoneer's applications. State regulators have called the structure a "Franken-charter." None of it has stopped a single approval.

Path C — the state route, which is broken

Issuers under $10 billion may elect state regulation if their state's regime is certified "substantially similar" to the federal one. State certifications were statutorily due July 18, 2026. Treasury's certification rule is still a proposal. The Certification Review Committee's expedited process does not exist. No state has been certified.

Seven senators — Lummis, Gillibrand, Ricketts, Cortez Masto, Cramer, Alsobrooks and Hagerty — wrote to Secretary Bessent on June 16 asking him to confirm that certification remains available on an ongoing basis rather than as a one-time window that has now closed, warning that a rigid reading would "likely foreclose future State participation." New York proposed its own aligned rule on June 9, with reserve segregation, two-business-day redemption, a 0.5% insured-deposit minimum for issuers above $25B capped at $500M, and a twelve-month transition — all of it keyed to the same January 18, 2027 federal date. Practically: below $10B you are in limbo, and above it you were always going federal anyway.

Path D — not a stablecoin: the banks build their own

The strongest competitive response is not another stablecoin. JPMorgan's JPMD deposit token has been generally available on Base since November 2025, and the Kinexys platform behind it processes on the order of $5–7 billion a day with $3–4 trillion cumulative. Citi Token Services runs in the US, UK, Singapore and Hong Kong. On June 5, 2026, JPMorgan, Bank of America, Citigroup and Wells Fargo announced a shared tokenized deposit network operated by The Clearing House, which they collectively own, targeted for the first half of 2027. Early Warning Services — owned by seven of the largest US banks and operator of Zelle, which moved $1.2 trillion in 2025 — announced ZelleUSD on June 11.

A deposit token is a claim on an insured bank, transferable 24/7 inside the regulated perimeter. It has no reserve-composition question, no §4(a)(11) yield problem, and no GENIUS effective date. If it ships on time it is the most serious threat to the institutional use case for public stablecoins — and I treat it as the leading counter-argument to this entire piece in §7.

Here is the chain that explains institutional stablecoin demand better than any narrative about crypto adoption.

First, trading went continuous — in mainstream products, not just crypto. CME launched 24/7 cryptocurrency futures and options on May 29, 2026, trading 7,200 contracts and roughly $50 million notional in the first weekend. On July 27, 2026 it took gold to 24/7: about 15,000 contracts and $60 million notional in the first weekend, on a contract averaging 87,000 lots a day. Kalshi's BTCPERP — the first true perpetual future ever approved by the CFTC, under a Reg 40.3 order rather than self-certification — went live June 3 and did $5.5 billion of notional in two weeks. Kraken listed US perps through Bitnomial on June 15. Cboe has run Continuous Futures since December 2025. In equities, 24X National Exchange is live, the Texas Stock Exchange began trading July 10, 2026, the SEC approved Nasdaq's 23/5 structure on April 10, and NYSE Arca's parallel rule is effective with an industry-projected December 6, 2026 start.

Second, the products themselves generate obligations on a continuous clock. A perpetual future has no expiry and settles a funding payment every eight hours. A margin call on a weekend crypto or gold position is not an edge case; it is the design. Positions that used to be revalued at a daily settlement are now revalued continuously in instruments with no natural pause.

Third, clearing caught up. Settlement did not. NSCC went live with 24x5 clearing on June 30, 2026, extending the CCP guarantee into overnight sessions. That closed the clearing gap. It did nothing to the settlement gap, because settlement still runs T+1 and, more fundamentally, because the Fedwire Funds Service and the National Settlement Service operate Monday through Friday only — Fedwire from 21:00 ET on the preceding calendar day to 19:00 ET, closed on all Saturdays, all Sundays and every Federal Reserve holiday.

Put those together and you get a measurable hole.

Fifty-eight hours. Thirty-four and a half percent of every week. Fedwire is open 110 hours out of 168 — five sessions of 22 hours — and the longest single stretch without it runs fifty hours, from Friday evening to Sunday night. Through all of it, CME's gold and crypto books are open and Kalshi's perpetuals are accruing funding every eight hours.

Now switch the market toggle in Figure 3 to Nasdaq's 23/5 structure, and the gap collapses to five hours. Switch it to NSCC's 24x5 clearing and it is ten. That contrast is the whole point, and it is worth being precise about: extended-hours equities were designed around the Fed's calendar and fit inside it almost perfectly. The 58-hour hole is created specifically and only by the genuinely continuous products — 24/7 futures, and perpetuals with an eight-hour funding clock. This is not a general infrastructure complaint. It is a targeted one, and it identifies exactly which instruments generate the demand.

The precise claim matters, so let me state it carefully. Commercial bank money moves during those hours — banks credit and debit their own customers' accounts continuously. What is unavailable is final interbank settlement in central bank reserves. An obligation between two institutions incurred at 3am on a Sunday cannot be extinguished with finality until Sunday night at the earliest, and if it arises Friday evening, not until Monday. Everything in between is credit exposure sitting on someone's balance sheet.

That is the demand driver. It is not ideological, it is not speculative, and it has nothing to do with anyone's view of crypto. It is a scheduling problem with a balance-sheet cost, and the instruments that solve it are the ones that transfer with finality on a ledger that does not observe Federal Reserve holidays.

And here is the part that changes the investment case

The obvious rebuttal is that this is temporary — the Fed is fixing it, so private settlement assets are a bridge to 2028 and then the problem goes away. That rebuttal is wrong, and the reason is in the Fed's own announcement.

On October 9, 2025 the Board announced its intent to expand Fedwire and NSS to six operating days — Sunday through Friday, including weekday holidays, targeted for 2028 or 2029. Daily hours are unchanged. Participation is voluntary. The Board described it as "an interim step" that provides foundational capability for potentially expanding to 24/7/365 later.

Run that through the grid. Adding a Sunday business day takes Fedwire from 110 open hours to 132. The gap falls from 58 hours to 36 — from 34.5% of the week to 21.4%. The longest continuous stretch without central-bank settlement falls from fifty hours to twenty-six. Saturday never opens at all.

This also dates the land-grab. Whatever occupies those hours between now and 2028–29 becomes the incumbent operational standard — the thing risk systems are wired to, the thing legal opinions are written about, the thing collateral schedules name. Displacing an incumbent settlement rail after it is embedded is far harder than winning the initial slot. That is why the calendar in §9 is dense with 2026 and 2027 launch dates and why so many of them are consortium plays rather than products.

The retail story about real-world assets — democratized access to private credit, fractional everything — is not where the institutional money went. Read the 2026 announcements in sequence and they are almost entirely about one thing: moving collateral faster, and on more days, than the existing plumbing allows.

The single most important transaction of the year happened on July 15, 2026, when DTCC processed its first live production tokenized trades with more than forty firms participating. Three of them matter:

• JPMorgan tokenized the Invesco QQQ ETF and posted it as margin at CME — the first time a central counterparty has accepted a tokenized asset for margin in live production. The timestamps are the story: tokenized at 9:00am, covering a real margin obligation at CME Clearing by 10:45am. A hundred and five minutes from custody to cleared margin.
• Société Générale posted tokenized US Treasuries as collateral with Citadel Securities.
• Citadel Securities and BNP Paribas completed a securities lending transaction on the same rails.

Assets were Russell 1000 constituents, major index ETFs and Treasuries, settled across DTCC's Hyperledger Besu chain and Canton. No volumes were disclosed — a point I return to in §7. The full service launches in October 2026, and the Collateral AppChain, announced with Chainlink on May 12, 2026, targets production in Q4 2026 for pricing, valuation, margining and settlement on a 24/7 near-real-time basis. More than fifty firms sit in the working group. DTCC and Finadium have put a number on the prize: intraday repo on a digital ledger could halve intraday funding costs at large dealer banks.

Notice the design goal in that sentence. Not new assets. Not new investors. Twenty-four-seven near-real-time — which is a direct response to §5.

Everything else lines up the same way

BNY launched tokenized deposits on January 9, 2026 and said explicitly it was starting with collateral and margin workflows. It added USDC to its Digital Asset Custody platform on June 29, launched a global Digital Transfer Agency on July 29 servicing $8.6 trillion, and is already reserve custodian for USDC, Ripple's RLUSD and Société Générale's USD CoinVertible, on top of the tokenized money-market platform it has run with Goldman Sachs since July 2025.

Binance began accepting Franklin Templeton's BENJI as off-exchange collateral on February 11, 2026, held with Ceffu and mirrored into the trading environment. BlackRock's BUIDL has been collateral at Crypto.com and Deribit since June 2025. WisdomTree got SEC exemptive relief on February 24, 2026 for 24/7 trading and instant settlement of a tokenized money market fund — a money fund whose shares change hands on a Sunday at a fixed dollar with dividends accruing by wallet-holding duration. Apollo's ACRED is levered two-to-three times on Morpho against RedStone NAV oracles.

And the regulator built the doorway. The CFTC's Digital Assets Pilot Program, launched December 8, 2025, permits FCMs to accept non-securities digital assets including payment stablecoins as customer margin, with weekly reporting by asset type; guidance issued alongside it covers tokenized real-world assets such as Treasuries — custody, valuation, haircuts, operational risk. Initially limited to bitcoin, ether and USDC, eligibility was extended on February 6, 2026 to GENIUS-compliant stablecoins issued by national trust banks, correcting a staff letter that had accidentally excluded them.

Set that against the pool it is aiming at. Non-cleared initial margin collected across the market was $1.6 trillion at the end of 2025; regulatory initial margin at major CCPs for cleared rates and credit was $423.5 billion. A Global Digital Finance industry survey published July 6, 2026 found 66% of firms plan to launch tokenized money market funds before the end of 2027 and 44% expect to accept them as collateral by the same date.

Total tokenized real-world assets stand at roughly $37.9 billion of distributed value, of which tokenized Treasuries and money funds are $16.2 billion — Circle's USYC at $3.0B, BlackRock's BUIDL at $2.7B, Ondo's USDY at $2.15B, Franklin's BENJI at $1.7B on one tracker and above $2.5B on Franklin's own reporting. Tokenized equities are $2.5 billion and growing fastest, up 18% in thirty days with $12.3 billion of monthly transfer volume.

Ondo, and two narratives that need retiring

Ondo is the cleanest single case study, and two things commonly said about it are now wrong. Ondo Chain is dead. On July 27, 2026 the company launched the Ondo Network — a TEE-based execution layer separating execution, verification and settlement — explicitly replacing the planned L1 on the reasoning that "execution speed, not settlement, was the limiting factor." And Ondo joined the DTCC consortium rather than competing with it.

The business is real: USDY at $2.15B, OUSG at $451M, tokenized equities of $864M across 406 assets for roughly 59–70% share of that market, more than $20 billion of cumulative trading volume, and the Oasis Pro acquisition completed October 7, 2025 bringing an SEC-registered broker-dealer, FINRA membership, an ATS and a transfer agent in-house. In July 2026 it rebranded Global Markets to Ondo Stocks, crossed $1B TVL and launched 24/7 trading on six tickers.

And the tell that ties it back to §1: Ondo Stocks is not available to US persons. Neither is the tokenized BlackRock IVV ETF it put on Ethereum on July 2, 2026. The registered-plumbing route into US investors runs through Oasis Pro's FINRA authorization, which is exactly the narrow, entity-by-entity path that exists because no general market-structure statute does. That is the shape of the world CLARITY was meant to change, and hasn't.

An argument this tidy deserves its strongest opposition stated properly rather than in a hedging paragraph at the end.

1. The banks may take the use case away entirely

This is the serious one. If The Clearing House network ships in the first half of 2027, institutions get 24/7 transferability inside the regulated banking perimeter, backed by insured deposits, with no reserve-composition question, no §4(a)(11) yield problem, no GENIUS effective date and no counterparty who might be a crypto company. JPMD already works. Kinexys already moves billions a day. If a treasurer can settle at 3am Sunday in a claim on JPMorgan rather than a claim on a reserve pool, most will.

The honest version of my thesis survives this only in a narrower form: public stablecoins keep cross-border, crypto-native and non-bank flows, and deposit tokens take institutional settlement between large banks. That is still a large business, but it is a different business from the one the $2 trillion projections describe. Watch the first-half-2027 date closely; a slip is the single most informative signal available.

2. The most obvious version of tokenized collateral was offered and refused

The Options Clearing Corporation filed a collateral rule change on February 27, 2026, approved April 10. It removed letters of credit and GSE debt, and added a wrong-way-risk haircut on self-custodied spot crypto ETPs. It added no tokenized assets and no stablecoins. More damning: OCC already permits money market fund shares as collateral, and that eligibility sits unused. If the operational case for a fund share as margin were compelling, the untokenized version would already be in use at the largest US equity-derivatives clearinghouse. It isn't. That is real evidence that the friction being solved is smaller than the marketing claims.

3. The numbers are still very small, and disclosure is thin

Tokenized Treasuries and money funds total $16.2 billion against $1.6 trillion of non-cleared initial margin — roughly one percent. DTCC's July 15 milestone disclosed no transaction volumes at all, which for a live-production announcement is conspicuous. Ondo's OUSG shrank 5% over thirty days. Eurex says tokenized collateral is "in production" without naming a single eligible asset. A great deal of what gets reported as adoption is a named first transaction, and a first transaction is not a business.

4. Flat float may mean saturation, not acceleration

My reading of Figure 1 is that float is flat because dollars are turning over faster. The alternative reading is that stablecoin demand has simply topped out near $300 billion, the April peak of $321 billion was the high, and rising volume reflects a smaller set of high-frequency users rather than a broadening base. Ethena's collapse from $14 billion to $3.9 billion is consistent with either story. I hold the velocity reading because the volume growth is corroborated inside Circle's income statement rather than only in on-chain aggregates — but it is one year of data on a metric with a short history, and I would not defend it past a second year of flat float.

Return to the durability point from §2. Every element of the 2026 build-out is a rule, an order, a staff letter, a no-action position or a pilot. Four facts about the current state of that foundation:

CME is suing the CFTC. Filed June 18, 2026 in the District of Columbia as 1:26-cv-02157, CME Group v. CFTC claims the Commission violated the Administrative Procedure Act in approving Kalshi's BTCPERP as a future without addressing whether perpetual contracts are swaps under Dodd-Frank. The relief sought is not narrow: vacatur of the order and the accompanying policy statement, plus a declaration that perpetual contracts are swaps. Terry Duffy's public framing is blunt — "when two parties exchange payments to each other, that is deemed a swap." If CME wins, the fastest-growing continuous-trading product in the US has to be re-plumbed through swap execution and clearing requirements, and the venue driving 24/7 demand loses its instrument. Note the irony: CME itself went 24/7 in crypto and gold rather than list perps, so it is simultaneously the largest beneficiary of continuous trading and the plaintiff against its most successful form.

The CFTC has one commissioner. Michael Selig was confirmed December 18, 2025 and sworn in December 22. Four of five seats are vacant, including both minority seats. The House Agriculture Committee urged the President in May 2026 to fill them before any crypto market-structure rollout. Every action described in this piece — the pilot program, the collateral eligibility extension, the Kalshi order, the no-action letter on offshore perps — currently rests on a single Senate-confirmed official, and is being challenged in court on exactly that kind of ground.

The SEC pulled its innovation exemption. Expected around May 18–22, 2026, it was shelved on May 22 after exchange officials objected to a provision permitting third-party tokens — digital representations of shares created without issuer consent — on the grounds that it would break dividend administration and shareholder vote counting. No new date has been set. Hester Peirce, who runs the Crypto Task Force and defended the exemption as covering "digital representations of the same underlying equity security," leaves the Commission in November 2026. Two SEC seats are also vacant.

Nothing under GENIUS is final. Not one rule, from any of six agencies. The OCC's affiliate-yield presumption — the provision that would actually determine whether exchange rewards programs survive — is a proposal with closed comments and no adoption date. Treasury's state-equivalence rule is a proposal. Foreign-issuer reciprocity has not been proposed at all.

This is the real cost of CLARITY's failure, and it is not the cost the sector talks about. Nobody is going to stop building. The DTCC service launches in October whatever the Senate does. What is missing is the thing that makes the build survive a change of administration, an adverse APA ruling, or a Comptroller with different views. The sector has spent a year getting what it wanted from agencies and calling it a win. It is a win with an expiry that nobody has priced.

Dates that are already fixed or announced. Anything described as a target is exactly that.

CLARITY is stuck, probably past this year, on an ethics provision rather than on anything to do with market structure. That matters less than it looks, because the institutional perimeter was built without it — by charter, no-action letter, exemptive order, pilot program and self-certification, across roughly eight months.

Stablecoins are being pulled into institutional use not by adoption narratives but by an arithmetic problem: US markets now trade fifty-eight hours a week with no central-bank settlement rail, and the Federal Reserve's own plan for 2028–29 leaves thirty-six. That makes a private 24/7 settlement asset permanent infrastructure rather than a stopgap, and it makes the next twenty-four months a land-grab for the operational default.

Tokenized real-world assets are the collateral layer of the same system, which is why the meaningful 2026 milestones are a tokenized ETF posted as margin at CME and a collateral chain targeting Q4, not a private-credit fund reaching retail. The float income that pays for all of it is flat while usage doubles, which is why the yield fight is the sector's central conflict and why the distributors have organized to take that income from the issuers.

And the whole structure rests on administrative discretion, a single-commissioner CFTC and a pending APA challenge aimed squarely at its newest load-bearing product. The sector got what it wanted without the statute. It has not yet noticed what the statute was for.