The Clarity Act and the Cost of Ambiguity in Crypto Regulation

There is something counterintuitive about watching an industry spend the better part of a decade asking Congress to regulate it. That is roughly where the digital asset industry sits, and the argument underneath it has very little to do with price charts and quite a lot to do with how legal certainty functions as infrastructure.

The intuition that regulation and innovation trade against each other is durable and mostly wrong at the margins that matter. Some of the most heavily prescribed corners of American finance — federally insured lending, where the forms are mandatory and the closing requirements are published in advance — are also the corners where capital moves in volume, precisely because every party knows the rules before the negotiation starts. Most of my practice sits in one of those corners, and the lesson it teaches transfers cleanly: what kills a transaction is rarely a demanding rule. It is not knowing which rule applies, or to whom, or whether the answer will hold long enough to matter. That is the frame worth bringing to the market structure debate, because it is the actual complaint.

The Digital Asset Market Clarity Act cleared the House in July 2025 by a 294–134 margin and advanced out of the Senate Banking Committee in May 2026 on a 15–9 vote, landing on the Senate Legislative Calendar as Calendar No. 423 on June 1.¹ Senate Republicans released consolidated text on July 22 merging the Banking and Agriculture Committee versions, which promptly reopened disputes over ethics, decentralized finance, and enforcement.² As of this writing the bill has no cloture motion, no scheduled floor vote, and no date on the calendar, and leadership has signaled that other business will occupy the chamber's remaining days before the August recess.³ The most advanced crypto market structure bill in American history is, at the moment, stuck.

Strip out the rhetoric and it does three fairly ordinary things. It draws a jurisdictional line, it gives spot exchanges a federal regulator, and it puts both in a statute rather than in guidance.

On the first: whether a given token is a security is currently answered after the fact, in litigation, under a test articulated in 1946.⁴ The bill sorts digital assets between SEC and CFTC oversight according to how centralized the underlying network is — a token launched by an identifiable team holding control and non-public information is treated as a security, with disclosures and lockups; a network that has crossed a defined decentralization threshold moves to commodity treatment. A tokenized stock remains a security either way, and the SEC has been saying as much on its own, most recently in reiterating that moving an activity onchain does not move it outside the securities laws.⁵ None of that is a novel theory. It approximates where the courts and both agencies have already landed. The bill's contribution is that a market participant could learn the answer without first being sued.

On the second: a domestic crypto exchange today holds state money transmitter licenses and carries enforcement exposure, but has no federal prudential supervisor in the sense the NYSE has one. Registration, custody segregation, audit requirements, and surveillance obligations are the ordinary apparatus that made the 2022 offshore exchange failures possible by their absence.

On the third: the agencies have moved substantially through guidance and no-action relief, but guidance survives only until the next administration withdraws it — and after the Supreme Court's June 29 decision in Trump v. Slaughter, which overruled Humphrey's Executor and held that for-cause removal protections for commissioners violate the separation of powers, the independence of any agency-level framework is a good deal less certain than it looked a year ago.⁶ Even the SEC's current chair, who has been active here by rule and by relief, has said the statute is what would future-proof the market.⁷ For an institution weighing whether to spend four years rebuilding settlement infrastructure, that distinction is the entire question. The stablecoin statute enacted last year is the proof of concept: once issuers had a one-to-one reserve requirement, monthly public disclosure, and Bank Secrecy Act obligations set out in law, adoption moved quickly.⁸ Not because the rules were permissive — because they were knowable.

The objections are not frivolous. The July draft carries restrictions on senior officials' digital asset holdings, which is unusual for an industry regulation bill; that is a legitimate structural objection, and also not an answer to the underlying concern. On illicit finance, the bill extends existing anti-money laundering and sanctions obligations to custodial intermediaries, and the harder question is how the perimeter treats non-custodial software, and whether a boundary drawn clearly invites the arbitrage it defines.

The yield fight is the one worth slowing down on, because it is the one that reaches ordinary lending. The GENIUS Act already bars permitted stablecoin issuers from paying interest or yield to holders, but it never defined "holder," and it does not plainly reach an exchange or an affiliate that pays rewards on balances parked with it.⁹ That gap is what the banking trades call the yield loophole, and closing it has been their central ask on Section 404. Their theory is straightforward: deposits fund loans, a stablecoin balance paying something near the federal funds rate is a savings product wearing a payment instrument's clothing, and money that leaves an insured deposit for a rewards-bearing exchange balance stops funding anybody's mortgage. The numbers they have marshaled are large. A Treasury advisory council identified roughly $6.6 trillion in transactional deposits as exposed, against a stablecoin market near$281 billion as of March.¹⁰ The ABA projects the market could reach $2 trillion if yield is permitted, largely at deposits' expense; the ICBA models a$1.3 trillion deposit decline translating into roughly $850 billion less community bank lending capacity; the joint trades have argued consumer, small business, and farm lending could contract by a fifth or more.¹¹ Citigroup, working the same question from outside the industry, projects a considerably narrower displacement.¹² The White House's own economists concluded in April that a yield prohibition would do little to protect bank lending while costing consumers a competitive return.¹³

Anyone who lends at the community level should take the deposit question seriously rather than wave at it. But the case has a soft center. Bank deposits may pay interest and largely do not; money market funds have competed for the same balances for four decades without anyone proposing to ban the yield; and an argument that a competing product must be prohibited because customers would prefer it is an argument about an incumbent's funding model rather than about financial stability. The honest framing is that this is a fight over who earns the float, dressed as a fight over credit availability — which does not make the credit concern fake, only secondary.

The proposal to impose broad downstream liability on developers for how their published code is later used sits in a different category altogether. Applied consistently, it ends open-source distribution — in digital assets, in AI, and in everything adjacent — because no individual contributor can price unlimited exposure for the conduct of strangers. Knowingly assisting a crime is already a crime. Publishing general-purpose software that someone later misuses is not, and the fight over export controls on strong cryptography is the cautionary version: source code treated as a munition, years of litigation and lobbying, and relief arriving only after foreign competitors had taken the market American firms were barred from serving.¹⁴

The alternative to a federal framework is not the absence of one. It is fifty state regimes, agency positions that turn over with each administration, and a competitive dynamic in which the compliant domestic operator carries costs its offshore counterpart does not. That is the worst available configuration, because the good actors bear the friction and the bad actors take the volume.

Ambiguity is a tax. It is paid in legal fees, in delayed launches, in transactions that die because nobody will opine on them, and eventually in activity that relocates somewhere with an answer. Whether this bill is the right answer is genuinely arguable. That the status quo is not is considerably harder to defend.