Coordinated Control

June 8, 2026

In June 2018, William Hinman delivered what would become the most consequential thirty minutes of digital-asset regulatory thinking of the past decade. The speech was carefully labeled personal views, not Commission policy, and was promptly absorbed into industry compliance memos as if it were a final rule. The core idea was simple. An investment contract under SEC v. W.J. Howey Co., 328 U.S. 293 (1946), turns on whether purchasers expect profits from the “efforts of others.” A token transacted on a network that has reached “sufficient decentralization” arguably fails that prong, because there is no longer a coherent “other” doing the work. Ether (ETH), Hinman said in the most-quoted passage, no longer represented a security transaction in its then-current state.

For seven years, “sufficient decentralization” sat at the center of every token launch and every secondary-market listing analysis. The SEC’s April 2019 Framework for “Investment Contract” Analysis of Digital Assets attempted to formalize the inquiry, but ran to dozens of considerations across multiple *Howey* prongs and was published as Staff guidance with no formal regulatory effect. *SEC v. Ripple Labs, Inc.*, 682 F. Supp. 3d 308 (S.D.N.Y. 2023), illustrated the cost of operating in this vacuum. Judge Torres distinguished institutional sales from programmatic exchange sales by reference to the economic reality of each transaction, but neither party could point to a statutory test that would have produced the same result with more certainty.

§104 of the engrossed Senate substitute is the legislative answer. It does not overrule Hinman. Hinman was not law. But §104 effectively buries the speech, replacing the SEC’s “sufficient decentralization” intuition with a five-factor statutory test, anchored by a single quantitative threshold, and operationalized through a certification procedure that produces final agency action reviewable under applicable law. This is the coordinated-control framework that Post #1 identifies as one of the four structural pillars of the bill’s architecture. CLARITY routes around *Howey* rather than overruling it. Decentralization stops being a vibe. It becomes a definition. Once a system is certified as not subject to coordinated control, the resale restrictions in §104(c) lift on a defined schedule that functions as a Rule 144 analog for tokens.

A word on procedural posture before the substance. §104(b)(1) directs the Commission to adopt rules, based on the §104(b)(2) criteria, to define when a distributed ledger system and its related ancillary asset are under coordinated control. The five criteria are not, themselves, the definition. They are the indicia Congress instructs the SEC to consider in rulemaking. The list is exhaustive, however (§104(b)(2) uses “the following criteria,” not “including”), and the §104(b)(3) safe harbors operate independently as bright-line outs. The SEC will fight on weighting and on attribution rules, but it cannot import factors outside the (A) through (E) list.

The Five Indicia

The §104(b)(2) criteria divide cleanly into two structural questions. Three of the five ask whether the protocol is technically open and functional. Two ask whether there is a control group still pulling levers. The five are not weighted on the face of the statute, but the architecture suggests, and the SEC’s rulemaking will likely confirm, that the technical-openness indicia are necessary but not sufficient. A system that ships open-source code under a permissive license and runs on a permissionless validator set can still be under coordinated control if a founding team controls 60% of the float.

§104(b)(2)(A) asks the extent to which the distributed ledger system is not (i) a publicly available protocol, (ii) a distributed ledger application whose source code is freely available via open-source code and recorded on a distributed ledger, or (iii) a Commission-determined analogue. The negative phrasing is awkward but the substance is familiar. A system fails this indicium to the extent the protocol or application source is closed, proprietary, or under license terms that meaningfully constrain forking. This is a low bar. Nearly every credible Layer 1 and Layer 2 in production today clears it. The edge questions are protocols with closed sequencers, MEV-extraction code that is not public, or proprietary fraud-proof systems. The Commission will need to address the extent to which non-public auxiliary infrastructure (block builders, relays, oracles) counts as part of the “distributed ledger system” for §104(b)(2)(A) purposes.

§104(b)(2)(B) is the permissionlessness indicium. It asks whether any person or group under common control has either (i) unilateral authority via operation of the distributed ledger system to restrict, censor, or prohibit use, including any applicable system-based user activity, or (ii) private permissions, hard-coded privileges, or similar capabilities granted by source code that provide preferential treatment compared to similarly situated persons. (i) covers admin keys with censorship power. (ii) covers protocol-level preferential treatment, which is the more interesting prong. A founding-team allocation that vests on a smart-contract schedule, with no special governance rights and no early redemption privilege, should not implicate (ii). A founder-only fee-distribution mechanism that takes a fixed percentage of protocol revenue indefinitely probably does, even if the rest of the protocol is otherwise permissionless. The Tornado Cash sanctions litigation surfaces a harder question on (B)(i): a protocol whose front-end voluntarily blocks sanctioned addresses but whose contracts do not presumably clears it, while a protocol whose contracts hard-code address blocklists arguably does not.

§104(b)(2)(D) is the autonomous-state indicium. The statute asks whether the system has not yet reached “an autonomous state” and whether a person or group under common control has unilateral authority to alter or change the functionality, operation, or rules of consensus or agreement. The term “autonomous state” is undefined and will be a major focus of SEC rulemaking. Compare the SEC and CFTC’s 2026 joint interpretive release (Release Nos. 33-11412, 34-105020), which characterized “functional” digital commodities by reference to whether the network’s native asset can be used “on the system in accordance with its programmatic utility.” §104(b)(2)(D) sits between functionality and immutability. A network can be functional but not autonomous, in the sense that core developers still hold upgrade keys allowing protocol-level changes. The SEC’s natural interpretive move will be to characterize autonomy as a function of governance distribution and upgrade-key custody, which folds (D) into the same inquiry as (B) and (C).

§104(b)(2)(E) is economic independence. The indicium asks whether the primary programmatic mechanisms that are intended to facilitate substantial value accrual to the ancillary asset are functional. This is a value-accrual readiness test. A token whose buy-and-burn or fee-distribution mechanism has not yet shipped fails it. A token whose value-accrual mechanisms are technically live but rely on continued promotional activity by the founding team probably also fails, though the line is fuzzier. (E) is the indicium most directly tied to the *Howey* “efforts of others” prong, and it will drive much of the SEC’s substantive analysis. A network can be open-source, permissionless, and well-distributed, but if its value accrues primarily because a centralized team is still building demand, it is not economically independent within the meaning of §104.

The Only Hard Number

§104(b)(2)(C) is the single quantitative threshold in the entire decentralization framework. It asks the extent to which a person or group under common control has beneficial ownership “of, in the aggregate, more than 49 percent of the total amount of outstanding units of the ancillary asset or voting power with respect to any governance system that relates to the distributed ledger system.” That “or” is doing significant work. A control group that holds 30% of token supply but controls 55% of governance voting power triggers the indicium. So does a group that holds 60% of supply but only 20% of voting power. The disjunctive structure forces issuers to manage decentralization on two axes simultaneously, which the Hinman framework never did with any precision.

Three structural questions follow from the 49% language. First, what does “person or group of persons under common control” mean as an attribution rule. Second, how does the indicium handle lockups, vesting, and time-restricted tokens. Third, how does “voting power” measure against the wide variety of governance structures in the wild.

The common-control attribution rule is the most consequential ambiguity. The statute does not define common control for §104 purposes, and the SEC’s rulemaking will need to import an attribution standard from somewhere. The Investment Company Act’s “control” definition (15 U.S.C. § 80a-2(a)(9), beneficial ownership of more than 25% of voting securities, with a rebuttable presumption) is the most natural analog. The Securities Exchange Act’s beneficial ownership rules (17 C.F.R. § 240.13d-3) are the next candidate. Section 318 of the Internal Revenue Code, with its constructive ownership and family attribution rules, is the most aggressive option. Practitioners should not assume that a majority-owned subsidiary of a foundation is treated as a separate person for §104(b)(2)(C) purposes, or that the SEC will reject family attribution out of hand for founder-controlled vehicles. *Sarcuni v. bZx DAO*, 664 F. Supp. 3d 1100 (S.D. Cal. 2023), analyzed DAO governance through a general-partnership lens, which suggests that joint enterprise principles may sweep token holders together as a single common-control group even without a contractual coordination mechanism. Where the SEC draws the line will determine whether well-distributed foundation-led launches can certify out of coordinated control within a reasonable post-launch window.

Lockup and vesting interplay is the second moving piece. The §104(b)(2)(C) test asks about “beneficial ownership… of total outstanding units.” Standard practice is to exclude unvested team and investor tokens from circulating supply but include them in total supply (the familiar fully-diluted-valuation versus market-cap distinction). “Outstanding units” is ambiguous between these two. The conservative reading is that fully diluted supply controls, in which case a team allocation of 30% counts toward the threshold from day one even if subject to a four-year vest with a one-year cliff. The aggressive reading is that only minted and unrestricted tokens count, in which case the threshold can be managed during the lockup period by burning, locking, or simply not minting reserved supply. The conservative reading is more consistent with the antifraud purposes of the statute and with §104(c), which itself treats unvested holdings as part of related-person exposure. The aggressive reading is more consistent with how the industry has measured decentralization for the last five years.

Governance-power measurement is the third moving piece, and the messiest. The statute says “voting power with respect to any governance system.” That language reaches DAO voting (one-token-one-vote and its quadratic and conviction variants), delegate-based governance (Optimism’s Citizens’ House and Token House, Arbitrum’s delegate system), validator-set voting, and multisig-as-governance (where a multisig executes off-chain governance decisions). It is not obvious that all four are measured the same way. A founding team that holds 40% of governance tokens but has delegated 35% to non-affiliated delegates might fall above or below 49% depending on whether delegation counts as a release of voting power. The Wyoming DUNA framework (Wyo. Stat. §§ 17-32-101 et seq.) and the Marshall Islands DAO LLC structure both address related questions of voting attribution among DAO members, but the §104 test will require its own answer.

The interaction with the §104(b)(3)(B) decentralized governance system safe harbor is also important. A DGS is, by definition, not a person or group under common control. §104(b)(3)(B)(i). The Commission cannot count DGS voting power toward §104(b)(2)(C) because the DGS is a separate legal person. But the participants in the DGS may themselves be under common control with each other or with the founding team. A foundation that funnels its governance influence through a nominally decentralized DAO does not automatically clear (C); the SEC will need to evaluate whether the DGS is operating with independence or as a pass-through for foundation control.

Multisig Councils Survive

The §104(b)(3) safe harbors are the operational counterpart to the indicia. Where the (b)(2) factors describe what coordinated control looks like, the (b)(3) safe harbors describe arrangements that are not, on their own, evidence of coordinated control. There are three. The general safe harbor in §104(b)(3)(A) directs the Commission to establish safe harbors more broadly, and §104(b)(3)(D) confirms that the enumerated safe harbors are not exclusive. The two enumerated safe harbors are the DGS carve-out in §104(b)(3)(B) and the emergency-measures carve-out in §104(b)(3)(C).

The emergency-measures safe harbor is the most thoughtful provision in the entire decentralization framework. Virtually every production DeFi protocol of significance has some form of multisig-based emergency response capability, and subjecting those councils to coordinated-control analysis under a strict reading of (B) would force them to be dismantled in ways that would harm users.

The safe harbor requires that the emergency measure be “pre-defined, temporary, [and] rules-based,” exercised by an “incident response or security council” exclusively in response to a “specific and documented cybersecurity incident or imminent threat,” pursuant to “publicly disclosed, on-chain authorization mechanisms,” “strictly limited in scope and duration solely to address that cybersecurity incident or imminent threat,” and exercised “without unilateral control by any single person.” §104(b)(3)(C). The procedural mechanism for invocation must be disclosed in publicly available written documentation reasonably available to the applicable Federal agency by a decentralized autonomous organization or similar legal entity sufficiently in advance of any exercise.

That language maps cleanly onto Compound’s Pause Guardian (a multisig that can pause specific contract functions in response to a threat, codified in the protocol and disclosed in governance documentation), Aave’s Guardian (similar structure, with explicit on-chain limits on what the Guardian can pause and for how long), and MakerDAO’s Emergency Shutdown mechanism (governance-triggered, with public documentation of the procedure). It is harder to fit the kind of governance-proposal-and-patch response Compound used in the 2021 reserve-distribution incident, which involved community proposals rather than a true incident-response council action. The safe harbor seems to require a council with pre-defined authority, not ad hoc community remediation.

The “publicly disclosed, on-chain authorization mechanisms” requirement deserves attention. It is not enough that a multisig exists with documented members. The authorization mechanism itself (which addresses are in the multisig, what they can do, what triggers their authority, and what limits apply) must be on-chain and publicly disclosed. This is a higher bar than the customary practice of disclosing multisig membership in governance forum posts while leaving the actual multisig contract opaque.

The DGS safe harbor in §104(b)(3)(B) sits alongside the emergency-measures carve-out and does separate work. It establishes that a DGS is not a person or group of persons under common control, and that a system is not precluded from being free of coordinated control “solely based on a functional, administrative, clerical, or ministerial action” of the DGS, including actions taken by a person acting on behalf of and at the direction of the DGS. The clerical-action language preserves operational flexibility for DGS implementations that delegate routine matters to identified service providers. A DUNA that retains a service provider to file annual reports does not become a person under common control with its members because the service provider acts at the DUNA’s direction.

Rule 144 for Tokens

§104(c) is the operational consequence of the coordinated-control framework. If a distributed ledger system is under coordinated control, related persons (broadly defined to include the ancillary asset originator, its subsidiaries, and entities under common control) face restrictions on resale of covered tokens. The restrictions function as a token-equivalent of Rule 144 of the Securities Act (17 C.F.R. § 230.144), and the parallels are deliberate.

Before certification of non-control, a related person may sell covered tokens acquired post-enactment only if (i) §4B(d) disclosures have been furnished, (ii) the holder has held the units for not less than twelve months, and (iii) the amount sold in any twelve-month period does not exceed a cap to be set by SEC rulemaking. §104(c)(1). After certification, the holding period drops to six months, and the volume cap is statutorily floored at no less than ten percent of total outstanding units per twelve-month period. §104(c)(2). For pre-enactment tokens, the holding periods are the same (twelve and six), but the §4B(d) disclosure requirement is waived for post-certification sales. §104(c)(3). Distributed ledger control persons (a narrower category of persons with unilateral authority over protocol functionality) face additional notice and disclosure requirements under §104(c)(4) even after certification.

The Rule 144 analogy is structural rather than mechanical. Rule 144 conditions secondary-market resales of restricted securities on current public information, a holding period (six months for reporting issuers, one year otherwise), volume limitations (the greater of 1% of outstanding or average weekly trading volume), and manner-of-sale restrictions for affiliates. §104(c) adopts the holding-period and volume-limitation concepts directly, conditions on §4B(d) disclosure rather than Exchange Act periodic reporting, and grants the SEC rulemaking authority to fill in the manner-of-sale and notice details. The most consequential difference is the volume cap. §104(c)(2)(C) sets a statutory floor of 10% of outstanding units per twelve months for post-certification sales, which is far more permissive by reference to absolute float than Rule 144(e). For most token launches with standard team allocations, the 10% floor will exceed the related person’s holdings, which means the statutory floor will not bind. The SEC will likely set the actual cap substantially lower in the rulemaking. Where it lands is one of the most consequential open questions in the entire rulemaking process.

Disgorgement is the enforcement mechanism. Under §104(e), profits realized by a related person from sales in violation of §104(c) inure to the holders of the ancillary asset, recoverable in a derivative action by the asset originator or by token holders themselves (with a 5%-of-supply standing threshold for derivative claims and a two-year statute of limitations). The structure mirrors §16(b) short-swing profit recovery under the Exchange Act (15 U.S.C. § 78p(b)) and will be familiar territory for plaintiff-side counsel.

§104(f) mandates a set of exemptions that reveal where the bill expects friction. Material hardship exemptions (death, bankruptcy, dissolution, tax liability arising from token receipt) under §104(f)(1). Liquidity provision exemptions for two-sided market making under §104(f)(2). Agency exemptions for custodians, brokers, dealers, and trading platforms under §104(f)(3). ETP and passive pooled investment vehicle exemptions, with explicit accommodation for authorized-participant create-and-redeem mechanics, under §104(f)(4). The §104(f)(4) exemption is particularly important for the spot Bitcoin and Ether ETPs that came online in 2024 and 2025; without it, related-person tokenholders of an ETP-eligible asset would face disposition restrictions on routine fund operations.

Hinman gave the industry a vibe. §104 gives it a test. The test has five indicia, one hard number, three safe harbors, and a certification procedure that produces final agency action. None of those pieces are self-executing. The SEC’s coordinated-control rulemaking will take eighteen to twenty-four months at best, and the resale restrictions will not lift on any current insider’s preferred timeline. But the architecture is different in a way that matters. Decentralization is no longer a litigation defense to be raised after an enforcement action. It is a statutory standard with a rulemaking path, a certification window, and a Rule 144 analog at the end of it. The “sufficient decentralization” theory survives in §104(b)(2), but only as one of five factors, and only one of those factors has a number attached to it. That number is 49. Hinman would not recognize the section that buried him.

Written by David Lopez Kurtz