The SEC’s 2019 token framework made security status a fact to recheck

The SEC staff’s April 2019 framework tied a digital asset’s status to facts about other people’s efforts, and said those facts could change. We argue that for banks, broker-dealers and advisers, that made classification a control with an owner and a trigger, not a memo.

Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.

Two-minute summaryThe dek and the first sentence of each section, quoted from the article

The SEC staff’s April 2019 framework tied a digital asset’s status to facts about other people’s efforts, and said those facts could change. We argue that for banks, broker-dealers and advisers, that made classification a control with an owner and a trigger, not a memo.

  1. The week On April 3, 2019, the staff of the Securities and Exchange Commission published a Framework for “Investment Contract” Analysis of Digital Assets.
  2. The question and who answers it The framework’s threshold question is whether a digital asset is a security.
  3. Who owns the record A control needs an owner, and the classification control raises a question about which function that should be.
  4. Where the facts come from The framework names the facts that matter.
  5. The facts the answer rests on The framework’s longest section concerns the third Howey element: a reasonable expectation of profits derived from the efforts of others.
  6. When the facts change The framework addresses change directly.
  7. Where the answer goes A classification is useful only where it changes what the firm does.
  8. Testing that the control works A control that exists on paper may still fail in operation.
  9. What a reviewer may ask We do not describe any regulator’s examination procedures here.
  10. Objections: staff views, small exposure and rules still to come The statement said plainly that the framework was not a rule and not binding on the Commission.
  11. What the control looks like on paper To make the argument concrete, we set out what we think a minimal record for one asset contains.
  12. With hindsight The framework did not last.

The week

On April 3, 2019, the staff of the Securities and Exchange Commission published a Framework for “Investment Contract” Analysis of Digital Assets. The framework opens with a direct instruction: if you are considering an initial coin offering, or otherwise engaging in the offer, sale or distribution of a digital asset, you need to consider whether the U.S. federal securities laws apply. It applies the test from the Supreme Court’s Howey case, under which an investment contract exists when there is the investment of money in a common enterprise with a reasonable expectation of profits to be derived from the efforts of others.

The framework’s longest discussion concerns the last element. It introduces the term “Active Participant,” a promoter, sponsor or other third party that provides essential managerial efforts affecting the success of the enterprise. It lists characteristics that make reliance on such a party more likely, among them that the party is responsible for developing, improving, operating or promoting the network, and that essential tasks are performed by that party rather than by an unaffiliated, dispersed community of network users. It lists characteristics that bear on an expectation of profit, among them that the digital asset is transferable or traded on or through a secondary market or platform, or is expected to be.

The framework also addresses change. It says that in evaluating whether a digital asset previously sold as a security should be reevaluated at the time of later offers or sales, there would be additional considerations, including whether the efforts of an Active Participant continue to be important to the value of an investment in the asset. In a footnote, it says the staff expects that analysis concerning digital assets as securities may evolve over time as the digital asset market matures.

A statement released the same day by the Director of the Division of Corporation Finance and the SEC’s Senior Advisor for Digital Assets and Innovation said the framework may apply to entities offering, selling or distributing; marketing or promoting; buying, selling or trading; facilitating exchanges; holding or storing; offering financial services such as management or advice; and providing other professional services. It said the framework represents staff views, is not a rule, regulation or statement of the Commission, and is not binding on the Divisions or the Commission. It described the framework as an analytical tool to help market participants assess whether the federal securities laws apply to the offer, sale or resale of a particular digital asset, and invited questions through FinHub, the SEC’s Strategic Hub for Innovation and Financial Technology.

The same day, the Division of Corporation Finance issued a no-action letter to TurnKey Jet, Inc. It said the Division would not recommend enforcement action if the company offered and sold its tokens without registration, noting among other things that the tokens would be immediately usable to buy air charter services, would be sold at one dollar each throughout the program, and could be transferred only to wallets on the company’s platform. The letter said any different facts or conditions might require the Division to reach a different conclusion.

The April 2019 framework’s opening sentence speaks to anyone considering an initial coin offering or otherwise engaging in the offer, sale or distribution of a digital asset, and token issuers deciding whether to register an offering are its most direct readers. This article reads it from a different seat: the bank, broker-dealer or investment adviser that does not issue tokens but touches them, by holding them for clients, letting clients trade them, advising on them or marketing products that include them.

For those firms, the framework carried a message that was easy to miss. It made a digital asset’s status under the securities laws depend on facts about other people, chiefly the promoter or developer the framework calls an Active Participant, and it said those facts could change. A status that rests on facts outside the firm and can change over time cannot be settled once in a legal memo and filed. It has to be owned, sourced and rechecked. In our view, that made classification a control.

The article’s core sections cover four parts of the control, and a later section takes up objections:

  • The question and who answers it: what the framework asked, and why the answer sat with the firm rather than the regulator.
  • The facts the answer rests on: which facts the framework said matter, and who holds them.
  • When the facts change: the framework’s reevaluation language, the no-action letter’s conditions, and what a recheck needs.
  • Where the answer goes: the business lines that depend on a classification and must move together when it changes.
  • Objections: staff views are not law, the exposure is small, and rules were coming anyway.

The four parts follow a classification through its life: asked, answered, changed and used. Other sections cover who owns the record, where the facts come from, how to test the control, what a reviewer may ask and what the record looks like on paper. Throughout, we describe the framework as it stood in 2019. The hindsight section explains that it has since been withdrawn and superseded, and what replaced it.

The question and who answers it

The framework’s threshold question is whether a digital asset is a security. Its focus is one kind of security, the investment contract, tested under Howey. The framework is clear that the analysis is about the asset and also about the circumstances: the Howey analysis looks at the manner in which the asset is offered, sold or resold, which includes secondary market sales.

The answer was the firm’s to give

The staff statement said the framework was an analytical tool to help market participants assess whether the federal securities laws apply. It was not a list of assets and their status. It did not say which tokens were securities. It gave firms a method and left the conclusion to them, with an invitation to contact the staff through its FinHub webform if questions remained after applying the framework.

That allocation matters for a regulated firm. When a regulator publishes a list, a firm can cite the list. When a regulator publishes a method, the firm owns the conclusion and the reasoning behind it. If an examiner or a court later disagrees, we think the firm’s strongest position rests on the quality of its analysis at the time it was made, and that analysis is only as good as the record that shows it.

Staff views, not Commission rules

The statement said the framework represents staff views and is not a rule, regulation or statement of the Commission. It said the Commission had neither approved nor disapproved its content, and that the framework was not binding on the Divisions or the Commission. It also said it did not constitute legal advice and did not modify or replace any existing laws, regulations or rules.

For a firm, those caveats cut both ways. They meant the framework could not be relied on as a safe harbor. They also meant the framework was the clearest public statement of how the staff thought about the question, and a firm that ignored it would have trouble explaining why. We think the reasonable reading for a compliance officer was to apply the framework, record that it had been applied, and record its status as staff guidance alongside the conclusion.

Which firms were in scope

The statement’s list of activities is broad. Offering, selling or distributing. Marketing or promoting. Buying, selling or trading. Facilitating exchanges. Holding or storing. Offering financial services such as management or advice. Other professional services. A bank that holds digital assets for clients is holding or storing. A broker-dealer that lets clients trade them is buying, selling or trading, and may be facilitating exchanges. An adviser that recommends them is offering advice. A firm that markets a fund holding them is marketing or promoting. Each of those activities may be regulated differently if the asset is a security, which is why the classification question reached past the issuer to the firms around it.

Who owns the record

A control needs an owner, and the classification control raises a question about which function that should be. Three candidates present themselves in a bank or broker-dealer: the business line that wants to offer the asset, the legal department, and the compliance function.

Why not the business line

The business line knows the asset best and has the strongest interest in a particular answer. A trading desk that wants to list an asset, or a custody team that wants to win a client who holds one, has a reason to prefer the classification that makes the activity easier. That is a conflict, not a judgment about anyone’s good faith. We think the business line should supply facts and argue its view, and someone else should decide.

The classification is a legal conclusion. The staff statement says the framework does not constitute legal advice, for which readers should consult their own attorney, and the framework itself encourages market participants to seek the advice of securities counsel. We think the conclusion belongs to legal counsel, inside or outside the firm. The record, the triggers, the schedule and the distribution to business lines are compliance work. Splitting the roles this way keeps the analysis with the people qualified to make it and the monitoring with the people whose job is to notice change.

Escalation and disagreement

Some classifications will be close. The framework says no single characteristic is necessarily determinative, which leaves room for reasonable people to differ. A firm can decide in advance who resolves a disagreement between legal and the business, and what happens to the activity while it is unresolved. A rule that the activity pauses, or continues under the more restrictive treatment, until the question is settled removes the incentive to delay a decision.

Board and risk committee

For a firm whose digital asset activity is material, the board or its risk committee will want to know the control exists and works. The measures we describe later in this article are designed for that audience: counts, review dates and the time it takes a change to reach every business line.

Where the facts come from

The framework names the facts that matter. It does not say where to find them, and the answer shapes how reliable the classification is.

The developer’s own statements

Developers and foundations may publish roadmaps, white papers, blog posts and governance proposals. These are the most direct evidence of the Active Participant’s role, and the least independent. A firm can record what the developer says it does, and should record separately any evidence that confirms or contradicts it.

Governance and technical records

A network may keep public records of who proposes and approves changes, who holds keys that can alter the network and who controls its code repositories. Those records speak to the framework’s question of whether essential tasks are performed by the Active Participant or by a dispersed community. They are technical, and a firm may need engineering help to read them.

Market data

Trading venues, volumes and price behavior speak to the framework’s questions about secondary markets and expectations of profit. The framework’s note that price appreciation resulting solely from external market forces is generally not profit under Howey means market data needs interpretation as well as collection. A firm can record where it got the data and what it concluded from it.

Outside data providers

A firm may buy classification opinions or risk ratings for digital assets from outside providers. They can save time. They cannot take the firm’s place as the party responsible for its conclusion, because the framework left the conclusion to the firm. We think a vendor’s view belongs in the evidence file as one input, labeled as such, with the vendor’s own basis attached where the vendor provides it.

The facts the answer rests on

The framework’s longest section concerns the third Howey element: a reasonable expectation of profits derived from the efforts of others. It does not give a formula. It gives characteristics, and it says that the stronger their presence, the more likely the element is met. For a firm running a classification control, those characteristics are the facts to collect.

Facts about the Active Participant

The framework says that when a promoter, sponsor or other third party provides essential managerial efforts that affect the success of the enterprise, and investors reasonably expect to derive profit from those efforts, the efforts-of-others element is met. It then lists characteristics, including that the Active Participant is responsible for the development, improvement, operation or promotion of the network; that essential tasks are performed by the Active Participant rather than an unaffiliated, dispersed community of network users; and that the Active Participant creates or supports a market for the asset or has a continuing managerial role in decisions about the network.

Every one of those facts is about someone other than the firm doing the analysis. The firm does not control the developer’s roadmap, governance or marketing. It can observe them through public statements, technical documentation, governance records and on-chain data, but it learns about changes only if it is watching.

Facts about the expectation of profit

The framework lists characteristics bearing on whether purchasers reasonably expect profit. Among them: the asset gives holders rights to share in the enterprise’s income or profits; the asset is transferable or traded on or through a secondary market or platform, or is expected to be; and the asset is offered broadly to potential purchasers rather than targeted to expected users of the goods or services or those who need the network’s functionality.

The framework adds that purchasers would reasonably expect an Active Participant to undertake efforts to promote its own interests and enhance the value of the network or asset in certain situations. Its examples include an Active Participant that can realize capital appreciation from the asset’s value, and one that distributes the asset as compensation to management or whose compensation is tied to the asset’s price in the secondary market.

Some of these facts are visible in trading data. Others depend on how the asset is marketed, including by people the firm has no relationship with. The framework also notes that price appreciation resulting solely from external market forces is generally not considered profit under Howey, which means the firm has to separate market-wide movements from asset-specific promotion when it reads the evidence.

Facts about consumption

The framework also describes characteristics that point the other way, toward an asset bought for use rather than investment. Among them: the distributed ledger network and asset are fully developed and operational; holders can immediately use it for its intended functionality; the asset’s creation and structure is designed and implemented to meet the needs of its users rather than to feed speculation; and any economic benefit from appreciation is incidental to obtaining the right to use it for its intended functionality.

For a classification control, these are the facts that would support a conclusion that an asset is not, or is no longer, an investment contract. They are as dependent on outside parties as the facts that point the other way.

The evidence file

Put together, the framework implies an evidence file for each asset: who the Active Participant is, what it does, how the asset is marketed and traded, and how it is used. We think the file should record where each fact came from and when it was observed. A conclusion that rests on a developer’s statement from a year ago is weaker than one that rests on last month’s governance records, and the file should show which kind it is.

When the facts change

The framework addresses change directly. It says that, in evaluating whether a digital asset previously sold as a security should be reevaluated at the time of later offers or sales, there would be additional considerations. On the efforts of others, it asks whether the efforts of an Active Participant, including any successor, continue to be important to the value of an investment in the asset. On expectation of profit, it asks whether purchasers no longer reasonably expect that the continued development efforts of an Active Participant will be a key factor in determining the asset’s value.

Change in both directions

The reevaluation language is written for an asset moving away from security status, as development finishes and a network becomes more dispersed. The characteristics listed elsewhere in the framework work in both directions. A developer could take on a larger managerial role. An asset could begin trading on a new platform. Marketing could shift toward price. A firm that treats its classification as fixed will miss movement either way.

Successors, forks and new versions

The framework’s reevaluation language refers to the efforts of an Active Participant, “including any successor AP.” That phrase points to a class of change a firm can easily overlook. A developer may hand the network to a foundation, a foundation may hand governance to token holders, or a company may be acquired. A network may split into two versions with different developers behind each. Each of these changes who the Active Participant is, and so changes the facts the classification rests on. We think a firm’s triggers should name them explicitly, and that a split should be treated as the creation of a new asset needing its own record.

Changes the firm makes itself

Some changes come from inside. A firm that starts offering staking or lending on an asset it already holds, or that begins marketing it differently, changes the circumstances in which the asset is offered or sold. The framework says the Howey analysis looks at the manner of offer and sale, including resales. A new product built on an existing asset is a reason to reopen the asset’s record, even if nothing about the developer has changed.

The no-action letter as a list of conditions

The TurnKey Jet letter shows what a staff position tied to facts looks like. The Division noted that the company would not use token sale proceeds to develop its platform, network or app, each of which would be fully developed and operational before any tokens were sold; that the tokens would be immediately usable for air charter services; that transfers would be restricted to wallets on the platform; that tokens would be sold at one dollar each throughout the program; that any repurchase would be at a discount to face value unless a court within the United States ordered it to liquidate the tokens; and that the token would be marketed in a manner emphasizing its functionality rather than the potential for an increase in market value.

The letter said its position was based on the representations made to the Division and that any different facts or conditions might require a different conclusion. Read as a control, the letter is a list of conditions that must remain true. If any one of them stopped being true, the basis for the position would change. The letter also said it expressed the Division’s position on enforcement action only and did not express any legal conclusion on the question presented. A firm relying on an outcome of that kind would need to watch each condition as well as the headline conclusion.

What a recheck needs

We think a recheck process has three elements. First, triggers: named events that reopen a classification, such as a change in the developer’s role, a new trading venue, a new marketing campaign, a governance change or a hard fork. Second, a schedule: a periodic review even when no trigger has fired, because the firm may not learn of every event. Third, an owner: a named person or committee who decides whether the classification still holds and signs the record.

The schedule matters because triggers depend on the firm noticing. A firm that holds a handful of assets for clients may follow each one closely. A firm with a long list may not, and a periodic review is the backstop.

Recording the guidance that applied

A recheck should also record which guidance it applied. The framework said analysis may evolve over time as the digital asset market matures. Guidance can be updated, withdrawn or replaced, as this one later was. A classification record that names the guidance it relied on, and the date, lets a later reviewer judge the decision by the rules of its time, and shows the firm when its earlier analysis needs to be redone under new guidance.

Where the answer goes

A classification is useful only where it changes what the firm does. In a bank, broker-dealer or adviser, the same asset may be handled by several business lines, each of which uses the answer differently. A change in the answer has to reach all of them at the same time, or the firm ends up treating one asset as a security in one place and not in another.

Custody

A firm holding assets for clients needs to know which legal regime governs that holding. Whether an asset is a security can affect the custody rules that apply, the books and records the firm keeps and the disclosures it makes. We do not set out those rules here; they differ by type of firm and organizations should check with counsel. The point for this article is that custody systems may tag assets by type, and the tag should come from the classification record rather than being set once at onboarding.

Trading and venues

A broker-dealer that lets clients buy and sell an asset, or a platform that matches orders, needs to know whether it is handling securities transactions. The framework’s list of activities includes buying, selling or trading and facilitating exchanges. A change in classification may change whether an asset can be offered on a given venue at all. We think that makes trading the business line where a late update does the most harm, because trades continue while the record is out of date.

Advice and suitability

An adviser recommending an asset, or including it in a model portfolio, relies on its classification for the obligations that attach to advice about securities. The framework names offering financial services such as management or advice among the activities to which it may apply. We think an adviser’s research process should cite the firm’s classification record rather than repeat its own analysis, so that both change together.

Marketing and disclosure

Firms that market products holding digital assets describe those assets to clients. The framework lists marketing or promoting among the covered activities. Marketing copy can also become evidence in someone else’s analysis: the framework asks how the asset is marketed, and the no-action letter noted marketing that emphasized functionality. A firm’s own descriptions should be checked against its classification so that the firm does not describe as an investment something it has classified as a consumptive asset, or the reverse.

One record, many users

The design conclusion is simple. One classification record per asset, owned by one function, read by every business line through a system rather than copied into each. When the record changes, each business line receives the change and confirms what it did. The confirmation is the evidence an examiner would ask for: that the firm knew, and acted, everywhere the asset was handled.

Testing that the control works

A control that exists on paper may still fail in operation. Three tests, each cheap, tell a firm whether its classification control does what it claims.

Reconcile the business lines

The first test is a reconciliation. Take the list of assets in the classification register and compare it with the assets each business line actually handles: those held in custody, those available for trading, those covered by research or advice, and those named in marketing material. Any asset that a business line handles and the register does not list is a gap in coverage. Any asset whose treatment in a business line’s system differs from the register is a gap in distribution. Both kinds of gap are findings, and both are easier to fix when found by the firm than by an examiner.

Fire a trigger on purpose

The second test is a drill. Pick an asset, pretend one of its listed conditions has changed, and follow the process. Who would learn of the change, and how? How long until legal reviews it? How long until each business line confirms the new treatment? The drill measures the path rather than the analysis, and the path is where delays hide. A firm that finds its trading system would keep offering an asset for a week after a reclassification has learned something worth knowing.

Sample the evidence

The third test is a sample review. An independent reviewer, such as internal audit, picks a few classifications and checks that each fact in the basis has a source and a date, that the guidance applied is named, and that the last review happened when the schedule said it would. The sample does not second-guess the legal conclusion. It checks that the conclusion rests on a record someone else can follow.

Who should see the results

The results of all three tests belong with the owner of the control and, for a firm with material activity, with the risk committee. A short note of what was tested, what was found and what was fixed is the evidence that the control is maintained, which is a different thing from evidence that it was designed.

What a reviewer may ask

We do not describe any regulator’s examination procedures here. We describe the questions we would expect a careful reviewer, internal or external, to ask of a firm that touched digital assets under the 2019 framework, and the record that would answer each.

  • Which assets did the firm touch, and in which business lines? The register and the reconciliation answer this.
  • On what basis did the firm conclude each asset was or was not a security? The classification record, with its sourced facts and the guidance applied, answers this.
  • How did the firm learn when the facts changed? The triggers, the schedule and the review history answer this.
  • When the classification changed, how fast did each business line follow? The distribution log and the drill results answer this.
  • Who decided, and who checked? The ownership split between legal and compliance, and the independent sample review, answer this.

None of those answers requires the firm to have been right about every asset. The framework was clear that the analysis depends on facts and circumstances and may evolve. A firm that can show a reasonable process, applied consistently and recorded as it went, is in a stronger position than one that reached the same conclusions without a record.

Objections: staff views, small exposure and rules still to come

Staff guidance is not law

The statement said plainly that the framework was not a rule and not binding on the Commission. A firm might decide that guidance with no binding force does not justify a standing control. Our answer is that the underlying law, the Howey test, was binding and remained so, and that the framework’s method was the staff’s own reading of it. The control is a way of applying the law with a record. The framework only described which facts the staff thought mattered.

The exposure is small

For some firms, digital assets may have been a small part of the business, and others may have held none at all. For a firm with no exposure, the control is unnecessary. For a firm with some exposure, we think the control can be sized to it: a short register, a named owner, a quarterly review. The cost of the record is low. The cost of discovering, after the fact, that one business line treated an asset as a security and another did not is higher.

Better to wait for rules

A firm might also wait for formal rules or legislation before building anything. Formal action did come, though not in 2019, as the hindsight section describes. Waiting has a cost: every asset the firm touched in the meantime still needed a classification, and the classification still needed a basis. A firm that kept records under the framework had something to compare against the new guidance when it arrived. A firm that did not had to reconstruct its past reasoning.

The control outlives the guidance

The strongest reason to build the control is that it does not depend on any particular guidance. The facts to collect may change when the guidance changes, and the categories may change. The structure stays the same: a record per asset, a basis in sourced facts, triggers and a schedule for review, an owner, and a path to every business line that depends on the answer. That structure fit the 2019 framework, and it fits the guidance that replaced it.

The control is too slow for a fast market

A last objection is speed. Digital asset markets trade around the clock, and facts can change in hours. A review process built on committees and quarterly schedules may seem too slow to matter. We think the answer is to separate two decisions. The full legal analysis can take the time it needs. The interim decision, what the firm does with the asset while the analysis runs, can be made quickly under a rule set in advance. A firm that has decided ahead of time to apply the more restrictive treatment when a trigger fires can act within hours, and can relax the treatment later if the analysis supports it. Speed comes from the rule written in advance, not from rushing the legal conclusion.

What the control looks like on paper

To make the argument concrete, we set out what we think a minimal record for one asset contains. It is not a regulatory requirement; it is our description of a practical control.

  • Identity: the asset, its network, its known identifiers and the date the firm first touched it.
  • Classification: the firm’s conclusion, the guidance applied, and the date and owner of the decision.
  • Basis: the facts relied on, each with its source and the date it was observed.
  • Conditions: the facts that, if they changed, would reopen the classification.
  • Users: the business lines and systems that read the classification.
  • History: every prior classification, with its basis, kept rather than overwritten.

The history line matters most in hindsight. When guidance changes, the firm needs to show what it concluded under the earlier guidance and why. A record that overwrites itself loses that. A record that keeps its history lets the firm answer the question an examiner will ask: what did you know, and when.

The record can also feed board and committee reporting. A short periodic view can show how many assets the firm touches, how many were reviewed in the period, how many classifications changed and how fast each change reached every business line. Those measures tell a risk committee whether the control is working without requiring its members to follow the legal analysis.

Clients who bring their own assets

One case deserves a line of its own. A custody or brokerage client may ask the firm to hold or trade an asset the firm has not reviewed. The control should say what happens then: whether the firm declines until a classification exists, accepts the asset under the more restrictive treatment while it reviews, or has some other rule. Whatever the rule, it should be written down and applied the same way to every client, because a firm that makes exceptions case by case ends up with classifications that depend on who asked.

With hindsight

The framework did not last. The SEC page that carried it, as checked on November 15, 2026, marks it as withdrawn and superseded by Application of the Federal Securities Laws to Certain Types of Crypto Assets and Certain Transactions Involving Crypto Assets, dated March 17, 2026. That release, in which the SEC issued an interpretation and the Commodity Futures Trading Commission provided guidance relating to it, states in a footnote that the interpretation supersedes the staff’s Framework for “Investment Contract” Analysis of Digital Assets of April 3, 2019. Its effective date is March 23, 2026.

The replacement is a Commission interpretation rather than staff guidance. It says the Commission and its staff will administer the federal securities laws consistent with it, including with respect to enforcement actions, and it describes itself as the Commission’s first step toward a clearer regulatory framework for crypto assets. It classifies crypto assets into five categories: digital commodities, digital collectibles, digital tools, stablecoins and digital securities. It says the first three are not themselves securities, that stablecoins may or may not be securities depending on their characteristics, and that digital securities are securities. It also says it does not supersede or replace the Howey test, which it calls binding legal precedent. It solicits public comment and says the Commission may refine, revise or expand the interpretation based on the feedback.

The idea at the center of this article survived the change. The 2026 release says a non-security crypto asset can be offered and sold subject to an investment contract, and it discusses how such an asset may become subject to, and cease to be subject to, an investment contract. One indicator it describes is the issuer’s fulfillment of the representations or promises it made to engage in essential managerial efforts. That is a fact about another party’s conduct, observed over time, which is the kind of fact a classification control exists to track.

The rules around holding these assets moved too. On January 23, 2025, the SEC staff issued Staff Accounting Bulletin No. 122, which rescinded the interpretive guidance in Topic 5.FF, Accounting for Obligations to Safeguard Crypto-Assets an Entity Holds for its Platform Users, effective January 30, 2025. On March 7, 2025, the OCC published Interpretive Letter 1183, confirming that crypto-asset custody, certain stablecoin activities and participation in independent node verification networks are permissible for national banks and federal savings associations. The OCC said the letter rescinds the requirement for OCC-supervised institutions to receive supervisory nonobjection and demonstrate that they have adequate controls in place before they can engage in those activities. Acting Comptroller of the Currency Rodney E. Hood said the OCC expects banks to have the same strong risk management controls in place to support novel bank activities as they do for traditional ones. The 2026 release also notes that Congress enacted the GENIUS Act in July 2025, creating a framework for payment stablecoins and excluding from the definition of security any payment stablecoin issued by a permitted payment stablecoin issuer.

What held up, as we read it, is the structure. The categories changed and the authority behind them rose from staff to Commission. A firm that had kept a dated classification record, with sources and history, under the 2019 framework could map each asset to the new categories and show what it had concluded before. A firm that had filed a one-time memo would have had more to reconstruct. The status of an asset still rests on facts that can change, so the recheck remains part of the job.

Spotted an error? Tell us and we will check it and correct the article where needed.

  • SEC
  • digital assets
  • Howey test
  • no-action letter
  • custody
  • crypto