The commission has put forward a crypto custody framework under the two federal investment statutes that govern advisers and funds. It is a proposal, not a finished rule. Public comments would stay open for 60 days after the text lands in the Federal Register, which means the clock does not start on the day a chairman posts about it. Still, the shape of the idea is clear enough to argue with. Conditional self-custody would be allowed when no permitted custodian exists. State trust companies could serve, but only after documented checks. Registered investment advisers, registered investment companies, and business development companies sit inside the perimeter.
Why Old Custody Rules Keep Missing New Coins
The custody regime that advisers live with was written for paper, then patched for electronic brokerage records. It assumes an asset has a home. A bank. A broker-dealer. A futures commission merchant. Someone with a vault, a control report, and a surprise exam. Crypto broke that assumption in a very practical way. A token can trade for months before any firm on the permitted list builds wallet support, insurance language, and an operations manual around it.
I have watched that gap play out in ordinary client conversations. Someone wants a small sleeve of a newly listed asset. The adviser can explain the thesis. The adviser cannot point to a custodian that will take the position this quarter. Under a strict reading of the old rule, advice starts to look legally awkward the moment the asset has to sit somewhere. That is not a theoretical puzzle. It is a calendar problem.
The chairman described parts of the Investment Advisers Act of 1940 and the Investment Company Act of 1940 as predating the internet. That line is a little theatrical, and also mostly fair. Those statutes were built around loss, theft, and misuse of traditional assets. They were not built around seed phrases, smart-contract admin keys, or a token that a custodian’s vendor has not whitelisted. “Unfortunately, our rules and regulations have not kept pace,” he said. Clients chasing what he called a multitrillion-dollar asset class have been living inside that lag.
A custody rule that only works after the market has already moved is a rule that arrives late to its own job.
A compliance officer I trust, after a very long Monday
The proposal is filed as S7-2026-35, with release numbers IA-7023 and IC-36353. Those numbers matter less to most readers than the practical bargain inside them. The agency wants to remove barriers that have kept regulated advice and fund strategies away from crypto, without pretending that “we will figure out the keys later” is a control environment. Whether that bargain holds is the whole argument of the comment period.
What The Proposal Actually Covers
Three groups sit at the center. Registered investment advisers. Registered investment companies. Business development companies. If you advise client accounts for a fee, or you run a fund that has to live inside the 1940 Act machinery, this is aimed at you. It is not a retail how-to for someone keeping coins in a phone wallet at home. A commissioner was careful on that point. In her account, self-custody here means the adviser holds assets for clients. It does not mean an investor skipping intermediaries altogether.
That distinction sounds fussy until you sit with it. Retail self-custody and advisory self-custody fail in different ways. A person who loses a seed phrase loses their own money. An adviser who loses a seed phrase, or lets an employee wander off with one, has a fiduciary mess, a client mess, and very possibly an enforcement mess. The proposal treats those as different animals. Good. They are.
Beyond digital assets, the amendments would also refresh expectations around financial statement audits for registered advisers and around broker-dealer custody services used by regulated funds. So this is not a pure crypto side letter. It is a custody update that happens to have crypto as the sharp edge. In my view that is the smarter way to write it. A rule that only mentions tokens tends to age badly the moment the next asset type shows up.
The Delay Problem, In Plain Language
Here is the scenario the chairman keeps pointing at. A new crypto asset launches. Custody vendors need time. Integration, legal review, insurance, chain support, sometimes a board approval. Several months can pass before a permitted custodian will take it. During those months, an adviser who wants to stay inside the old rule has two ugly choices. Decline the exposure. Or hold the asset in a way the rule never really contemplated.
The framework tries to give a third door. Hold it yourself, but only after you have actually looked for a permitted custodian and failed to find one. Then look again every quarter. If a qualified home appears, the excuse expires. That quarterly loop is the part I would not skim. It turns a one-time exception into an ongoing duty. Firms that treat it as a permanent workaround are going to have a bad meeting with their own compliance file.
Conditional Self-Custody Is Not A Blank Check
The phrase conditional self-custody is doing a lot of work. Conditional is the word that should make operations teams sit up. An adviser would first have to determine that no permitted custodian is available for that asset. Then the adviser would repeat the assessment every quarter. Miss the repeat, and the rationale for holding the keys yourself starts to rot.
I like the instinct. I am less sure about the paperwork. “No permitted custodian is available” can mean several things, and advisers will want the release to say which ones count.
- No firm on the permitted list supports the chain at all.
- Firms support the chain but refuse this specific token.
- A firm will take it, but only above a size your client will never reach.
- A firm will take it next quarter, and “next quarter” keeps moving.
- A firm will take it, at a fee that makes the position pointless.
Fee and timing are where comment letters usually get sharp. If a custodian exists but charges a rate that swallows the allocation, is that custodian “available”? If onboarding takes five months, is the asset unsupported today? A rule that does not answer those questions will be interpreted by the most aggressive counsel in the room. Perhaps that is inevitable. It is still worth forcing the issue now, while the text can still move.
There is also a human version of this test. Someone has to sign the memo. Someone has to keep the emails from custodians who said no. Someone has to calendar the quarterly redo so it does not die in a shared inbox. I have found that controls fail less often on policy language than on the Tuesday when the person who owned the calendar leaves the firm. Self-custody makes that Tuesday more expensive.
What Holding The Keys Would Actually Require
The proposal’s public summary is thinner than an operations manual, which is normal at this stage. Still, anyone who has stood up a wallet program can sketch the minimum. You need generation of keys in a controlled setting. You need storage that is not a laptop in a drawer. You need a way to approve a send that does not depend on one person. You need a record of who touched what. You need a plan for the day a signer is unavailable, and a different plan for the day a signer is the problem.
Think of it like a spare house key, except the house is a bearer asset and the key can be copied without a locksmith ever knowing. That metaphor is a little crude. It is also the one clients understand. Advisory self-custody has to be built so that copying the key is hard, noticing a copy is possible, and moving funds requires more than one tired person at 11 p.m.
A sensible internal standard, even before the final rule tells you the exact one, would cover at least these pieces.
- A written finding that no permitted custodian can take the asset, with names and dates.
- A quarterly refresh of that finding, stored where an examiner can actually find it.
- Key generation and backup that do not live in one employee’s password manager.
- Dual control on withdrawals above a modest threshold.
- A lost-key and a rogue-employee playbook that has been rehearsed, not just filed.
- Client disclosure that says, in ordinary words, who holds the asset and what can go wrong.
None of that is exotic. Private funds and a few larger advisers already run versions of it. The shift is that a federal proposal would pull the practice out of the gray and pin a duty to it. Gray was comfortable. Duty is clearer, and heavier.
State Trust Companies Get A Narrow Door
The second door is outside custody through a state trust company. This one has been a sore spot for years. Some states charter trust companies that genuinely custody digital assets, with capital rules, exams, and segregation requirements. Other charters look thinner from a distance. Advisers have not always known which pile a given firm belonged to, and the old federal language did not give them a clean way to decide.
Under the proposal, advisers and funds would assess the company’s state authorization and its written safeguards before appointing it. Then they would repeat those checks every year. The safeguards have to speak to theft, loss, misuse, and misappropriation of crypto assets and of related cash. Related cash is easy to forget and rude to ignore. A lot of real losses happen in the fiat leg, the wire, the omnibus account, not in the cold wallet everyone photographs for the pitch deck.
Annual is slower than the quarterly self-custody test, which makes sense if you think the trust company is a standing vendor rather than an emergency shelf. It does not make sense if the trust company’s control environment is deteriorating in March and your next review is scheduled for January. Good firms will not wait for the anniversary. The rule, if adopted as sketched, would set the floor, not the ceiling.
| Path | When it fits | Review rhythm | Main risk |
| Permitted custodian | Asset is supported and onboarding is real | Ongoing vendor oversight | Support lags a new listing |
| Conditional self-custody | No permitted custodian will take the asset | Quarterly availability test | Key loss or insider misuse |
| State trust company | Charter and safeguards check out | Annual authorization and safeguard review | Thin charter, weak segregation |
| Do not hold | Neither door is honest | Not applicable | Clients go elsewhere, or around you |
That last row is the one marketing teams dislike. Sometimes the compliant answer is still no. A proposal that creates two new paths does not obligate a firm to use either of them. Fiduciary duty does not flip into a product mandate because a comment file opened.
How A Diligence File Should Read
If I were building the state-trust review from scratch, I would want the file to answer questions a skeptical partner could ask without notice. Who chartered this company? What does the charter actually authorize, in verbs, not slogans? Who examines it, and when was the last exam? Where do client coins sit versus the company’s own coins? What happens in insolvency, on paper? Who holds the admin keys, and can a single cloud login move funds?
Written safeguards should be more than a PDF with the right nouns. Theft, loss, misuse, misappropriation. Those four words are the commissioner’s list, and they map onto different failures. Theft is an outsider. Loss is a bad backup. Misuse is a process that allows a “temporary” transfer. Misappropriation is an insider who decides the coins are a bonus. A safeguard document that only discusses hackers has done a quarter of the job.
Cash is the quiet twin. Fiat on-ramps, redemption accounts, and the bank that holds them need the same sour gaze. I have seen digital-asset programs with beautiful cold storage and a sloppy operating account. Examiners notice the sloppy account. Clients notice it later, which is worse.
Funds Are Not Just Advisers With A Logo
Registered investment companies and business development companies live under a tighter set of habits than a typical advisory account. Boards, custody agreements, fidelity bonds, audit cycles, affiliated transactions. Letting those vehicles offer more crypto strategies is explicitly part of the agency’s pitch. The pitch only works if custody is solvable. A fund that cannot point to a lawful place for the coins cannot honestly offer the strategy, no matter how clean the prospectus language looks.
Self-custody inside a fund is a sharper edge than self-custody inside a separately managed account. The keys secure a pool. Shareholders who never met the portfolio manager still own the outcome. Boards will want to know who can sign, what the disaster plan says, and whether the auditor can actually test the balance. “Trust the wallet screenshot” is not going to survive a serious audit committee.
Business development companies add another wrinkle. Many of them already live in private markets where custody of a loan or a share certificate is its own headache. Crypto does not replace that complexity. It stacks on top. A BDC that wants a digital-asset sleeve will need the old controls and the new ones, and a board that can tell them apart.
There is a client-facing version of this too. People buy a fund because they do not want to babysit keys. If the fund itself is babysitting keys, the disclosure has to say so without hiding behind jargon. Qualified custodian is a term of art. “We hold this ourselves because nobody else would” is a sentence. The second one belongs in the document, somewhere a human might read it.
Reporting And Recordkeeping Will Carry The Weight
The docket also flags changes to reporting and recordkeeping. That is the unglamorous half, and it is the half that decides whether the glamorous half is real. A conditional right to self-custody is only as good as the file that proves the condition was met. If the records are vague, the right collapses the first time someone asks.
I would expect examiners to want, at a minimum, the availability memos, the vendor declines, the key-ceremony notes, the signer lists, the withdrawal approvals, and the annual trust-company reviews. They will also want to see that related cash moved through accounts the firm can explain. None of this requires a new philosophy. It requires not losing the email.
Audit language is part of the same package. The amendments touch financial statement audits for registered advisers. Crypto balances have embarrassed audits before, usually when the auditor could not get comfortable with ownership or with the entity that claimed to hold the coins. A rule that invites more on-balance-sheet digital assets without a parallel audit path would be a strange gift. The fact that audits are in the release is a small reassuring signal. The comment file should test whether the signal is backed by procedures an auditor can actually perform.
This Is Not The Safeguarding Rule That Died
Context helps, because this proposal did not appear from a clear sky. A broader safeguarding effort was withdrawn in June 2025. A custody review later tracked a submission to the Office of Management and Budget in late August, described at the time as a separate rulemaking rather than a revival of the withdrawn text. If you only remember the headlines, it is easy to mash those projects together. They are related in topic and not the same document.
The withdrawn effort had tried to widen who counts as a custodian and how client assets, including crypto, had to be protected. It drew a loud response from custodians, advisers, and funds, much of it about cost and about whether the net had been cast so wide that ordinary practices became noncompliant. Pulling it back was an admission that the first draft could not carry the weight. This newer framework is narrower in ambition and more specific about the gap it wants to close. Newly launched assets that custodians cannot support yet. State trust companies that might be good enough if someone checks. That is a smaller bite. Smaller bites sometimes pass.
Still, do not confuse sequence with consensus. A proposal that follows a withdrawal can be wiser, or it can be the same fight in a shorter coat. Commenters who hated the earlier draft will read this one looking for the old problems in new paragraphs. Commenters who wanted any lawful path for new tokens will read it looking for holes they can live with. Both groups will be in the file. That is how this process is supposed to work.
Where This Sits In A Wider Policy Run
Custody is one tile. The chairman has been pointing at a run of earlier moves, and they matter because custody rules do not operate in a vacuum. Staff issued a December 2025 no-action letter for a voluntary securities tokenization pilot at the Depository Trust Company. A January 2026 classification framework for tokenized securities followed. The commission later interpreted which crypto assets are securities, and when an asset might stop being wrapped in an investment contract.
The tokenization pilot, under a three-year no-action letter, covered a defined set of U.S. securities. Russell 1000 shares. ETFs tracking major indexes. Treasury bills, notes, and bonds. The plan described for the fall of 2026 would let participants create tokenized representations of eligible securities already in traditional custody, move those representations to approved wallets, and convert back. That is a different animal from an adviser holding a native token nobody will custody. It is still part of the same policy weather. Records, wallets, and the line between a security and a representation of a security are all in motion.
Staff also spoke in April about broker-dealer registration implications for certain interfaces used to prepare tokenized securities transactions. In August the commission proposed an offering framework for certain investment contracts involving crypto assets. In September it introduced a five-year innovation exemption for trading tokenized versions of National Market System stocks. Qualifying tokens under that relief have to carry the same company interest, dividend, voting, and liquidation rights as the conventional share. Synthetic stock products are out. Venues have to stop token trading when the underlying stock is halted. Third-party tokens need issuer notice and at least 30 calendar days, and an issuer objection stops the listing.
Why mention all of that in a custody piece? Because an adviser reading only the custody release will miss the perimeter. A tokenized share with real voting rights is not the same object as a meme coin with no issuer and no custodian. The controls you owe a client depend on which object you bought. Mixing them in one “crypto sleeve” is how disclosure goes mushy.
The Comment Clock Is Slower Than Social Media
A practical note, because people will get this wrong. The chairman shared the statement in an early October post. The commission had published the same statement on October 1, alongside the proposed rules. The response period is 60 days after Federal Register publication, not 60 days after the post. Until the Register date exists, anyone claiming the comment window is “already half gone” is guessing.
That lag is annoying if you run a firm and want certainty before year-end planning. It is also the ordinary machinery. Use it. The most useful comments I have read on custody rules are not speeches about innovation. They are scenarios. A two-page walkthrough of a token that three custodians refused, with dates. A description of what an annual trust-company review can realistically verify. A note on how a fund board would oversee self-custody without pretending every director is a cryptography specialist.
If you write, write about the quarterly test, the meaning of available, the cash leg, and the audit trail. Those are the joints where a final rule either becomes usable or becomes another memo that nobody can operationalize.
Who Checks The Keys If The Adviser Holds Them
This is the question that should bother anyone who likes the proposal’s direction. External custodians are not saints. They fail, they get hacked, they sometimes commingle in ways their marketing never mentioned. They do, however, sit in a business where custody is the product. Exams, insurance, and reputation attach to that product. When the adviser becomes the custodian of last resort, the checker and the holder move closer together. That proximity is the risk.
So who checks? Several people, if the system is healthy, and nobody, if it is not. The adviser’s own compliance program checks. The surprise exam or the internal control report, where the existing rule still requires one, checks. The auditor checks, if the engagement is scoped to actually test ownership. For a fund, the board checks. The commission’s exam staff checks later, on a sample, after something looks odd or simply because the cycle came due. Clients check only what they are shown.
That stack can work. It works better when self-custody is rare, documented, and temporary. It works worse when a firm decides the quarterly memo is a formality and the wallet is a business line. I would rather see a narrow exception that sunsets asset by asset than a cultural shift in which advisers become casual custodians because the marketing deck says “full stack.”
The point of a custodian was never the logo on the statement. It was a second set of incentives standing between the adviser and the asset.
Conditional self-custody removes that second set for a while. The conditions are an attempt to put something else in its place. Whether memos and quarterly reviews are a worthy substitute is a fair fight. I lean toward yes for a thin sleeve of unsupported assets, and no as a general model. You can disagree. The comment file is the place for that disagreement to become text.
What Clients Should Ask Before They Cheer
Most clients will never read release IA-7023. They will hear a shorter version. “We can hold crypto properly now.” That sentence is not quite what was proposed, and it is not yet true in any final sense. A client who actually cares where the coins sleep can ask better questions without becoming a lawyer.
- Is this asset at a permitted custodian, a state trust company, or with you?
- If it is with you, when did you last confirm that nobody else would take it?
- Who else has to approve a withdrawal?
- What happens if you, the firm, cannot operate for a week?
- Are the dollars tied to this position in a separate account I can understand?
- Will my statement say where the asset sits, in words rather than a ticker alone?
None of those questions are hostile. A firm with a real program will answer them without flinching. A firm improvising will get poetic. Poetry is a bad sign in custody.
There is a cost conversation too. External custody is not free, and self-custody is not free either once you price people, hardware, insurance, and the exam you hope never comes. Clients who demand both low fees and bespoke new-token exposure are asking for a tension the proposal does not magically dissolve. Someone pays for the control. If the pitch says nobody does, read it again.
A Worked Example, Without The Fairy Tale
Picture a mid-sized adviser. Forty staff. A few hundred households. A compliance lead who already has a full calendar. A client wants two percent of a taxable account in a token that listed six weeks ago. Three well-known custodians have said they might support it “in a future release.” None will take a transfer today. A state trust company will take it, and the charter looks real, but onboarding is ten weeks and the minimum is above the client’s sleeve.
Under the sketch in this proposal, the adviser could document the three declines and the unfit minimum, then hold the position itself, with the quarterly redo on the calendar. Or the adviser could wait for the trust company and tell the client the sleeve starts in ten weeks, if it starts. Or the adviser could say no. All three can be defensible. Only the first one creates a key-management duty the firm may not have staffed.
Now change one fact. A custodian emails in week seven and says support is live. The quarterly test, if done honestly, should push the coins out of the adviser’s wallet and into that custodian. The awkward part is operational. Moving a live position has gas fees, timing risk, and a client who liked the idea of “our own custody” once someone explained it badly. The rule, as described, does not care that the client liked it. Availability ends the exception. Firms should say that out loud at the start, so the later transfer does not feel like a bait and switch.
One more twist. Suppose the asset is not a native coin at all, but a tokenized representation of a Treasury bill inside a controlled pilot, convertible back to the traditional record. That position may never need the self-custody door. Treating it with the same waiver as an unsupported altcoin would be sloppy. Different pipes, different duties. The policy run described earlier is useful here only if advisers actually separate the pipes.
Small Firms And Large Firms Will Feel This Differently
Scale changes the honesty of the option. A large adviser can staff dual control, buy the hardware, pay for an outside review of the key ceremony, and still have someone whose only job is vendor surveillance. For that firm, conditional self-custody is an inconvenient tool on a shelf. Useful when a listing outruns the custodians. Not a personality.
A four-person shop does not have that shelf. The same person who builds the portfolio may be the person who would hold the signer card. Dual control becomes a diagram rather than a practice. I do not think the proposal should ban small firms from the exception. I do think small firms should be slower to pick it up. “We are nimble” is not a control. If the only way to serve the client is to become a custodian you are not built to be, the kinder answer may still be a referral or a wait.
Large fund complexes have the opposite temptation. They can build the machinery, then start to like it, then quietly expand the set of assets they decide “nobody really custodies well.” That drift is how an exception becomes a platform. Boards are the brake. They should ask, once a year at least, which positions are in the exception bucket and why each one is still there.
Insurance, Insolvency, And The Sentences Nobody Reads
Two topics will dominate serious comments even if they stay quiet in the first news cycle. Insurance and insolvency. Self-custody shifts loss toward the adviser in a way client agreements may not currently describe. A crime policy that covers “computer fraud” sometimes does not cover a signer who was tricked, or a backup that was never tested. Reading the endorsement is dull. Skipping it is how a covered story becomes an uncovered one.
Insolvency is the darker page. If an adviser fails, client coins held in the adviser’s own wallet structure need a legal path that is not “hope the bankruptcy estate is polite.” State trust companies at least arrive with a trust vocabulary, which is not the same thing as safety, but it is a vocabulary courts already know. Self-custody arrangements need equally plain words about ownership. Whose asset is this on the worst day? If the document shrugs, the coins will be argued over by people who did not build the wallet.
Related cash shows up again here. Wires in flight, redemption balances, and exchange balances waiting to settle are where insolvency cases get ugly. A crypto rule that forgets the dollar leg is only half a custody rule.
What I Would Watch Between Now And A Final Text
A few markers will tell you whether this stays a narrow fix or swells into something advisers cannot staff.
- How “available” is defined, including fee, minimum, and onboarding delay.
- Whether the quarterly test has a paper standard or just a vibe.
- What annual trust-company review must include, and what “written safeguards” must name.
- How fund boards are expected to oversee an exception they did not design.
- Whether audit and surprise-exam language is updated in the same breath.
- How long self-custody can last once a custodian appears.
- The Federal Register date, because everything else is rumor until then.
I would also watch the tone of the comment file. If custodians write that the exception will be gamed, and advisers write that the exception is still too narrow to use, the final rule may land in a boring middle. Boring middles are often the ones that survive contact with exams. Dramatic rules make better posts and worse operations manuals.
A Cleaner Way To Think About The Trade
Strip the acronyms off and the trade is simple. Markets invent assets faster than custody desks can adopt them. Fiduciary rules hate a gap. The proposal tries to fill the gap with a temporary, documented, self-held pocket, plus a checked path through state trust companies. In exchange, advisers and funds take on duties that used to sit, at least partly, with someone whose entire business was holding other people’s assets.
That trade can be a grown-up one. It can also be a way to launder a wish. The wish is that regulated products should be able to touch every liquid token the week it lists. The statute is older than that wish, and client protection is older than both. A framework that respects the lag, instead of pretending the lag is not there, is more honest than the shrug it replaces.
Will it be adopted as written? I doubt the quarterly wording survives untouched. Comment letters have a habit of sanding down verbs. Will something in this neighborhood become the way serious firms handle the unsupported-asset problem? That feels likely, because the alternative is to keep advising with one eye shut. Shut eyes do not age well once examiners decide the topic is in season.
If you run money for other people, the useful move between now and the Register date is not a victory lap. It is a dry run. Pick one asset you cannot custody today. Write the availability memo as if the rule already existed. Price the key ceremony. Call the state trust company and ask what their safeguard packet actually contains. You will learn more from that afternoon than from another summary of the press statement.
And if you are the client on the other side of the table, ask where the coins would sleep. Then ask who else has to agree before they move. The proposal, at its best, is an attempt to make those answers exist. It has not made them final. It has made them harder to dodge.
Practical Prep While The Text Is Still Soft
Nothing in a proposing release obligates a firm to rebuild its desk tomorrow. Waiting for a final rule before you buy hardware is reasonable. Waiting to learn whether your current agreements even allow you to hold a client asset is not. A lot of investment management agreements forbid the adviser from taking custody, full stop, because that sentence was the safe sentence in 2016. If the firm later wants the conditional path, those agreements have to be read, and some of them have to be changed, with the client’s eyes open.
Disclosure is the other quiet workstream. Form ADV brochures talk about custody in a section many clients skip and examiners do not. If self-custody of unsupported crypto becomes a real service, the brochure has to describe it without sliding into either fear or cheerleading. Say when you would do it. Say when you would stop. Say that related cash has its own account. Say that a state trust company is not “the government.” Those four sentences would already put a firm ahead of the average draft I expect to see in the first wave of updates.
Training is cheaper than a incident. The people who approve wires already know how to slow a request down. They need a version of that instinct for a wallet approval. A rushed signer is the same failure mode as a rushed wire, with a faster exit. Fifteen minutes in a staff meeting, using a fake token and a fake urgent client, will surface the gaps. I have sat in those meetings. The gaps are rarely technical. They are social. Someone does not want to annoy a senior portfolio manager. The rule cannot fix that. The culture can, if anyone bothers to practice.
A usable exception, stripped down: Find no permitted home Write it down with names and dates Hold only what that memo covers Redo the search every quarter Leave when a real custodian appears Review any state trust once a year Keep the cash leg as dull as the keys
That card is not the rule. It is a way to brief a committee without losing the plot. If a final release cannot be explained at that altitude, it is not ready for a trading desk.
The Part Worth Arguing About
Reasonable people can want this proposal to pass and still dislike pieces of it. The availability test may be too easy to satisfy with a single unreturned email. The annual trust review may be too slow for a vendor that is clearly sliding. Fund self-custody may need a brighter board role than the summary describes. Broker-dealer custody tweaks, mentioned alongside the audit updates, may matter more to daily fund operations than the crypto paragraphs that drew the posts.
Reasonable people can also want the whole thing slowed. Custody failures are not abstract. Clients have lost coins to exchanges, to bridges, to insiders, to backups that were never backups. A federal permission slip, even a conditional one, will be marketed harder than it is written. That is not a reason to keep the shrug. It is a reason to write the conditions so they cannot be skipped by a footer.
I keep landing on the same modest hope. Let advisers serve a client who wants a new asset, without inventing a custodian that does not exist yet. Make them prove the absence. Make them leave when the absence ends. Let a state trust company in only if the charter and the safeguards survive a reread. Then publish the Register date, take the comments, and sand the verbs. That is not a revolution in the 1940 Acts. It is a late repair. Late repairs are still repairs, provided someone checks the keys after the speech is over.
]]>
