What September 15 actually decides, and what it doesn't
I'm Theresa Gosiengfiao. I build financial infrastructure, and for the past year that has meant building RiverForge: compliant DeFi rails for institutional clients — registered investment advisers, funds, the people who have a duty of care attached to every decision they make.
Before RiverForge I spent most of two decades in data infrastructure. SQL Server, ETL pipelines, reporting systems — the unglamorous machinery underneath other people's dashboards. That is a strange background for a crypto letter and it is exactly why I'm writing one. When you have spent years being the person who finds out what a system actually does, as opposed to what its documentation claims, you develop a specific allergy. Most crypto writing triggers it.
I am not an investment adviser, and this letter will never tell you what to buy. I want to be direct about that, because the distinction is the whole design. I build the plumbing. You decide whether to run water through it. Those are different jobs, and the second one is harder — you carry the fiduciary duty, the client conversation, and the examination file. What I can offer is an unusually clear view of how the machinery works, from someone who has to make it work.
What this letter is. Every other Wednesday, one piece of on-chain financial infrastructure, explained at the level of detail a compliance officer would need. How it works. What it costs. What breaks it, and what happens to client assets when it does. Sourced to primary documents — statutory text, filings, no-action letters, audits — with the link, so you can check my reading against your own.
What it isn't. No price commentary and no market calls. No recommendations of any asset, protocol, or product. No news roundup; you have five of those already and they aren't helping. And nothing tailored to any individual reader's situation — I'll explain a mechanism to anyone who writes in, but I won't tell you what to do with your book.
Why it exists. In a 21Shares and FUSE study of 350+ advisers released this May, 84% said the digital-asset education material available to them is insufficient. That finding should be embarrassing to my industry, not yours. The people building this infrastructure have spent fifteen years talking to each other in a private dialect and then acting surprised when fiduciaries won't touch the results. That's a writing failure, and it's a fixable one.
RiverForge sells infrastructure into this market, which means I have a commercial interest in you eventually being comfortable here. You should read everything below knowing that. My defense is that this letter is more useful to you if it's honest about what doesn't work, and I intend to prove that over twenty-four issues rather than assert it in issue one.
Let's start with the thing everyone is watching.
On Tuesday, September 15, at 2:15 p.m. Eastern, the Senate votes on cloture for the motion to proceed to H.R. 3633 — the Digital Asset Market Clarity Act. Majority Leader Thune filed the cloture motion on August 8, just before recess, and it ripened over the break. Patrick Witt, executive director of the President's Council of Advisers for Digital Assets, has framed the date as effectively decisive — the administration's read being that a bill that can't clear this gate before the calendar closes in won't clear it at all.
The trade press is treating the date as the moment U.S. crypto regulation gets settled. Advisers I talk to have absorbed that framing and are waiting on the outcome before doing anything.
The framing is wrong in two separate ways, and both of them matter to how you spend the next twelve months.
First: September 15 is not a vote on the bill.
It is a cloture vote on the motion to proceed — a vote about whether to begin debating whether to take up the legislation. It requires 60 votes. Republicans hold 53 seats, so at least seven Democrats or independents must cross over, and more if any Republican is absent or defects.
If cloture succeeds, here is what remains: up to 30 hours of debate on the motion to proceed, then a vote on the motion itself, then a second 60-vote cloture motion to close debate on the bill, then final passage. There is also the Senate Agriculture Committee's Digital Commodity Intermediaries Act (S. 3755), which advanced out of that committee on January 29 on a 12–11 party-line vote and has to be squared with the Banking Committee's product — analysts disagree about whether that happens before or during floor consideration, but it has to happen. Then reconciliation with the House text. Then a signature.
September 15 is the first of roughly six gates.
Prediction markets are considerably more skeptical than the trade press. As of September 3, Polymarket's contract on H.R. 3633 being signed into law in 2026 sat around 16%, and Kalshi's comparable market on a qualifying crypto market-structure bill by January 2027 near 20%. Both move fast — Polymarket traded as high as the low 80s earlier this year — so treat those as a snapshot, not a forecast.
The sticking points are not technical. They are stablecoin rewards, where banking groups argue that activity-based rewards are deposit interest wearing a costume; an enforceable ethics provision barring senior officials and their spouses from issuing or sponsoring digital assets; illicit-finance language; and the CFTC's capacity to absorb the authority the bill hands it, given commissioner vacancies and a staff that has shrunk by roughly a fifth.
None of these is close to resolved. And the two Democrats who voted the bill out of Senate Banking 15–9 in May said so on the record at the time. Senator Gallego: "my vote here does not guarantee a vote on the floor." Senator Alsobrooks: "my vote today is a vote to keep working in good faith. It does not mean that I'll be voting for the passage of the clarity act on the floor."
Second, and more important: even total success changes nothing next year.
This is the part that gets lost, and it should govern your planning. The CLARITY Act does not switch on when signed. From the House Financial Services section-by-section:
That trailing clause is the one to internalize, because the statute hands the SEC and CFTC a large volume of joint rulemaking: key term definitions, delisting procedures, mixed digital asset transactions, portfolio margining relief, issuer disclosure standards, insider reporting, custodian requirements. Joint rulemakings between two agencies with different cultures and a history of jurisdictional friction do not finish early.
Run the arithmetic optimistically anyway. Signed in December 2026 puts Title IV live around September 2027 and Titles II and III around December 2027 — and that second date floats later by however long the rulemaking slips. My own read, and I'll flag it as inference rather than fact, is that the year your operating environment actually changes is 2028.
Third: DeFi specifically is deferred, not decided.
If you take one thing from this issue, take this. The bill does exempt a set of non-custodial activities from registration — validating transactions, publishing or updating software, developing wallets, providing user interfaces, and developing and publishing a blockchain system itself. The SEC retains anti-fraud enforcement over that activity (Sec. 309); the CFTC retains anti-fraud and anti-manipulation authority (Sec. 409).
But the bill also orders a joint CFTC/SEC/Treasury study of decentralized finance — its size, scope, benefits, risks, integration with traditional markets, and illicit-activity exposure — with a report to Congress due one year after enactment (the DeFi study provision in Title V).
Congress ordering a study is what Congress does when it has not reached agreement. The definitional question that matters most to you — when a decentralized protocol becomes an "exchange" requiring registration — is among the most contested items in the Senate negotiation and is not resolved by this text.
Nor are the texts themselves reconciled. The House version uses a "mature blockchain system" test built around a 20% voting-power threshold; the Senate Banking draft uses a common-control standard and leaves the details to SEC rulemaking. Those are different tests that will classify different assets differently. Anything built against the House definitions is built against language two chambers have not agreed to.
What has already changed, without any legislation at all.
Here is the asymmetry worth noticing. While the market waits on a bill that will not bind anyone before 2028, the development that actually opened a path for RIAs happened last year and required no act of Congress.
On September 30, 2025, SEC staff issued a no-action letter permitting state-chartered trust companies to serve as qualified custodians of crypto assets for registered investment advisers and registered funds. That is narrower than it sounds and narrower than a statute — it is staff-level relief addressing one specific category of custodian, it can be withdrawn, and Commissioner Crenshaw published a pointed dissent arguing the staff had poked holes in the custody rule without the rulemaking that should accompany it. Her objection is worth reading precisely because it identifies where the relief is thin.
But it is operative today, it speaks directly to the ambiguity that has made compliance officers say no, and it does not depend on anything happening on September 15.
Three general observations, none of which turn on the vote. As always, none of this is advice about your particular clients or firm.
1. The custody question is answerable on today's framework. If you are evaluating a custodian, the diligence is the same diligence you'd run on any counterparty, and it can be run now: which specific assets and chains are supported; how keys are generated, held, and recovered; what the insurance actually covers and what the sub-limits are; whether client assets are segregated or held omnibus; and what happens to client assets in the custodian's own insolvency. That last question is the one most often skipped and the one that matters most when it matters at all.
2. Firm policy is a document, not a statute. In the same 21Shares study, 46% of advisers reported operating under restrictive firm policy — but the split by channel is stark: 14% at RIAs against 48% at wirehouses. At an independent RIA, the binding constraint is frequently that nobody has written the document yet, rather than that anyone has decided against it. An IPS section, a suitability standard, and a documented rationale do not require enabling legislation. I'll work through actual drafting language in a later issue.
3. There is asymmetric risk in restructuring around unreconciled text. The House and Senate versions classify assets by different tests. Operational changes built to one of them may need unwinding — which is a different proposition from reading the bill closely and knowing what you'd do under each version.
And on the 15th itself: the pass/fail outcome is the least informative thing that will happen. Watch instead whether the Sec. 309 and 409 exemption language survives amendment, and how the Senate Agriculture product gets folded in. Those two shape what the framework actually looks like in 2028. The cloture count only tells you the calendar.
Section-by-Section Summary, Digital Asset Market Clarity Act of 2025 — House Financial Services Committee, July 10, 2025 (PDF).
A short document. If you read nothing else, read the effective dates at Sections 206, 313, and 414, and the DeFi provisions — the non-custodial exemptions at Sections 309 and 409, and the DeFi study in Title V. Those provisions tell you more about your real timeline than any amount of coverage of the vote will.
Worth reading alongside it: Commissioner Crenshaw's dissent on the September 2025 custody relief. Dissents tell you where a framework is load-bearing and where it isn't.
Figures and prediction-market prices in this issue are current as of September 3, 2026.
RiverForge Research is published on a regular schedule for a general audience. It is not investment, legal, or tax advice, contains no recommendation to buy or sell any asset, and is not tailored to any reader's circumstances. Theresa Gosiengfiao is the founder of RiverForge, which builds digital-asset infrastructure for institutional clients and may have a commercial interest in topics discussed. No compensation is received from any issuer, protocol, or platform mentioned. Reply with questions about how something works — but I can't advise on your specific book or clients.