Last week I sat with a fund prospectus open in one window and a decompiler in the other, and the two documents were saying the same thing in different languages. The prospectus announced that, under a new SEC proposal, retail investors could soon buy into private funds, and that managers could charge "up to 20% performance fees." The decompiler, as it happened, was chewing through an on-chain yield vault whose fee logic I had been asked to review. Same structure. Same incentive. Same shadow.
I trace the shadow before it casts. And this shadow is twenty percent wide.
Because here is the detail the headlines swallowed: under the current Rule 205-3 of the Investment Advisers Act, there is no percentage cap on performance fees at all. There is only a gate โ the client must be "qualified," meaning roughly $1.1 million in managed assets or $2.2 million in net worth. The SEC has never told a private fund manager, in writing, "you may not charge more than twenty percent." So when a proposal suddenly introduces a hard ceiling, that ceiling is not a loosening. It is the first time the regulator has reached into the fee structure itself and set a numeric limit. The twenty percent is a new constraint wearing the costume of a new freedom.
That inversion โ a rule that looks like access but behaves like control โ is the entire story. And it is the same inversion I have been auditing on-chain for six years.
Let me set the mechanics before the argument, because the mechanics are where the argument lives.
The proposal sits at the NPRM stage โ a notice of proposed rulemaking โ which means nothing is law yet, only drafted intent. Three separate rulebooks are in play. The first is the Investment Advisers Act, Section 205 and its Rule 205-3, which governs when a manager may charge performance-based compensation at all. The second is Regulation D, Rule 501(a), which defines who counts as an "accredited investor" and therefore who may buy into the private market without the protections of registration. The third is the Investment Company Act, Sections 3(c)(1) and 3(c)(7), the exemptions that let private funds exist without registering as public investment companies. Three rulebooks, one direction.
The SEC under Paul Atkins has pivoted from the Gensler-era posture of regulation-by-enforcement toward what it calls capital formation. Atkins is a known market-friendly chair; his very appointment signaled the turn. His language around the proposal โ "expand participation while guarding against fraud" โ is the tell. It is a dual mandate: widen the door at the front, keep a guard at the back. Access loosens; enforcement does not disappear, it relocates.
This matters because the market is sideways. In a consolidation regime, capital hunts for yield that public markets are not offering, and the private market โ historically walled off โ becomes the object of desire. The SEC is responding to that desire. But it is responding with a structure that quietly imports the discipline of the very thing it is trying to compete with: on-chain, permissionless, transparent yield. I have watched this movie. In 2020 I ran ten thousand simulated arbitrage attacks against Curve's stableswap invariant to prove a point about slippage, and the lesson was never that the math was beautiful. The lesson was that when you formalize an incentive, the incentive stops hiding. The SEC is now formalizing an incentive โ retail access to private funds โ and the incentive will stop hiding.
Start with the arithmetic of a performance fee, because the arithmetic is the vulnerability. A carried-interest model โ the classic "2 and 20" โ charges two percent of assets under management annually and twenty percent of profits above a hurdle. The twenty percent is not twenty percent of the fund. It is twenty percent of the gain, which, compounded over a fund's life, is a very different animal. On a fund that returns fifty percent gross, a twenty percent carry captures a fifth of the upside before the investor sees a dollar of it. The number sounds small. The number is not small. I once rebuilt the fee waterfall of a mid-tier fund as a spreadsheet and watched a manager's realized take rise from fourteen percent to twenty-six percent of investor profit simply by moving the hurdle and adding a catch-up clause. Nothing about the headline fee changed. Everything about the economics did.
Now introduce the proposal's ceiling. If the SEC caps performance fees at twenty percent, it is not capping carried interest โ carried interest already lives near twenty. It is capping the upper tail of fee structures: the thirty-percent-carry boutique, the tiered waterfall that accelerates above a threshold, the catch-up clauses that let a manager sprint to a full share of profits once a hurdle is cleared. The cap does not touch the median fee. It amputates the tail. And the tail is where the most aggressive, most retail-hostile structures live.
This is why the framing of "deregulation" is wrong on its face. A deregulation removes a gate. This proposal adds a ceiling while loosening a gate. It is a simultaneous opening and closing โ the classic move of a regulator who wants the political credit of access without the reputational risk of blowups. In protocol terms, it is a system that lifts its whitelist while introducing a hardcoded fee cap in the same upgrade. Permissionless entry, parameterized exit. I have audited that exact pattern. In 2025, co-authoring the security framework for AI agents executing on-chain transactions, we found that the dangerous upgrades were never the ones that added features. They were the ones that added features and constraints together, because the constraints masked the features in the changelog. The twenty percent is a constraint masking an access grant. Read the changelog carefully.
The accredited investor definition is where the second mechanic lives, and it is the more consequential change. The current standard is pure wealth โ a number, a threshold, a bank balance. The proposal floats an alternative: a knowledge standard, potentially satisfied by holding Series 7, 65, or 82 licenses, or by demonstrated investment experience. This quietly rewrites the regulatory DNA. A wealth standard is objective. A machine can verify it. A knowledge standard is subjective, and subjectivity is where compliance systems go to die. How do you prove knowledge? A test? A license? A signed attestation? Each answer creates a new attack surface. The moment knowledge becomes a gate, it becomes a gate to game. Knowledge, unlike money, cannot be counted โ only claimed and assessed.
This is where the on-chain lens sharpens the analysis. DeFi solved the accredited investor problem by deleting it. A permissionless vault does not ask whether you are sophisticated; it asks whether you can sign a transaction. That is the purest knowledge standard in existence โ not a credential, but a capability. The SEC, by moving toward a knowledge standard, is inching toward the on-chain model without admitting it. And the on-chain model carries a lesson for the SEC: capability-based access is elegant, but it is also adversarial. Anyone can acquire capability. Anyone can practice. The gate that admits the capable admits the adversary who is willing to become capable first. I learned this in 2021, reviewing the random seed entropy source behind a generative art collection, where the "secure" gate depended on a block hash that a patient observer could eventually predict. So the knowledge standard is not a door. It is a filter with a documented bypass.
Then there is the product structure, and here the conflict becomes structural rather than cosmetic. Private funds are built on illiquidity โ long lockups, capital calls, gates on redemption. Retail investors need the opposite. You cannot sell a seven-year lockup to someone whose emergency fund lives in a checking account. The proposal implicitly forces the creation of what I would call half-liquid products: interval funds, tender-offer funds, structures that offer periodic redemption windows rather than daily liquidity.
I have seen this structure before. It is the on-chain vault with a withdrawal queue. In 2022, reverse-engineering the Terra/UST collapse, I built a model showing how a lopsided incentive structure made the system fragile regardless of sentiment. The fragility was not in the peg. It was in the redemption queue โ the assumption that not everyone would exit at once. Interval funds make the same assumption. They are redemption queues with a compliance department. And redemption queues fail in exactly one direction: when everyone arrives at the window at the same time, which is precisely when liquidity matters most. The half-liquid product is the same bet as the on-chain vault: it works until it doesn't, and it doesn't exactly when you need it to.
Compliance cost is the part the enthusiasm skips. Serving retail investors is not serving smaller institutional investors. It is serving a different species. An institutional LP with fifty million dollars does not file a suitability complaint; a retail investor with fifty thousand does. The obligations that attach to retail โ suitability assessment, accredited-status verification, plain-language disclosure, complaint handling โ scale with the number of investors, not the size of the assets. A fund that adds ten thousand retail accounts adds ten thousand complaint surfaces. Do the math. A single institutional relationship might carry a compliance cost of a few thousand dollars a year. A single retail relationship might carry a compliance cost that is proportionally enormous โ KYC, suitability, disclosure, servicing โ against an account one-hundredth the size. The unit economics of retail private funds invert the unit economics of institutional private funds. This is not a rounding error. It is the whole business case, and it favors scale. Only managers large enough to amortize the compliance cost across billions will find retail profitable. The proposal, marketed as broadening access, structurally concentrates the industry. I have watched this concentration logic before, and it has a name in security work: the economies-of-scale attack. The vulnerability is not in the code. It is in the deployment โ only the largest operators can afford to run it safely, so the smallest operators run it unsafely, and the unsafe runs become the incidents.
The fee disclosure is the final mechanic, and here the proposal does something almost nobody has priced. If performance fees are capped and regulated, they must be disclosed โ how the twenty percent is computed, on what base, before or after expenses, gross or net of the tax gross-up on carried interest. Disclosure of a fee structure is disclosure of a business model. For managers who have treated their fee waterfalls as trade secrets โ and the best ones do โ this is an involuntary open-sourcing of their economics. In the void, the bytes whisper truth, and the truth here is that a regulator who makes you publish your fee math has taken a piece of your intellectual property without calling it that. The bug hides in the beauty. The beautiful, simple, investor-friendly fee cap carries a hidden payload: the forced transparency of the most closely guarded numbers in private finance.
Now the counter-intuitive angle, the one the consensus misses.
Everyone is debating whether the proposal will pass. Almost no one is asking what happens to enforcement when it does โ and enforcement is where the real cost lands. The consensus assumes that loosening access loosens oversight. The opposite is true. The moment retail investors are inside private funds, the SEC's enforcement posture must shift from gatekeeping to guardianship. It can no longer say "you shouldn't have been in this fund." It must say "the manager should have protected you." That is a change in the burden of proof, and it tilts toward the investor.
Here is the blind spot. In institutional disputes, sophisticated parties bear their own risk. In retail disputes, the regulator presumes the manager's responsibility. When a retail investor loses money in a private fund and there is any flaw in the suitability file, the SEC does not investigate whether the investor understood the risk. It investigates whether the manager documented that they did. Suitability, in the retail context, is not a check โ it is a chain of custody for consent. And a chain of custody is only as strong as its weakest link, which is always the human who filled out the form.
This connects to a piece of litigation the industry has not yet absorbed: the Supreme Court's 2024 decision in SEC v. Jarkesy, which limited the SEC's ability to use in-house administrative courts for fraud claims and pushed some cases toward federal jury trials. The practical effect is that retail-fraud enforcement becomes more expensive and more procedural. Which means the SEC will pick its retail cases carefully โ and the cases it picks will become the templates for the whole industry. One bad suitability file, litigated in federal court, becomes the standard every manager must meet. The regulator will not define the standard through rulemaking. It will define it through the first jury verdict, and the industry will reverse-engineer its compliance program from a trial transcript.
The deepest blind spot, though, is data. Retail access means retail KYC. Retail KYC means a flood of personal information flowing into entities that previously held almost none โ institutions, not individuals. A private fund manager who has never touched consumer data becomes, overnight, a controller of personal information, subject to state privacy statutes and, for non-US investors, the GDPR. This is a compliance obligation the proposal never mentions and the industry has not staffed. It is a derived vulnerability โ not in the fee cap, not in the access gate, but in the exhaust of the whole system. Vulnerability is just a question unasked, and no one has asked the data question. Finding the pulse in the static means listening for the obligations that no one put in the press release.
There is a cross-border dimension too, and it cuts against the proposal's own logic. If the United States loosens retail access to private funds while the European Union's AIFMD and the United Kingdom's FCA framework retain strict retail restrictions, capital has a new reason to route toward American vehicles. That is regulatory arbitrage, and it invites a reverse measure โ ESMA or the FCA tightening the treatment of US-domiciled funds sold into their markets. The proposal's authors frame it as domestic. It is not. A rule that changes who may buy a fund changes where the fund is built, and the map redraws itself around the money.
So here is the forecast. Over the next eighteen to twenty-four months, the proposal will move through comment, and the comment file will fill with objections from state regulators and consumer advocates who see a knowledge standard as an erosion of their authority. The final rule will likely land in a hedged form: a retained wealth threshold, an optional knowledge path, a fee cap softer than it sounds. The cap will be the headline. The knowledge standard will be the substance. The data obligations will be the bill nobody saw coming.
And the industry will discover that it has been handed the on-chain playbook without the on-chain transparency. Permissionless entry. Parameterized fees. Redemption queues. Suitability as chain of custody. The private fund is becoming a vault โ but a vault you cannot inspect, because the fee math is disclosed only to the regulator and the suitability file is disclosed only to the court.
The question is not whether retail access arrives. It is whether the access arrives with the one property that made on-chain systems auditable in the first place: the willingness to let anyone read the rules. Logic blooms where silence meets code โ and right now, the silence is the proposal itself, and the code is still unwritten.