The Algorithm Was a Person: Inside the Linqto SPV Fraud and the Myth of Retail Access

The most dangerous lie in finance is rarely the one that takes your money. It is the one that takes your judgment first. On October 11, the SEC unsealed a complaint against two former Linqto executives, William Sarris and Joseph Endoso, alleging securities fraud. The headline figure is $430 million in sales. But the detail that should stop any engineer cold sits lower in the complaint: Linqto told retail investors its prices were set by a dynamic algorithm. The regulator alleges they were set by hand. I have spent enough hours in audit rooms to know what that sentence means. It is not a rounding error. It is a confession about what the founders believed — that the word "algorithm" had become a permission slip, a technology costume draped over an old, ordinary act of human discretion. Solitude is the only auditor that never sleeps. Somewhere in the Linqto story, nobody sat alone with the question of whether the words on the screen were true. That is the failure. Everything after it is arithmetic. To understand why this matters beyond one company, you have to understand the structure Linqto sold. The platform's core product was not a stock. It was a special purpose vehicle — an SPV — that held an interest in a private, pre-IPO company. Retail investors did not buy shares in the startup. They bought a slice of a shell that claimed to own a slice of the startup. This is the architecture that has quietly become the standard vehicle for "democratizing" private markets. The logic is seductive: venture capital has outperformed public markets for a decade, so why should the returns be reserved for institutions and the already wealthy? The SPV is the answer the industry offers. It pools small checks into something large enough to negotiate a position in a late-stage company. But an SPV is also a legal instrument, and legal instruments carry obligations. According to the SEC, Linqto's vehicles failed on several of them at once. The complaint alleges the firm sold unregistered securities, operated as an unregistered investment company, and — most seriously — misrepresented material facts to the people buying in. Among those alleged misrepresentations: that pricing was algorithmically driven when it was manual, and that certain offerings had sold out when they had not. Thousands of retail investors participated. The SEC says the company's own legal counsel warned that the business model ran afoul of the securities laws. The executives continued anyway. That single fact — the ignored warning — is the spine of the case, and it is the detail I want to spend the most time on, because it is the difference between a compliance failure and a fraud. Start with the statute, because the statute tells you how the government is thinking. The SEC is not charging one thing. It is charging three families of violations in parallel: Section 5 and Section 17(a) of the Securities Act of 1933, Section 10(b) and Rule 10b-5 of the Exchange Act of 1934, and Section 7 of the Investment Company Act of 1940. Control-person liability under Section 20(a) pulls the individuals into the frame. When a regulator stacks registration violations on top of antifraud violations on top of investment-company violations, it is not improvising. It is building a case that does not depend on winning any single argument. Registration is a strict-liability regime — you either registered or you did not. Antifraud requires intent, but it is the charge that carries the moral weight and the heaviest penalties. Investment-company status is the structural charge that says the entire vehicle was mislabeled. The pivot of any SPV case is the Howey test: is this an investment of money in a common enterprise with an expectation of profit derived from the efforts of others? For a vehicle that pools retail money to hold pre-IPO equity, the answer is almost certainly yes. And once an instrument is a security, the question is no longer whether you can sell it to the public — it is whether you did so lawfully. Here is where the retail narrative breaks. The accredited-investor line — the rule that limits private placements to people above certain income or net-worth thresholds — exists precisely because private markets are opaque. There is no audited quarterly filing, no liquid price, no continuous disclosure. The law's bargain is simple: sophisticated parties can bear information asymmetry; unsophisticated ones cannot. The SPV "democratization" story is, structurally, an attempt to route around that bargain. I want to be careful here, because there is a version of this argument that is just paternalism, and I do not accept it. Retail investors are not children. The problem is not that ordinary people are incapable of understanding a private-market bet. The problem is that the disclosure they were given was, allegedly, false. Access without truth is not democratization. It is the appearance of it — the costume without the body. Now look at the two misrepresentations the SEC chose to foreground, because they are chosen with intent. The first is the algorithm. The second is the sold-out claim. Both are what I would call "pricing theater." Pricing theater is a pattern I have watched mature over several cycles. The mechanism is always the same: take a discretionary human decision, wrap it in the language of automation, and let the reader's assumptions do the rest. Nobody audits an algorithm they believe is neutral. The word does the work of a warranty without the obligations of one. This is why the allegation is so damaging. It is not merely that a price was wrong. It is that the credibility of the price rested on a claim about how the price was made — and that claim, per the SEC, was fabricated. In a private vehicle with no market price to check against, the story of the pricing is the pricing. Remove the story and there is nothing left to trust. The "sold out" claim follows the same logic from the demand side. Scarcity is the oldest sales tool in existence. In a market where the whole promise is access to something exclusive, manufactured scarcity is not a small embellishment — it is the product. If offerings were not actually sold out, then the urgency that drove purchases was manufactured too. Then there is the fact that will decide the criminal exposure: scienter. Antifraud liability under Rule 10b-5 requires intent or at least recklessness. It is often the hardest element for the government to prove, because you cannot subpoena a state of mind. But you can subpoena a memo. The complaint alleges the company's legal counsel warned that the business model violated securities law and that the executives proceeded regardless. If that is proven, the government does not have to infer intent from circumstances. It has a document that says the defendants were told. I have been on the other side of a warning like that. In 2017 I refused to sign off on a contract audit because the encryption standards would not protect user metadata, and the founders wanted to ship anyway. I left over it. I am not telling that story to claim virtue. I am telling it because I know exactly what the pressure feels like — how easy it is to tell yourself the warning is a formality, that the lawyers are paid to worry, that the market will not wait. The decision to proceed past a legal warning is not one dramatic choice. It is a hundred small ones, each of which feels survivable. But from the government's chair, it is not a hundred small choices. It is one fact, and it converts a compliance problem into a fraud case. It also, as I read it, closes the door on the kind of deferred-prosecution or leniency path that well-counseled defendants sometimes walk through. You cannot claim good faith when your own lawyer's warning is sitting in the file. Then consider what the SEC is asking for. The requested relief is not just disgorgement and penalties. It includes an officer-and-director bar. That is the tell. A bar is not a fine you pay and move past. It is a statement that the regulator believes the conduct was severe enough to justify removing a person from the profession. Combined with the fact that the SEC named two former executives rather than only charging the entity, the message is unambiguous: this is enforcement against people, not just paperwork. The disgorgement math deserves its own paragraph, because it is where the law's limits matter. The SEC cites roughly $430 million in sales. That is a revenue figure, not a profit figure, and after the Supreme Court's decisions in Liu v. SEC and AMS v. SEC, disgorgement is cabined by equitable principles — it cannot exceed the wrongdoer's net profits and, where feasible, should be returned to investors. So the headline number and the recoverable number are not the same thing. This is why the complaint also seeks prejudgment interest: it is a way to recover value that disgorgement doctrine might otherwise leave on the table. There is one more angle the complaint does not foreground, and it is the one I would flag to any founder reading this. A platform that matches retail buyers to unregistered securities is arguably operating as an unregistered broker-dealer. That charge is not the centerpiece here, but the conduct that would support it is fully described. In the coming cycle, I expect the broker-dealer registration question to become the quiet second front in every SPV enforcement action, because it reaches the distribution channels and not just the issuer. And then there is the signal that should worry any defendant more than any of the civil counts. The SEC's San Francisco office led the action, but the U.S. Attorney's Office for the Southern District of New York and the FBI assisted. When criminal authorities are in the room, the exposure is no longer a fine. It is liberty. Civil disgorgement and criminal indictment run on separate tracks, and a defendant who settles civilly has not settled anything that matters most. I want to place this case in a longer arc, because I think the arc is the real story. We spent a decade being told that code is law. The phrase was meant as a promise of neutrality — that software would execute impartially, free of human discretion. But a promise of neutrality is only as good as the honesty of the people who invoke it. Code is law, but conscience is the interpreter. When the interpreter is corrupt, the law it renders is corrupt too, no matter how elegant the syntax. That is the inversion this case represents. Consider the Tornado Cash sanctions, which I have criticized sharply: there, the government penalized developers for code that actually worked — that did exactly what it was designed to do, transparently, on-chain, for anyone to inspect. Whatever you think of the outcome, nobody accused the code of lying. Here, the allegation runs the other way. The "code" that was supposed to set prices did not exist. The automation was a person, and the person was the sales pitch. Here is the part I think most coverage will miss, because it is uncomfortable for everyone involved. The instinctive reading of Linqto is that a few bad actors exploited a good idea. I do not think that is quite right, and the difference matters for what comes next. The loudest voice in this saga is the one promising access. And the loudest voice is rarely the most aligned. "Democratizing private markets" has become the most seductive phrase in the industry precisely because it is partly true — and partly true is the most dangerous kind. The structural problem is that the SPV model's economics depend on volume from investors who, by definition, cannot satisfy the accredited-investor tests that the law uses as a proxy for sophistication. The model does not have a compliance problem at the edges. Its growth engine sits on the far side of the line the law draws. That means the fix is not a better disclosure template. If you tighten disclosure and verify accreditation honestly, the retail volume that makes the model attractive collapses. If you preserve the volume, you have to loosen the gate, and the gate exists for a reason. This is a genuine tension, not a moral failing of one company — but Linqto is what happens when an industry pretends the tension does not exist and ships the product anyway. The fraud is the symptom. The vehicle is the disease. So watch the signals, not the headlines. Does the SEC bring a second and third SPV case? Does SDNY convert its assistance into an indictment? Does a major compliant platform with real broker-dealer registration suddenly look more valuable than its unlicensed peers? The answers will tell you whether Linqto was an outlier or the first domino. The honest question is not whether retail investors deserve access to private markets. They probably do. The question is whether anyone in this industry is willing to build that access on truth rather than theater — and whether we will keep letting the word "algorithm" stand in for the work of telling it.