The Self-Certification Gap: Citadel Securities and the Jurisdictional Vacuum Beneath Single-Stock Event Contracts

A single sentence crossed the wire. Citadel Securities urged the SEC to assert regulatory authority over event contracts tied to public companies, and in doing so, challenged the CFTC's self-certification process. No date. No dollar figures. No named contracts. One fact in, and the entire regulatory architecture shivers.

That is the signal. When one market participant's filing can rattle two federal agencies, it is not because the sentence is loud. It is because the sentence lands on a structural fault line that has been load-bearing and unexamined for fifteen years. The question is not whether Citadel is right. The question is what happens to a system when the boundary conditions of two overlapping jurisdictions are finally tested in public.

I have spent enough time inside contract bytecode to know that the most consequential bugs never live in the function bodies. They live in the interfaces — the places where two systems each assume the other is handling the edge case. Regulatory architecture has the same failure mode. The Commodity Exchange Act and the Securities Exchange Act each assume the other is catching the contracts that fall between them. Every so often a product appears that neither is actually catching, and both discover it at the same moment. Single-stock event contracts are that product. The mechanism that let them list — self-certification under CEA §5c(c) — is now the mechanism under attack. Not because it is broken. Because it works. Its heart.

The mechanism under attack

Start with the plumbing, because the politics is downstream of the plumbing.

Under Section 5c(c) of the Commodity Exchange Act, a Designated Contract Market — a DCM, the CFTC's licensed exchange category — can list a new product by self-certifying that it complies with the Act and the Commission's regulations. The exchange files the terms. The CFTC gets a review window. If the Commission does not affirmatively find a violation inside that window, the product lists, trades, and settles. No pre-approval. No permission slip. The default state is permission.

This is not a loophole in the pejorative sense. It is a deliberate design choice rooted in the CFTC's historical role as a principles-based regulator of derivatives. The agency's comparative advantage has never been adjudicating the legal character of every instrument. It has been policing price discovery, surveillance, and position limits after the fact. Self-certification externalizes the first-line legal judgment to the exchange and reserves the Commission's authority for objection.

The CEA hands the CFTC two levers to pull against a listing it dislikes. Section 5c(c)(5)(C) lets the Commission prohibit event contracts that involve unlawful activity, that constitute gambling, or that run contrary to the public interest. That third prong — public interest — is the one that has done the heavy lifting in the event-contract fights, and it is also the one with the softest edges. "Public interest" is a standard, not a rule. It invites the agency to substitute its judgment for the market's without a bright line to anchor the substitution.

Now layer the second statute on top. The Securities Exchange Act of 1934 defines a security in Section 3(a)(10), and the definition is broad enough to sweep in any "investment contract" — the term the Supreme Court operationalized in SEC v. W.J. Howey Co., 328 U.S. 293 (1946). Then Dodd-Frank added Section 3(a)(68), which carves out a category called a security-based swap: a swap that references a single security or loan, or a narrow group of them, or an event relating to a single issuer of a security, where the value depends on a financial, economic, or commercial consequence of that event.

That last clause is the whole ballgame.

The pivot everyone is missing

Here is where the coverage of this story, thin as it is, tends to stop. Citadel wants the SEC involved. The natural shorthand is that event contracts touch "investor protection," which is the SEC's remit. That framing is true and almost useless, because it does not name the mechanism.

The mechanism is the single-issuer event clause of §3(a)(68).

Read it again. A security-based swap includes a swap on an event relating to a single issuer of a security. Consider what a single-stock event contract actually is. It is a contract whose payoff is conditioned on whether a specific public company does something — misses earnings, completes a merger, replaces its chief executive, gets delisted, suffers a specific regulatory outcome. The settlement value derives from a commercial consequence experienced by one identifiable issuer. That is not adjacent to the definition of a security-based swap. That is the definition, read at face value.

If that is right, then the classification question is not "does this touch investor protection" but "does this instrument fall inside §3(a)(68)." And if it does, the jurisdictional consequence is not subtle. Section 3(a)(68) products are securities-based swaps. Securities-based swaps are regulated by the SEC and, since the post-Dodd-Frank division of labor under Title VII, the CFTC's role over them is residual at best. Two agencies cannot both be the primary regulator of the same instrument, and the statute already picked.

I want to be precise about my own confidence here. The statutory text is clear. The application of that text to a specific contract's terms is where the fight lives, because a single-stock event contract can be drafted to reference an issuer's stock price — squarely a security — or to reference a corporate event in a way that the issuer's lawyers will argue is not a "swap" at all. The drafting matters enormously. I audited the proxy pattern in 0x Protocol v2 in 2017 and watched a forty-percent gas-cost edge case get dismissed as "premature optimization" because the core team had already decided the surface reading of the standard was the real reading. Regulators make the same error. They read the label, not the bytecode. The winner of this dispute will be whoever forces a literal reading of §3(a)(68) before the other side has time to reframe around "public interest."

Self-certification as a latency-optimized settlement layer

The reason this dispute is structural rather than episodic is that self-certification is a latency-optimized mechanism. It is engineered to minimize the time between a product idea and a live market. That is its entire value proposition, and it is a genuine value proposition — in a bear market, when listing velocity itself becomes the scarce commodity, the exchange that can bring a contract to trade in weeks rather than quarters has a durable advantage.

But latency optimization always externalizes something. In smart contract design, shaving confirmation latency usually trades away finality guarantees. In regulatory design, shaving approval latency trades away the pre-listing validation that would have caught a misclassified product. Self-certification does not eliminate the classification question. It defers it. The classification question sits unresolved underneath every self-certified event contract, accumulating like unsettled state, until someone with standing — a broker-dealer, a competitor, an issuer — forces settlement.

Citadel is forcing settlement.

Look at who Citadel is. It is one of the largest market makers on the planet. It is registered as a broker-dealer and, through affiliates, as a swap dealer. Its exposure is not abstract. If a single-stock event contract is reclassified as a security or a security-based swap after listing, every market maker that provided liquidity in it may be found to have transacted in an unregistered security. That is not a fine. That is a category of liability that attaches to the act of quoting a price.

So the filing is, before anything else, defensive. Citadel would rather clarify the boundary now, while the products are small, than discover the boundary in an enforcement action against its own desk. This is the same calculus I watched play out in the DeFi Summer of 2020, when I simulated Compound Finance's interest-rate model in Python and found a theoretical liquidation cascade in the oracle pricing path. The model held in live conditions, and the response from founders was dismissive. But the institutional risk managers — the people who actually hold the tail — took it seriously, because they understood that an unpriced tail is a liability whether or not it has realized. Citadel is behaving as an institutional risk manager for its own book.

What the enforcement record actually shows

The instinct is to assume the CFTC has been crushing these products and the SEC has been passive. The reality is closer to the opposite, and that inversion is why the current moment is dangerous.

On the CFTC side, the Commission has repeatedly lost ground. Its attempt to block event contracts on congressional-control questions — the Kalshi matter — ended with a district court siding against the agency, concluding that it had exceeded its authority in rejecting the listing. The reading that circulated afterward was celebratory: courts will protect event contracts from an overreaching CFTC. But that reading commits a category error, and the category error is now the most expensive misunderstanding in the sector.

Kalshi established that the CFTC cannot block a contract merely because it disapproves of the underlying event. It did not establish that no other agency has jurisdiction. It did not establish that a single-stock event contract is not a security-based swap. It did not establish that the SEC is barred from acting. A procedural win against agency overreach was interpreted, by a market eager to interpret it that way, as a substantive win on classification. Those are different holdings, and conflating them is exactly the kind of interface assumption that produces the bug.

The precedents share a structure. Courts have been skeptical of the CFTC's power to say no on "public interest" grounds. They have not been asked, in a binding way, whether the SEC must say yes. That question — the affirmative jurisdiction of the SEC under §3(a)(68) — is the one Citadel is now teeing up. If a court answers it in the SEC's favor, the Kalshi line of cases does nothing to protect single-stock products, because those cases were about the CFTC's power to prohibit, not the SEC's power to regulate.

There is a second-order effect that almost nobody is pricing. The self-certification fight has an asymmetric blast radius. Macro event contracts — elections, central bank decisions, macroeconomic prints — sit comfortably inside the CFTC's commodity and general-event tradition. They are the product line least exposed to a §3(a)(68) reclassification. Single-issuer contracts are the most exposed. A rational platform, reading the regulatory tea leaves, does not exit the business. It reallocates the product line away from the exposed segment. The compliance-driven product restructuring is already latent in the economics, and Citadel's filing is the trigger that raises its probability.

The two-agent problem

Step back and look at the architecture, because the architecture is the story.

Since Title VII of Dodd-Frank, the United States has run a two-agent derivatives regime. Swap execution and clearing of "swaps" fall to the CFTC. "Security-based swaps" fall to the SEC. The line between them was drawn by statute and refined by rulemaking, and it works for the instruments the drafters had in mind — interest-rate swaps on one side, single-name credit default swaps on the other. Event contracts are a category the drafters did not have in mind. They emerged from a technological and cultural substrate — prediction markets, blockchain settlement, the collapse of the cost of creating a bespoke contract — that arrived years after the statute was written.

This produces what I think of as an enforcement race condition. Two agents, each with a partial mandate, each conditioned to act only when the other has not, each reluctant to be first because the first mover absorbs the political cost of constraining an innovative product. The race condition is stable as long as the volume is small. It resolves the moment volume grows enough that the loser of the boundary can no longer ignore it. We are at that moment, and Citadel — a participant whose losses scale with volume — is the agent that withdrew first.

There is a further complication the 2024 Loper Bright decision injected into the mix. That ruling, holding that the Chevron deference doctrine no longer requires courts to defer to an agency's interpretation of ambiguous statutes, removes a cushion that both agencies relied on. Under Chevron-style deference, the SEC and the CFTC could each advance plausible readings of their own jurisdiction and expect a court to go along if the statute was ambiguous. After Loper Bright, the court reads the text itself. This matters asymmetrically. The CFTC's expansion into event contracts ran substantially on the "public interest" standard — a quintessentially ambiguous, agency-defined term that Chevron once sheltered. The SEC's claim, by contrast, runs on §3(a)(68), and §3(a)(68) has specific words in it. When deference falls away, the party with the more determinate statutory text wins. That is a structural advantage for Citadel's position, whether or not Citadel framed it that way.

Compliance cost as a moat

Now the part the sector does not want to hear, and the part I would put at the center if I were writing the rule.

The immediate effect of SEC assertion over single-stock event contracts is not a ban. It is a registration requirement. An exchange listing a securities-based swap, and the brokers and dealers transacting in it, would need to come inside the SEC's regime. That means broker-dealer registration, market-surveillance systems calibrated to detect manipulation of securities, and — the expensive part — information barriers to prevent misuse of material nonpublic information about the referenced issuer.

That last requirement is where the cost curve bends. A single-stock event contract pays off on an issuer event. Anyone who knows the event is coming knows the payoff. The information-barrier burden is therefore not incidental; it is the core of the regime. And building it is not a matter of hiring a compliance officer. It requires data governance, segregation across trading and research functions, and monitoring infrastructure that can distinguish informed trading from noise in an instrument with no underlying economic exposure to hedge against.

I ran into the limits of this kind of monitoring during the eight months I spent auditing an AI-agent framework's interface with smart wallets. The finding there was a race condition that let autonomous agents bypass multi-signature requirements under specific latency conditions — a control that looked absolute in the specification and was porous at the execution layer because no one had modeled the timing. Detection of insider exploitation in event contracts has the same structure. The rules look complete on paper and leak at the boundary where human judgment about "materiality" cannot be automated. There is no mature RegTech for deciding, in real time, whether a corporate event is significant enough to have attracted informed trading. Compliance here is human-plus-machine, and it is slow and expensive.

Which brings the competitive structure into focus. Compliance cost that scales with fixed investment favors the capital-rich. A double-registered, well-surveilled platform can absorb the regime; a small venue cannot. The instinct is to read this as anti-innovation. The accurate reading is that it is anti-small. The head of the field benefits. Citadel, one of the most regulated participants in the market, is not harmed by a higher bar. It is advantaged by one — because a higher bar is a moat, and the firms already inside the wall are the ones who can afford to build it higher.

What the bulls got right

Credit where it is owed, because the celebrants of the Kalshi outcome were not entirely wrong, and the strongest version of their case deserves an honest hearing.

The strongest version is that the CFTC's power to prohibit event contracts on "public interest" grounds is genuinely dangerous and unduly vague, and that the courts were right to constrain it. "Public interest" as a prohibition standard lets an agency kill a product because it dislikes the product's subject matter, with no limiting principle. That is a real threat to prediction markets of every stripe, and the Kalshi line of cases has done the market a real service by forcing the agency to articulate something firmer than discomfort. On this point, the bulls are correct and the CFTC's litigation posture has been indefensible.

The second thing the bulls got right is subtler and more important. The market's reflexive framing — SEC bad, CFTC good — is wrong, but the underlying intuition that clarity is better than ambiguity is right, and Citadel is, however unwittingly, a clarity-maximizing force. A jurisdictional vacuum is not neutral. It is a tax on every participant, because every participant has to price the probability of reclassification, and the price of that uncertainty is borne by the smallest players with the least ability to hedge it. Someone forcing the boundary to become explicit is doing the market a service, even if the service is delivered by a firm whose motives are self-interested. The blind spot in the bulls' position is not their love of event contracts. It is their belief that the CFTC's silence is a stable equilibrium. It is not. It is an unresolved race condition, and unresolved race conditions do not stay unresolved once volume makes them expensive.

The shape of the next eighteen months

The most probable trajectory is not a ban and not a courtroom loss. It is a rulemaking. The SEC's most efficient tool is a concept release or proposed rule inviting comment on whether single-stock event contracts fall within the securities laws — a move that asserts jurisdiction, freezes the listing pipeline while the comment period runs, and imposes the political cost on the CFTC of being the agency that stranded the products. A rulemaking is slower than an enforcement action and harder to reverse, which is precisely why it is the tool of choice for a regulator who expects to win on the text.

The trigger that would accelerate it is volume. A single-stock event contract that grows large enough to attract a manipulation case — an insider with foreknowledge of an issuer event taking a position — converts an academic jurisdiction question into a live investor-protection question, and investor protection is the SEC's home turf. The second trigger is legislative. The gap in Title VII that event contracts expose is a drafting gap, and drafting gaps invite Congress. A hearing that forces both agencies to state their positions on the record would produce a public record of exactly how little agreement exists on the boundary.

The window to act is between those two triggers. For a platform, the rational move is not to litigate the boundary but to choose a regulatory home deliberately — CFTC-inside products in one line, SEC-registered products in another, and a documented classification methodology for every contract before it lists. For a market maker, the rational move is to know its exposure to each product line cold, because the reclassification, when it comes, will not be gradual. It will be a single settlement event, and the inventory held on the wrong side of it will not have a hedge.

The sector spent a decade assuming the CFTC's jurisdiction over event contracts was a fact. It was a convention. Conventions are not enforced by statute. They are enforced by the absence of a party willing to challenge them. Citadel just became that party, and the convention is now in settlement. Its heart.

Somewhere in a filing room, a contract term that references a single issuer's stock price is waiting to be read literally for the first time. When it is, the question will not be whether the product was innovative. It will be whether the agencies ever actually agreed on who owned it. They did not. The gaps in the architecture were always there. The market mistook them for floorboards. They were load-bearing. Its heart.