CME chair's staff-size defence exposes the self-certification gap

Terry Duffy's answer to Luana Lopes Lara on Thursday was, by any measure, the most clarifying moment of the CFTC's first Innovation Advisory Committee meeting. She had asked whether CME had faced problems with market manipulation. He told her he employs more people in his regulatory department than she has in her entire company. She told him he might want to learn something about efficiency. The room laughed. The problem the exchange revealed was not funny.

Twenty-five hundred event contracts have been self-certified since January 2025. Duffy said none were opposed by the CFTC. That number, uncontested in the room, is the load-bearing fact in this story, and it is not the number Selig wants discussed.

The CFTC chairman spent most of his prepared remarks casting the agency as a bulwark against overreaching state attorneys general — New York's Letitia James in particular, whose office is seeking thirty-six billion dollars in fines and restitution from Kalshi. Selig's framing is defensible as a jurisdictional argument. Federal preemption over designated contract markets has a legal foundation, and the more than twenty active lawsuits between state regulators and prediction market operators will eventually produce a cleaner answer on where the line sits. But the James lawsuit is also, at its most basic level, a consumer protection complaint — age restrictions, unlicensed operation, retail exposure. Selig's response to that complaint was a three-part regulatory roadmap that does not yet have a second meeting scheduled to advance it.

Lopes Lara's position is that federal oversight produces stronger consumer protections than a patchwork of state rules. She said she had never heard a single argument for why state-by-state is better. That is a clean line, and she may be right in principle. The difficulty is that the argument for federal superiority requires the federal regulator to actually regulate. A process that has approved twenty-five hundred contracts without opposing one is not a framework that can carry that argument convincingly against a state AG who has found a live case.

Duffy flagged three specific contracts he believes violate CFTC Core Principle 3 — the Maduro ouster contract, several sports-related derivatives, and a trade connected to Gabriel Perez, the former Trump teleprompter operator whose mention-market activity is now under active investigation. Selig responded that the products were not listed in the US, a correction his own agency subsequently had to walk back for at least one of the three. That error, small in isolation, matters here because it happened on the record, in the room where Selig is supposed to be the authority on what the agency has and has not cleared.

The self-certification mechanism was designed for speed. Lopes Lara is right that a prediction market without the ability to list contracts quickly is a prediction market that cannot function. Duffy's implied remedy — a larger compliance operation, presumably resembling CME's — misses the point about what prediction markets are built to do. But the answer to that tension is not zero opposition to twenty-five hundred contracts. The answer is a designed triage process that trades some speed for defensibility, which is precisely what Selig's roadmap gestures at without committing to a timeline.

What I have seen in markets that operate on self-certification without effective review is not fraud, usually. It is drift — the gradual expansion of what counts as acceptable until the first visible failure resets everything at cost. The Gabriel Perez investigation is not that failure yet. It is the signal that the drift has begun.