Developing The CFTC Did Not Answer the Bingo Problem. It Exempted Bingo

Editor's note: TrueEdge builds odds tools and earns affiliate commissions from licensed sportsbooks, including two named in this piece. This piece argues a position on prediction-market regulation. Weigh it accordingly.

On Friday afternoon the Commodity Futures Trading Commission issued two documents about the same question and gave two answers. A proposed rule, RIN 3038-AF82, would amend the Commission's regulations to say that event contracts — "including those based on sports-, politics-, cultural-, and weather-related events" — are swaps. An interim final rule, RIN 3038-AF81, effective the moment it hits the Federal Register, says that a wager on the same outcome placed at a state-licensed sportsbook is not.

Both rules were at the White House for under two weeks. Both are about the Supreme Court. TD Cowen's Jaret Seiberg said the quiet part out loud, calling the interim final rule a document "designed to improve the agency's position in court." Kalshi's response in Flaherty v. KalshiEX, LLC, No. 26-299, is due November 9. The comment windows on both rules close inside thirty days of publication. The sequencing is not subtle.

We read both filings rather than the press releases, and the interesting thing is not the headline. It is where the line got drawn, and what the Commission concedes about how long that line will hold.

The line is plumbing

Two circuits have now said sports event contracts are not swaps, and both got there the same way. The Ninth Circuit, in KalshiEX, LLC v. Assad on August 28, worried that a plain-text reading would mean "bingo games and pingpong matches" fall under CFTC jurisdiction, and that "every person placing a sports bet at Caesar's Sportsbook (or anywhere else for that matter) is violating the CEA." The Sixth Circuit, in KalshiEX LLC v. Schuler on September 25, put it more starkly: on the government's reading, "every sports wager placed in a casino, on an online sports book, or between two friends at a bar" would have to "take place on federal markets," and gambling off a designated contract market would become a felony. The court said it was reading the statute narrowly to avoid "attach[ing] criminal penalties to a breathtaking amount of commonplace gambling activity."

That is the bingo problem, and it is a good argument. The interim final rule does not refute it. It legislates around it.

The exclusion has two parts, and neither is about what the bet is. The Provider Test asks who is offering it: the counterparty must be licensed under state law or by a tribe under the Indian Gaming Regulatory Act, must be acting within that licence, and the transaction must be regulated as gaming under that law. The Product Test asks how it trades: the transaction must not be "traded on a board of trade, an organized exchange, a swap execution facility, or any other market in which trades are executed multilaterally or subject to individual negotiation."

So the Chiefs at minus-130 is a swap or not a swap depending on whether a house set the price as principal or an order book discovered it. That is a real distinction — the CFTC builds it carefully out of the 2012 insurance carve-out, out of CEA Core Principle 9's requirement that a designated contract market run "a competitive, open, and efficient market," and out of the clearing requirement at CEA §2(h). Gambling products are not cleared; a wager obligation is not novated to a central counterparty. Nobody is pretending otherwise.

It is still plumbing. The economic exposure is identical and the Commission says so in terms: "a person can achieve similar economic exposure by either entering into a gambling wager or purchasing an event contract. However, that does not make the two products one and the same."

The proposal proves too much

Read the notice of proposed rulemaking and the restraint evaporates.

To establish that sports outcomes carry a "potential financial, economic, or commercial consequence" under prong (ii) of the swap definition, the Commission assembles a list. The Knicks' 2026 championship run generated roughly $380 million for New York City's economy, per the city's economic development corporation. Madison Square Garden Sports doubled in market value over a year. The Bucks' 2021 playoff rounds were worth $2.7m, $12.8m, $14.1m and $28m to Milwaukee. Fine.

Then it keeps going. A close game draws better television ratings than a blowout, "and thus, the spread of a game is associated with potential economic consequences." Player contracts carry per-touchdown incentives, so individual performance has economic consequences. A furniture store in Houston refunds purchases if the Astros win the World Series. A New York bar covered tabs when the Knicks won a Finals game and used Kalshi to hedge the promotion. Chick-fil-A gives Reds ticketholders a free sandwich when Cincinnati homers in the fourth. Caribou Coffee runs a buy-one-get-one when the Vikings intercept a pass. Delta awards one loyalty mile per Seahawks passing yard.

"The list of examples could extend indefinitely," the Commission writes. That is the problem, not the proof. If a coffee promotion supplies the economic consequence that converts an event into a commodity, then the fourth-quarter interception is a commodity, the free sandwich makes the fourth-inning home run a commodity, and there is no outcome inside any sporting event that fails the test. Which is exactly what the Ninth Circuit said would happen. The CFTC's answer is not that the logic stops somewhere. Its answer is the interim final rule: the logic does not stop, so we are carving out the casino by regulation.

Prong (iv) is stranger still. The statute includes anything "commonly known to the trade as a swap." The Commission reads "the trade" to mean derivatives-market participants — "including the CFTC, its registrants, and traders, rather than some broader group, such as the general public." The regulator has placed itself inside the definitional community whose usage determines the scope of its own jurisdiction. It is not a frivolous reading; prong (iv) has to mean something, and the proposal marshals real evidence, including CME's own comment letter calling event contracts and swaps "not new," Cboe's statement that they "generally fall within the CEA's broad definition of 'swap,'" and thousands of self-certified contracts. But a test that counts the agency's vote is a test that will always come out one way.

The overlap the rule says does not exist

Here is the sentence that will be quoted back at the Commission. In the interim final rule, explaining why the Product Test is needed at all:

"Although there is not currently a practical overlap between the casino-style gambling products offered by State and Tribe-regulated entities (i.e., bilateral and non-tradable) and swaps offered by CFTC-registered entities (i.e., multilateral and tradable), the Product Test is designed to nonetheless acknowledge the functional distinction..."

There is already a practical overlap, and it is the largest sportsbook in the country. DraftKings holds state gaming licences and runs DraftKings Predictions, which posted a record $218 million of volume on a single NFL Sunday — up roughly 50% from earlier in the season. Bank of America upgraded the stock to buy on October 6 on the strength of it, modelling about $400 million in prediction-market fees next year plus $200m to $400m from market making. Jason Robins told investors the market-making arm is "one of the fastest to profitability business lines we've ever launched." Flutter is doing a gentler version of the same thing on FanDuel's side.

So one company, on one Sunday, sells you the Chiefs two ways. Through the sportsbook, where it sets the price as principal, the transaction is excluded from the swap definition by a rule that took effect on publication. Through Predictions, where it is a market maker on somebody else's order book, the same exposure is a federally regulated derivative under a rule that has not been finalised. The Provider Test sorts by institution. The institutions are already on both sides.

The Commission half-sees this. The Product Test, it says, "ensures that State and Tribal-regulated entities do not, in the future, begin offering products properly within the Commission's jurisdiction." That is an anti-circumvention clause, which is an admission that the circumvention is coming.

The strongest case against us

The best version of the other side is not that the distinction is fake. It is that market structure is the only distinction that has ever done any work in this statute, and we should be glad someone finally wrote it down.

Insurance pays out on the occurrence of an event associated with a financial consequence. It is not a swap, and nobody finds that absurd, because insurance is not traded on an organised market and is supervised by state insurance commissioners. The CFTC and SEC said so jointly in 2012. The gambling carve-out is built on the identical two-part structure, down to the phrasing. That is not improvisation; it is precedent applied.

And there is one consumer-protection fact buried in a footnote that the states' briefs do not engage with at all. A sportsbook may lawfully limit or refuse a winning customer. A designated contract market may not: 17 CFR 38.151 imposes an impartial-access requirement, and a DCM cannot bar a trader for being profitable. If you think the sharpest practice in American sportsbetting is the quiet limiting of winners — and we do — then the exchange model is the one that forbids it. The structural distinction the CFTC is drawing has a consumer-facing edge that runs against the states.

We take that seriously and still land where we land. The tradability line is coherent as applied to a 2024 market in which Kalshi ran an order book and DraftKings ran a book. It is not coherent as applied to a 2027 market in which DraftKings runs both and is the market maker on its own venue. When the operator of the exchange is also the dominant liquidity provider, "price discovery through the competitive interaction of multiple buyers and sellers" starts to describe a house with extra steps. The Commission asks for comment on exactly this — question (3) invites "better ways to draw an objective, functional distinction." It is the right question and the rule does not answer it.

What would falsify this

If the final rule arrives with a test keyed to something the operator cannot restructure its way around — a hard separation between venue and market maker, say, or a concentration limit on liquidity provision — the criticism here dissolves. We would say so.

Two procedural notes we cannot resolve today. The CFTC invoked Administrative Procedure Act good cause under 5 U.S.C. 553(b)(B) to make the exclusion effective immediately, resting partly on Pew's finding of $24 billion in April 2026 volume across two prediction markets and on the prospect of states closing those markets state by state. Courts read good cause narrowly, and "the litigation is going badly" is not a traditional emergency. We do not know how that lands.

And a small thing that is not small. Footnote 38 of the proposal states that "the SEC joins in proposing this rule in satisfaction of the statutory requirement in Section 712(d)(1)" — the Dodd-Frank provision requiring the two agencies to further define "swap" jointly. The SEC issued no release on October 9 and is not named in the CFTC's. When the two agencies went out jointly on derivatives definitions on June 18, the SEC put its own number on it, 2026-57, with a quote from Chairman Paul Atkins. This time the joinder appears in a footnote. We checked the SEC's newsroom for October 2026 and there is no corresponding release; the six the agency put out between October 1 and October 9 concern cross trading, crypto custody, an exam handbook, World Investor Week and an enforcement judgment. That may be entirely regular — the proposal says the SEC joined, and we have no reason to doubt it. But we have not seen the SEC say so in its own voice, and we are not going to assert a defect we cannot document.

What we will say is that a rule written to settle a question the Supreme Court has been asked to decide is a rule written for an audience of nine, and that the audience will read the footnotes too.