The Provision Both Sides Avoided Is Now the Agency's Argument

On 28 August 2026 a panel of the United States Court of Appeals for the Ninth Circuit decided KalshiEX, LLC v. Assad, No. 25-7516. Most of the attention went to the holding on the product. The part worth reading twice is shorter and sits at the end: a separate concurrence in which Circuit Judge Kenneth K. Lee identified a statutory provision that cut against the result, explained why neither side had pressed it, and then declined to resolve it because a Commission regulation disposed of the appeal anyway.

Forty-two days later the Commodity Futures Trading Commission published two rulemaking documents. One of them cites that concurrence by name, for a proposition close to the opposite of the one the judge was entertaining, and states in terms that if the panel's reasoning rested on a Commission regulation, the proposal would remove the basis for it. This article is not about who is right. It is about what the sequence teaches anyone running a venue: a legal position is only as stable as the instrument underneath it, and the instruments move on calendars that belong to other people.

The provision that gave one judge pause

The provision is the Special Rule for review and approval of event contracts, codified at 7 U.S.C. section 7a-2(c)(5)(C). Its first clause is permissive. In connection with the listing of agreements, contracts, transactions or swaps in excluded commodities, the Commission may determine that such agreements, contracts or transactions are contrary to the public interest if they involve one of six listed things, among them activity unlawful under Federal or State law, terrorism, assassination, war and gaming. The second clause then prohibits anything the Commission has so determined from being listed or cleared.

Judge Lee read the permissive verb as doing real work. In the opinion as published by the Ninth Circuit he wrote that the statute "appears to give the CFTC discretion whether to ban altogether gaming contracts" and that it "does not seem to categorically bar all gaming contracts, despite the text and contextual clues suggesting otherwise." The consequence he drew from that is the interesting one. If the ban is discretionary rather than automatic, then, in his words, "perhaps some unique sports events can be part of a swap trade if they meet the statutory requirements." That is a route to the exchange winning on the product, and it was sitting in the statute the whole time.

Both sides had a reason to leave it alone

He also explained why it had gone unargued, and the explanation is more instructive than the point itself. Kalshi had not relied on the Special Rule because the Commission, acting under that same provision, had already issued a rule banning gaming contracts. Nevada had been quiet because the provision complicates the state's own statutory analysis. One side could not use the door because the agency had closed it; the other side did not want to point at a door at all.

So the judge parked it. He wrote that the court did not need to resolve "this thorny statutory interpretation question" in that appeal "because 17 C.F.R. section 40.11 bars gaming contracts," adding that "while CFTC has proposed revising that regulation, it remains in the books and controls the outcome of this appeal." That sentence is doing two jobs. It decides the case, and it records, in the opinion itself, that the thing deciding the case is under revision.

The regulation that decided the appeal has not changed a word

As of the text of 17 CFR 40.11 published in the electronic Code of Federal Regulations, current through 8 October 2026, the rule reads as it did before any of this started. Paragraph (a)(1) prohibits a registered entity from listing for trading or accepting for clearing any agreement, contract, transaction or swap based upon an excluded commodity that involves, relates to or references terrorism, assassination, war, gaming, or activity unlawful under any State or Federal law. Paragraph (b) is still marked as reserved. Paragraph (c) still carries the ninety day review. Nothing in the rule has been amended, suspended or stayed.

For a venue, that is the first useful distinction. A regulation in force is the most durable kind of legal fact available, because it changes only by a published amendment with a date on it. We have written before about the twelve factors the Commission proposed for that determination and about the procedural half of the same proposal. Both of those pieces described a rule that was going to move. Six weeks after a federal appellate judge noted the same thing, it had not moved.

The proposal he mentioned stopped moving in July

The revision Judge Lee referred to is the rulemaking the Commission published in the Federal Register on 12 June 2026 under the title Prediction Markets; Public Interest Determinations, RIN 3038-AF65. Its own abstract describes amendments that would specify the types of event contracts subject to a public interest determination, set out the factors the Commission would apply, reorganise the procedure, and add a definition of the term gaming together with a rule on when an event contract "involves" an underlying activity.

Its comment period closed on 27 July 2026. As of 11 October 2026, no final rule has issued. That is seventy six days after the record closed and a hundred and twenty one days after publication. The proposal is neither dead nor in force. It is a document that the Commission may adopt, modify or abandon, and a venue reading it cannot tell which, because nothing in a pending proposal carries a commitment about timing.

The instrument that did arrive never mentions that rule

On 9 October 2026 the Commission did act, and it acted somewhere else. Press release 9309-26 from the CFTC announced an interim final rule excluding certain activity from the definition of a swap. Chairman Michael S. Selig is quoted in it saying that "casino-style gambling products are not derivatives." The release states that the rule "will be effective immediately upon publication in the Federal Register," with comments due within thirty days of that publication.

Read the text of the interim final rule as approved by the Commission and two things stand out. First, the only amendment it makes is to section 1.3, where it adds a new paragraph (7) to the definition of the term swap, excluding agreements offered by an operator licensed under State law or under the Indian Gaming Regulatory Act, acting within the scope of that licence, where the product is not traded on a board of trade, an organised exchange, a swap execution facility or any other multilateral market. Second, the string 40.11 does not appear anywhere in the document. The instrument that would reshape the definitional perimeter leaves the rule that decided Assad entirely untouched.

There is a third point, and it is a matter of dates rather than text. Effectiveness is tied to publication in the Federal Register. As of 11 October 2026, the Federal Register's own index of Commission documents shows nothing of the kind: the agency's two most recent entries are an information collection notice published on 6 October and a crypto asset proposal scheduled for 13 October. The exclusion exists as an approved Commission document and is not yet law.

A footnote answers the concurrence by name

The companion document is the one that closes the loop. Press release 9310-26 from the CFTC announced a notice of proposed rulemaking, RIN 3038-AF82, that would expressly define the term swap to include event contracts based on sports, politics, cultural and weather related events. In the proposal as approved by the Commission, a passage deals head on with the Ninth Circuit's use of the regulation. The Commission observes that Rule 40.11 gives no basis for concluding that event contracts are not swaps, and adds that "to the extent that the Ninth Circuit panel's reasoning was based on a CFTC regulation, this Proposal would serve to clarify by federal regulation that event contracts are swaps."

The sentence after that is the one a venue should sit with: "Accordingly, there would no longer be a valid basis to interpret CFTC regulations to conclude that event contracts are not swaps." The supporting footnote cites Assad at the concurrence, with the parenthetical "Lee, J., concurring," alongside the Special Rule itself. The provision that neither party wanted to argue in August is, in October, part of the agency's affirmative case, and the judge who flagged it is cited in support of a reading he had not reached.

One subsection, three readings, six weeks

Line up what three institutions did with the same subsection and the picture is not a dispute about facts. It is three serviceable readings of one sentence.

  • Ninth Circuit, 28 August 2026. The permissive verb suggests the Commission has discretion, not a categorical bar, so the category might survive on the merits. Question reserved.
  • Sixth Circuit, 25 September 2026. In the opinion in KalshiEX LLC v. Schuler, Nos. 26-3196 and 26-5235, as published by govinfo, the panel went the other way on the inference. Because the Special Rule reaches "agreements, contracts, transactions, or swaps," the fact that the Commission may prohibit gaming related event contracts "does not mean Kalshi's sports-event contracts are necessarily swaps." Breadth cuts against the exchange.
  • The Commission, 9 October 2026. The Special Rule is affirmative textual evidence that event contracts are swaps, because the word swap in it would otherwise be surplusage, and because Congress gave the agency jurisdiction over contracts involving games without giving it authority over gambling products themselves.

The Third Circuit had already read the surrounding statute in a fourth way in KalshiEX, LLC v. Flaherty, No. 25-1922, which Judge Lee expressly declined to follow on method. We walked through the split over the swap definition in a separate piece on the three circuits and the two October rulemakings. The point here is narrower and more practical. A provision that supports three different inferences is not a foundation. It is a variable.

An argument is only as durable as the instrument beneath it

Set the four instruments side by side and they sort themselves by how fast they can turn, which is the opposite of how they sort by apparent weight.

  • The statute, 7 U.S.C. section 7a-2(c)(5)(C). Slowest to change, and the least determinate, because four courts and one agency have read it four ways.
  • The regulation in force, 17 CFR 40.11. Determinate, verifiable to the day, and the only one of the four a venue can quote without qualification.
  • The pending proposal, RIN 3038-AF65. Fully briefed since 27 July 2026, with no scheduled outcome. Useful for anticipating direction, worthless as authority.
  • The interim final rule, RIN 3038-AF81. Written to take effect without waiting for comment, and not yet effective, because the trigger is a publication date the Commission does not set by itself.

A compliance position that cites a pending proposal is not wrong so much as undated. A position that cites a regulation in force is dated by definition, which is why the first question to ask about any internal legal memo is not whether it is right but as of when.

Dating a legal basis the way a venue dates a price

Every venue in this category already understands shelf life on the market side. A resolution source is pinned before a market opens. A price is stamped. A settlement rule is published with a version. The same discipline applied to the legal basis for a listing decision costs very little and is close to absent across the category.

Four things make the difference, and none of them require a rulemaking to finish first.

  • A classification record per contract family, with a date and an instrument version. Not "we believe these are swaps" but "assessed on this date against 17 CFR 40.11 as in force on that date, and against RIN 3038-AF82 as proposed." When the instrument changes, the record shows what has to be looked at again.
  • A written note of what changes under each outcome. If the exclusion in section 1.3 takes effect, which listed contracts are affected, and which are not? If the inclusion proposal is adopted instead, does anything about resolution, margining or disclosure change for an open position? A reader who has to ask is a reader who was not told.
  • A changelog for the legal basis, published like any other rulebook change. Holders of open positions learn about amendments to the rules governing their money through filings they have no reason to read. Telling them directly is voluntary and cheap.
  • The parts that hold whichever way this goes. Restricted participant lists, location controls, plain disclosure of the venue and the clearinghouse on the market card. None of these depend on the swap definition at all, which is exactly why they are worth building first.

The parts of this that are yours to decide

Our fifth directive says that standards should be set against the jurisdiction that regulates next, not only the one that has not yet. The October documents are a good illustration of why we wrote it that way. Nobody at any exchange chose the date on which an interim final rule reaches the Federal Register, or the date on which a proposal from June becomes a final rule, or whether a circuit panel reserves a question or answers it. Those are other people's calendars.

What a venue does choose is whether its own account of its legal footing is written down, dated and tied to a named instrument, so that when one of those calendars moves, the work is a revision rather than a reconstruction. Judge Lee's concurrence is a short lesson in the difference. He knew exactly what his conclusion rested on, said so in the opinion, and flagged that it was being rewritten. Six weeks later he was right, from a direction nobody had marked. If you can say the same about your own position, in writing, with dates, you are in a better place than the argument that won in August. Firms that want to hold that standard together rather than separately can read the commitment and sign it.

Share X LinkedIn Email