On 30 April 2026 the Commodity Futures Trading Commission received two comment letters about prediction markets from people who do not trade on them. One came from the National Cattlemen's Beef Association, signed by its senior vice president for government affairs, Ethan Lane, who described an association representing more than 178,000 members through direct membership and 44 state affiliates. The other was a joint filing on behalf of twenty one agricultural trade organisations, from the American Farm Bureau Federation and the National Corn Growers Association to USA Rice and the North American Millers Association, filed by Edmund Perry on the advance notice docket as comment CFTC-2026-0331-3522.
Neither letter asks for anything on behalf of its members' positions on a prediction market, because its members hold none. Both are about the exchange next door. The argument is that a binary contract referencing an agricultural futures price changes the behaviour of that futures market, and that the cost lands on producers, merchants and processors who never opened an account anywhere near it. This is a different shape of harm from the one the category usually argues about. It is not the trader who loses money. It is the hedger who never traded and pays more to hedge.
Twenty one associations signed a letter about an exchange they do not use
The joint letter opens by saying what it is not. The signatories write that they "strongly support the CFTC in its mission to promote and foster innovation in the derivatives markets", and that their concern is that novel market structures "might have unintended impacts on the trading of agricultural derivatives on traditional platforms". They then set out six mechanisms: settlement source opacity, trading hours that run past the close of the reference market, strike prices outside that market's daily price limit, the absence of hard position limits, the gap in published position reporting, and what leverage would do to any of the above.
The Commission's account of this appears once. In the cost-benefit section of its proposed prediction markets rule of 12 June 2026 (91 FR 35806, RIN 3038-AF65), the Commission writes that some commenters asserted that prediction markets built around agricultural events could shift liquidity and affect "the cost and reliability of hedge execution for producers, merchants, processors, and allied participants". Footnote 311 names the two letters. In a document of roughly 76,400 words, the word agricultural appears six times and agriculture three, and footnote 311 holds two of them.
The settlement source was not in the public filing and still is not
The coalition's sharpest sentence is an admission of what it could not find out. The signatories write that "it is not clear from the publicly available contract specifications" how the contracts settle, and that they understand, from conversations with Kalshi rather than from any document, that the markets settle to an average of quotes obtained from liquidity providers in the underlying futures contract on CME or ICE.
The public filing confirms the gap. KalshiEX LLC self-certified the contract "Will the <price type> price of <commodity> be above, below, between, exactly or at least <value>?" on 19 February 2026 under the rulebook code COMMODITIES, signed by head of markets Xavier Sottile. Appendix A says the Source Agency "refers to a commodity pricing data provider specified by the Exchange" and gives two examples, ICE Data Services and TradingView. Which provider applies to a given market is designated at issuance, and Appendix D, headed Source Agency, is one of three appendices for which the exchange requested confidential treatment. The same appendix set holds the analysis of the contract's "risk mitigation and price basing utility", which is the analysis that would answer the coalition's question directly.
The product definition is not narrow. Appendix A says <commodity> may be "a physical commodity or commodity futures contract" and lists wheat, corn, soybeans, coffee, sugar, cocoa and cotton among the examples. Directive 02 of our manifest asks that settlement sources be stated in plain language on the surface where the trade happens. An example vendor in the public filing, with the real one in a confidential appendix, is not that.
The fix came from a conversation nobody is obliged to have
Three footnotes in the joint letter describe what actually resolved part of this, and they are the most useful paragraphs in the whole filing. The signatories write that on learning of the contracts, members of the coalition "affirmatively reached out to Kalshi, and Kalshi engaged in a dialogue with the coalition". They record that the exchange amended certain aspects of the contracts as a result, that it indicated it would align trading hours with the hours of the underlying reference market and "appear to have made this change on some contracts already", and that it has said it will engage with the agricultural community before listing new contracts of this kind.
That is a good outcome and the coalition says so. It is also an outcome with no mechanism behind it. Nothing in part 40 required the call, nothing recorded what was agreed in a form a third party can check, and nothing makes the next exchange do the same. The coalition writes in the same footnote that it hopes "in the future these conversations can occur prior to listing new contracts". A standard that depends on the complainant finding the right phone number is not a standard. It is a courtesy that worked once.
The event market stays open after the reference price stops moving
The coalition's worked example is specific enough to check. It writes that the Kalshi "Soybeans price tomorrow at 5 pm EDT" contract settled at 5:00 p.m. Eastern while the underlying CME July 2026 soybean futures contract closed at 2:15 p.m. Eastern, and that the event market traded through the gap. The February self-certification supports the general point, stating that under Rule 5.16 of the exchange rulebook "trading shall be available at all times outside of any maintenance windows".
The weekend version matters more. The coalition notes that the CME soybean contract closes on Friday afternoon and does not reopen until Sunday evening, while a market referencing its price stays open throughout. If that market grows large enough for its prices to be read as sentiment, the reopening futures market has a published number to react to that was generated while its own book was shut. The coalition's concern is not that this has happened. It is that no rule addresses it, and that a meaningful move in the futures price reaches forward contracting and spot pricing, and from there the people who sell the crop.
A twenty five thousand dollar accountability level is not a position limit
The two regimes can be laid side by side from primary documents. Under Commission Regulation 150.2, federal speculative position limits apply to referenced contracts based on the core referenced futures contracts listed in paragraph (d), at levels fixed in appendix E to part 150. The February Kalshi filing sets, in place of that, a "Position Accountability Level" of 25,000 dollars per strike, per member. The August filing discussed below sets the same figure.
The coalition's objection is that accountability is a reporting trigger and a limit is a ceiling, and that the two are not interchangeable where both instruments reference the same underlying. It writes that the divergence "risks creating an unregulated sidecar venue in which large, directional positions can accumulate in instruments that are economically related to core agricultural futures". The letter also asks whether prediction market exchanges should carry volatility controls comparable to the daily price limits the reference exchanges enforce, noting that strikes were offered beyond the soybean daily limit band. Price limits are exchange governance rather than Commission rule, which the coalition acknowledges, and which is why the request is aimed at the exchanges as much as at the regulator.
The Commission asked this exact question in March and never answered it
The advance notice of proposed rulemaking published on 16 March 2026 (91 FR 12516) put the question better than the coalition did. In question 2.e, on Core Principle 5, the Commission asked what factors it should consider "regarding how position limits across similar event contracts should be aggregated (e.g., whether there is the same underlying reference, or whether there are similar references)". Aggregation across a shared underlying reference is the sidecar problem, stated by the regulator six weeks before the agricultural coalition stated it.
The June proposal does not take it up. The phrase position limit appears three times in its 76,400 words. Twice it sits in footnotes describing decades-old staff no-action letters for the Iowa Electronic Markets and PredictIt. The third is in the cost-benefit discussion of addictive potential, where position limits appear alongside cooling off periods and self-exclusion as protections for retail traders. That is a user protection use of the term, and a legitimate one, but not the market integrity use question 2.e was about. The string 24/7 does not appear in the proposal at all.
The report commercial hedgers read does not reach these trades
The reporting point is narrow and verifiable. The coalition writes that trades on whether futures prices will finish above or below a level "do not fit cleanly into COT reporting and are not, to our knowledge, being publicly disclosed by the CFTC", and asks the Commission to fold prediction market activity into the Commitments of Traders report or publish a new one. Hedgers use that report to see how categories of trader are positioned. A position taken on the same price on another venue is invisible in it.
What has moved since is frequency rather than scope. In remarks on 29 July 2026, Chairman Michael S. Selig said he had directed staff to begin circulating the Commitments of Traders report twice weekly instead of once, with a target of year end. A report that arrives more often still reports the same venues. The separate reporting proposal that does address event contracts sets its thresholds so high that, by the Commission's own arithmetic, almost nobody clears them, which we went through in detail in August.
The registry cannot be searched for the commodity you grow
There is a mechanical reason the coalition had to telephone the exchange. The Commission's public product registry lists KalshiEX self-certifications by product name, and the names are parameterised templates. Across the 194 KalshiEX entries dated between 2 April and 31 July 2026, not one product name contains the word cattle, corn, soybeans, wheat, coffee or sugar. The relevant entries are the February 2026 template quoted above and an earlier one certified on 14 November 2025. A trade association that wants to know whether anything references its benchmark cannot search for its benchmark. It has to open every filing.
The February filing adds a smaller oddity in the same direction. Its header gives the Kalshi contract category as Commodity and the internal category as Financials, and the opening line of the public analysis says the contract "is a contract relating to Financials". None of this is withheld. It is simply arranged so that the people most affected are the least able to find it, which is the practical content of Directive 02 when the audience is another market rather than a retail user.
The contract that came back in August is the better one
The agricultural price contracts did not survive the spring. Writing for the American Farm Bureau Federation on 22 June 2026, economist Bernt Nelson recorded that Kalshi's agricultural commodity contracts "were later removed and were not among its other listed commodity contracts as of June 18". Agricultural event contracts as a class did not disappear. The registry shows KalshiEX certifications through 2025 for crop yield per harvested acre, for the spot price of a crop, and for whether more corn or soybeans would be planted in Indiana.
Then on 19 August 2026, after the comment window closed, KalshiEX self-certified "Will Kansas wheat production be above, below, exactly, between or at least <count> bushels?" under the rulebook code KSWHEAT. It is worth reading because it is built the way the coalition asked. The underlying is total Kansas wheat production in bushels. The source agency hierarchy names the United States Department of Agriculture first, and the filing points to the National Agricultural Statistics Service quick stats service. The restricted list names USDA and NASS officials, employees and contractors, anyone compiling the estimates before publication, and their households. It settles on a published government statistic rather than another exchange's price, so the whole class of objection in the April letters does not arise.
That comparison is not a story about one exchange behaving badly. The same exchange filed both. One design routes around the neighbouring market and the other leans on it, and nothing in Regulation 40.2 distinguishes them. Both reached the market the business day after certification. The approval route that would let the Commission look first exists and has gone unused for decades, which we covered in September.
What an exchange could publish before it lists somebody else's price
The statutory hook is not missing. Under 7 U.S.C. 5, Congress found that the transactions the Act covers serve the national public interest through trading in "liquid, fair and financially secure trading facilities", and made it a purpose of the Act to "promote responsible innovation and fair competition among boards of trade, other markets and market participants". The proposal quotes that second phrase and then runs the analysis one way only, asking whether restricting an event contract would push demand offshore. The words other markets point the other way too. Nothing in the three general public interest factors of proposed Regulation 40.11(a)(5), covering hedging and price basing utility, manipulation and settlement integrity, and strain on the exchange's own compliance tools, asks what a listing does to the market it references.
An exchange does not have to wait for that. Four things could be published on the day such a contract is certified, and none require a rulemaking. Name the settlement source for each market, specifically, before the first trade. Match the trading hours of the reference market, or publish the reason for not doing so. Publish a position limit as well as an accountability level, and say how it aggregates across contracts sharing the same reference. And talk to the sector that uses the reference market before listing, then publish what changed as a result.
The coalition got three of those four out of one phone call, which tells you the cost is low. The reason to write them down is that the next sector to find a contract on its benchmark will not know who to call. Concentrated liquidity is what agricultural hedgers are defending, and it is the same property that makes a prediction market price worth reading. We have made that argument about this category's own order books. It does not become less true when the book belongs to somebody else.