The CFTC's Kalshi Order and the Registrant Caught in the Middle
By R Tamara de Silva | De Silva Law Offices, LLC | July 19, 2026
A registered exchange that receives conflicting commands from a state court and its federal regulator has, until now, been able to assume that compliance was possible. The path might be expensive and the sequencing delicate, but the regulatory framework at least contemplated an answer: declare an emergency, file the responsive rule with the CFTC, and manage the transition in an orderly way. On July 14, 2026, the CFTC closed that path. The order arises out of Kalshi’s dispute with the State of Michigan, and its mechanics can be stated in three steps. A Michigan court ordered Kalshi to cancel certain executed trades. Kalshi, as the CFTC’s regulations require, filed an emergency rule setting out how it would comply. The CFTC then declared that Kalshi’s filing itself constituted a market emergency, stayed the rule, and ordered Kalshi to fulfill the very trades the Michigan court had ordered cancelled. The filing the regulations compelled Kalshi to make became the basis for a federal order directing Kalshi to defy a state court. The order resolves that conflict for Kalshi. For every other platform, intermediary, and new entrant in the event contract markets, it remains open.
Most of the commentary this week has treated the order as another chapter in the federalism contest between the CFTC and the states. It is that. But the order read on its own terms, rather than through the press release, tells a second story about what happens to a registrant that follows the rules when the rules point in two directions at once.
How Kalshi Got Here
On June 29, 2026, the Circuit Court for Ingham County, Michigan entered a temporary restraining order in Nessel v. KalshiEX LLC. The TRO prohibits Kalshi from offering, listing, matching, executing, clearing, or settling anything that constitutes internet sports betting under Michigan law to any person located in Michigan. It requires Kalshi to retain a geolocation provider licensed by the Michigan Gaming Control Board. And it imposes a fine of $120,000 for every day Kalshi fails to comply with the geolocation requirements.
Kalshi moved to dissolve or modify the TRO. The court responded by going further. In a verbal modification, later confirmed in a July 6 letter to the parties, the court directed that certain trades already entered into by Michigan residents be voided, cancelled, and refunded.
It is astonishing that a state trial court ordered a federally regulated exchange to unwind executed derivatives transactions governed by a federal regulatory regime.
The district courts are genuinely split on the merits of Michigan's underlying gambling claim, as we discussed in Kalshi's NY Loss: The Preemption Question Judge Torres Skipped. But whatever one thinks of that question, cancelling executed trades is a different order of intervention than blocking new ones. However, cancelling executed trades is a different order of intervention than blocking new ones. Every state that has litigated against Kalshi so far has sought to stop the platform from taking new business. Michigan reached into the order book and interfered with already executed trades.
What Kalshi Did Next
Kalshi did what the regulations tell a DCM to do when a government action threatens its market. Its own rulebook defines an emergency to include any action by a state governmental body that may have a direct impact on trading.
CFTC Regulation 40.6(a)(6) requires a registered entity that adopts an emergency rule to file it with the CFTC, before implementation if practicable, and within twenty-four hours if not.
So on July 6, Kalshi notified the CFTC of an imminent market emergency. On July 12, it filed its Emergency Rule. Under that rule, the positions of the specific Michigan users identified by the court would be force-liquidated on the central limit order book at current market value. Where the liquidation value fell below a user’s original cost of entry, Kalshi would pay the difference out of its own operational funds. Kalshi told the CFTC it would absorb the entire shortfall, and that it could afford to because the number of affected positions was limited.
The design of the rule merits attention as no trader loses money. The order book absorbs the liquidations through ordinary market mechanics. Kalshi bears the entire cost. As court-ordered unwinds go, it is difficult to imagine a compliance plan more carefully constructed to minimize market impact. Kalshi was attempting to thread the needle between a $120,000-per-day contempt exposure in Lansing and its obligations as a DCM in Washington.
The CFTC cut the thread.
What the CFTC Did
Two days later, the CFTC issued its order doing two things at once. First, it stayed the Emergency Rule under Regulation 40.6(c)(1). Second, invoking its emergency authority under Section 8a(9) of the CEA, it directed Kalshi to fulfill the open trades in the ordinary course of business. Fulfill, that is, the very trades a Michigan court has ordered voided, cancelled, and refunded.
The federalism stakes are obvious, and the CFTC’s order does not hide them. It states flatly that state courts cannot order the unwinding of executed swap transactions, and it warns that if they could, nothing would stop them from reaching futures, options, and forwards next. That is the argument the CFTC has been making in its lawsuits against nine states and in its amicus filings, and it is the argument commentators have spent the week debating.
But the federalism argument is not what makes this order remarkable. Three features of the text are.
Elegant architecture in choosing the emergency
Section 8a(9) lets the CFTC act when it has reason to believe an emergency exists, and the statute defines emergency to include a major market disturbance which prevents the market from accurately reflecting the forces of supply and demand. The order’s operative finding reads: “the Emergency Rule, adopted in response to a Michigan circuit court’s unprecedented order requiring Kalshi to unwind open, previously executed trades constitutes an emergency.”
The Emergency Rule constitutes the emergency. Not the Michigan TRO. The registrant’s own compliance filing is the disturbance the CFTC acted upon.
This is an elegant argument and it solves a problem for the CFTC. Had the CFTC named the Michigan TRO as the emergency, it would have been ruling on the validity of a state court order, and that is the preemption question currently dividing the federal courts. The CFTC would have been litigating on ground it does not control. By locating the emergency in Kalshi’s Part 40 filing instead, the CFTC never leaves its own territory. A DCM rule filing is a creature of the CFTC’s regulations, submitted under the CFTC’s procedures, and subject to the CFTC’s undisputed authority to review and stay. Acting on the filing raises no preemption question at all.
The consequence of this design is the striking part. The CFTC never holds the Michigan order invalid. It does not need to. It simply stays the rule Kalshi adopted to comply with that order and directs Kalshi to fulfill the trades instead, which means that compliance with the state court is now a violation of a federal directive. Kalshi is left commanded to defy a court whose order the CFTC has pointedly declined to call unlawful. The state order stands, untouched. Obeying it is what became impossible.
The order concedes the numbers and finds an emergency anyway.
The statutory trigger is the major market disturbance. The order acknowledges that the court’s directive applies to a limited number of trades and that Kalshi can absorb the associated costs, then holds that this “does not obviate the market emergency” because “the unprecedented forced liquidation of even one executed trade risks market distortions.”
The CFTC’s underlying concern is legitimate, and it sits near the core of the agency’s mission. The finality of executed trades is what makes a derivatives market a market. Nor does a trade exist in isolation. Every executed trade feeds the price at which the next trade is struck; the Michigan positions helped form the prices on which thousands of subsequent trades, hedges, and related contracts were built. Retroactively voiding them does not simply erase a handful of positions, it falsifies an input to price discovery that the rest of the market has already relied upon.
The order makes this point itself, noting that participants may hold corollary contracts exposed to the resulting volatility. If a state court can unwind one cleared trade, then every cleared trade on every exchange carries unwind risk, and traders will price that risk whether the affected positions number five or five thousand. On that logic, a single trade really is the correct unit of analysis, because the injury runs to the rule of finality rather than to the positions themselves.
But that is a defense of the policy, not of the statutory fit. Section 8a(9) speaks of a major market disturbance "which prevents the market from accurately reflecting the forces of supply and demand." Prevents, in the present tense. The provision describes a market that is malfunctioning now: a squeeze, a default, a delivery failure, the kinds of fast-moving breakdowns in which waiting for a court means the damage is done before the ruling arrives. That is why Congress made the power immediate, and it is the premise on which the Seventh Circuit was willing to hold its exercise unreviewable.
The disturbance the CFTC describes in this order has not happened. Kalshi's markets were functioning normally on July 14. Prices were clearing, trades were settling, and the handful of Michigan positions at issue would have been liquidated at market value with every trader made whole. The CFTC's concern is prospective: that if a state court's cancellation order is allowed to stand as precedent, participants will begin pricing unwind risk into contracts that today trade without it. That is a genuine injury, but it is an injury threatened by a legal precedent, not one being inflicted by a malfunctioning market. And the ordinary remedy for an unlawful state court order is litigation, a remedy the CFTC is already pursuing against nine states. The question the order leaves unanswered is why this dispute, of all the CFTC's pending conflicts with the states, required resolution through an emergency power the CFTC itself describes as beyond judicial review, rather than in the courts where it is already making the identical preemption argument.
The Mismatch Between the Stay Ground and the Command.
Regulation 40.6(c)(1) gives the CFTC two grounds to stay a certified rule: the rule presents novel or complex issues requiring more time to analyze, or the rule is potentially inconsistent with the Act. The CFTC chose the first. It did not find Kalshi’s rule inconsistent with the CEA. It found only that it needs more time to study the question, and the regulation gives it 90 days to do so, with a 30-day public comment period inside that window.
Yet paired with that provisional stay is a permanent, self-executing command under Section 8a(9) to fulfill the trades. And under Regulation 40.6(c)(5)(ii), the stay itself becomes presumptive evidence that Kalshi cannot truthfully re-certify the same rule. That is a great deal of substantive freight loaded onto a procedural pause.
One further detail deserves mention for readers following the litigation. The order’s Authority section recites, quoting the Seventh Circuit’s 1979 decision in Board of Trade of the City of Chicago v. CFTC, that the merits of the CFTC’s emergency determination are precluded from judicial review.¹ An agency that cites, in the body of its own order, the authority for why that order cannot be reviewed is signaling the fight it anticipates. Whether a 47-year-old Seventh Circuit holding concerning a wheat contract insulates a 2026 order directing defiance of a state court is a question that will not be answered in the Seventh Circuit, because this dispute does not sit there.
What This Means for Platforms and Intermediaries
The lesson for the industry is an uncomfortable one.
Kalshi’s Part 40 emergency filing was not a safe harbor. It was the transmission mechanism. Consider what the filing actually accomplished for Kalshi. The regulations required it, so Kalshi had no choice but to make it. Once made, it handed the CFTC both notice of Kalshi’s compliance plan and a procedural vehicle for stopping that plan, because a filed rule is something the CFTC can stay. The filing Kalshi was obligated to submit is the reason the CFTC’s order exists and the mechanism through which it operates. That is the trap, and it generalizes. A registrant caught between a state court and the CFTC has no filing available to it that satisfies both sovereigns. If it adopts a rule to comply with the state court, it must file that rule, and the CFTC can stay it and order the opposite, which is precisely what happened here. If it complies with the CFTC instead, it accrues contempt exposure at $120,000 a day. Every door leads to the same room.
Kalshi can withstand that position for a considerable time. It is well capitalized, it volunteered to absorb the liquidation shortfall from operational funds, and this fight is existential for its business model, so it will fund the litigation to its conclusion. The same fact pattern applied to a smaller platform, a guaranteed introducing broker, or any of the new entrants building in this space produces a very different result. A $120,000 daily fine is not a rounding error for those firms. Neither is the litigation budget required to fight a two-front war against a state attorney general and, potentially, the firm’s own federal regulator. The CFTC’s order protects the trades on Kalshi. It does nothing to resolve the underlying conflict for anyone else, and the underlying conflict is now live in every state where the CFTC has filed suit.
There is also a structural observation here that connects to the comment letter this firm filed with the CFTC earlier this month on vertical integration in the event contract markets. The order directs Kalshi “and its affiliates” to fulfill the trades. In a vertically integrated structure, where the exchange, the clearinghouse, and the brokerage function sit inside one corporate family, a single order to the DCM reaches the entire trade lifecycle. There is no independent FCM or clearing member in the stack with its own obligations, its own counsel, and its own view of whose order controls. Whatever the merits of integration as a business model, this episode shows what it looks like under stress- one company, alone, holding the entire conflict between two sovereigns.
The Comment Window Is Open
Because the CFTC stayed the rule on novel-or-complex grounds, Regulation 40.6(c)(2) requires a 30-day public comment period during the 90-day review. That window is an opportunity for market participants to put on the record what the order leaves unanswered. what a registrant is supposed to file the next time a state court orders it to do something the CFTC forbids, whether an emergency finding can rest on a single compensated trade, and how the CFTC intends to reconcile a procedural stay with a permanent substantive command.
The states, for their part, are unlikely to treat the CFTC’s assertion of unreviewability as the last word. Wisconsin has already demonstrated what a determined state court can do to the preemption theory, as this firm discussed last month. Michigan has now demonstrated what a determined state court can do to an order book. The CFTC has demonstrated what it will do in response. What none of the three has yet provided is a workable path for the registrants standing in the middle.
Endnotes:
¹ Board of Trade of the City of Chicago v. CFTC, 605 F.2d 1016, 1025 (7th Cir. 1979), cert. denied, 446 U.S. 928 (1980).
De Silva Law Offices, LLC advises exchanges, intermediaries, and market participants on CFTC and NFA regulatory matters, including event contracts, registration, and enforcement defense. Questions about this article or the firm’s practice may be directed to info@desilvalawoffices.com.
De Silva Law Offices, LLC | 110 North Wacker Drive, Suite 2500, Chicago, IL 60606 | 312-500-8424 | desilvalawoffices.com
NB: This article is for informational purposes only and does not constitute legal advice. Reading this article does not create an attorney-client relationship.