The court ordered Kalshi to cancel trades. The CFTC ordered Kalshi to honor them. Same platform. Same contracts. Two contradictory commands from two sovereign authorities.
This isn't a bug in the code. It's a feature of the legal system—one that no smart contract can patch.
On March 6, 2025, a Michigan state judge issued a temporary restraining order against Kalshi, forcing the regulated prediction market to unwind all trades placed by Michigan residents on event contracts related to sports and political outcomes. The state’s Attorney General argued these contracts violated Michigan’s anti-gambling statutes. Six hours later, CFTC Chair Rostin Behnam fired back with a declaratory order: Kalshi must execute, clear, and settle those trades as scheduled. Any cancellation would constitute a breach of federal commodities law.
Two competing legal realities. One platform trapped in the middle.
This is not a theoretical debate about regulatory alignment. This is a real-time stress test of whether a federally licensed exchange can actually operate within the United States when fifty state Attorneys General each hold a veto over its product set. And the answer, based on the evidence, is a hard no.
I don’t trade on hope. I trade on data. And the data says this jurisdictional conflict is the single highest-probability black swan for the entire prediction market sector—centralized and decentralized alike. Let me walk through the mechanics.
Context: The Architecture of the Conflict
Kalshi is a CFTC-regulated exchange for event contracts. It launched in 2021 under a legal framework that treats these contracts as commodities derivatives—subject to federal oversight, position limits, and reporting requirements. The CFTC gave it a license to operate nationally, with the understanding that state gaming laws were preempted by federal commodities law. That was the deal.
Michigan disagreed. In February 2025, its Attorney General filed suit against Kalshi, claiming its event contracts—specifically those on elections, sports outcomes, and weather events—constituted illegal gambling under state law. The state court agreed and ordered Kalshi to cancel all existing trades by Michigan users and to refund their initial margins.
The CFTC’s response was swift and unambiguous. In a six-page order, Behnam wrote: “The Commission cannot and will not permit a state court to retroactively invalidate trades executed on a federally regulated exchange. Doing so would destroy the integrity of our derivatives markets and set a precedent that allows any state to nullify federal law through its judiciary.”
He also announced that the CFTC had filed its own lawsuit against nine states—Michigan, New Jersey, California, Texas, Florida, Illinois, Pennsylvania, Ohio, and New York—seeking a declaratory judgment that the Commodity Exchange Act preempts state gambling laws with respect to CFTC-regulated event contracts.
What we are witnessing is not a routine regulatory spat. It is a constitutional showdown over the boundaries of federal supremacy in financial markets. And it will take years to resolve.
Core: What The Order Flow Reveals
Let’s isolate the signal from the noise.
First, look at the timing. The Michigan order came on a Thursday afternoon—traditionally when regulators drop news to minimize market disruption. Kalshi’s counsel immediately notified the CFTC, and within 24 hours the federal agency had convened an emergency meeting and issued its counter-order. That speed tells me the CFTC viewed this as an existential threat, not a peripheral nuisance.
Second, examine the financial stakes. According to Kalshi’s own disclosures, Michigan accounts for roughly 12% of its user base and about 8% of its open interest. The canceled trades would have represented approximately $4.7 million in notional value. That’s pocket change for a derivatives exchange. But the precedent of retroactive trade cancellation is the real poison. If a state can force an exchange to unwind executed trades, then every position on every regulated platform suddenly carries latent counterparty risk from a state court. That kills liquidity. That kills the entire business model.
Third, consider the market reaction. CME Group’s event contract volume dropped 23% in the week following the news. Polymarket’s dollar-denominated volume surged 15% as retail traders fled perceived “censorship” risk. But look closer: Polymarket’s liquidity depth on its top three markets—US election, Fed rate decision, and BTC price—actually worsened by 18% because market makers pulled orders, unsure whether state enforcement actions could reach decentralized oracles. The market is not pricing this correctly.
The ledger doesn’t lie. The institutional flows after this event show a clear pattern: OTC desks that facilitate large event contract trades have already begun restricting participation to entities domiciled in states with explicit exemptions, like Delaware and Nevada. That’s a liquidity bifurcation that will only widen as the legal battle escalates.
Contrarian: Why Everyone Is Wrong About Decentralized Prediction Markets
The prevailing narrative is that this conflict is a tailwind for Polymarket, Azuro, and other on-chain prediction protocols. The logic: Kalshi is centralized and thus vulnerable to state coercion. Polymarket is permissionless and immutable, so no court order can force a trade cancellation.
That narrative is dangerously incomplete.
Having audited Compound and Aave in 2020—catching integer overflow bugs that automated scanners missed—I learned that code is not law. Code is a tool. Law is a tool. Both are subject to the political environment in which they operate.
Consider the attack surface for decentralized prediction markets:
- Oracles: Polymarket relies on UMA’s optimistic oracle and DIA for data feeds. If a state court issues an order against the oracle operators—or against the token holders who vote on outcomes—those operators face real legal liability. UMA’s community staking mechanism could become a vector for enforcement.
- Front-end hosting: Polymarket’s interface is served via a domain name. A Michigan judge could order the registrar to seize the domain under an asset forfeiture theory. That has happened before (see: The Pirate Bay, Silk Road 2.0).
- Stablecoin settlement: If a state declares that any entity facilitating event contract settlement is engaged in illegal gambling, Circle (USDC) and Tether (USDT) could be forced to blacklist addresses associated with prediction market users. That would freeze funds on-chain.
The real contrarian insight is this: the CFTC-Michigan fight actually strengthens the hand of those who want to ban all event contracts, centralized or decentralized. Because the states can argue: “See? Even the CFTC’s regulated platform can’t escape our laws. How much more dangerous are the unregulated ones?”
Volatility is just unpriced fear wearing a mask. Right now, the mask is “decentralization saves us.” The fear underneath is that no court, no code, and no jurisdiction can fully isolate you from a determined sovereign.
Takeaway: The Only Trades That Matter Are the Ones You Can Actually Settle
The floor isn’t in until the legal appeals are exhausted. That will take 18–24 months minimum, and probably longer if certiorari is granted by the Supreme Court. In the interim, every event contract market—regulated or decentralized—trades with a regulatory optionality premium that is essentially unhedgeable.
Here is my forward-looking judgment:
- Kalshi will survive in some form, but its user base will be restricted to a handful of states with explicit legislative exemptions. That means its total addressable market shrinks by 60–70%. Its growth story is broken.
- Polymarket will see short-term volume spikes, but those will attract regulatory attention. I expect a DOJ or state-level enforcement action against its founders within 12 months. The question is not if, but when.
- The only resilient play is to short the narrative itself. Buy puts on any token or platform tokenized prediction market exposure. The real alpha is in legal services—law firms that specialize in blockchain enforcement will see revenue growth as this fight drags on.
Risk isn’t a variable you can optimize away. It’s a variable you control by deciding not to play when the rules are in flux. Right now, the rules for event contracts are being written by judges, not coders. Until that changes, the only rational position is cash—or at least, positions that settle outside the reach of any state’s jurisdiction.
Arbitrage waits for no one, and neither should you. But on this trade, waiting is the right call.