How an asset class is made

Here is a fact that should be more unsettling than it sounds: between September 2023 and March 2026, the prediction-market contract did not change at all. A contract that pays $1 if an event happens and $0 if it doesn't, trading at the crowd's odds — the same instrument we've called a derivative, not a wager¹ — was exactly the same object in both years. What changed was the verdict. In 2023 the Commodity Futures Trading Commission moved to ban these contracts as unlawful "gaming." In 2026 a CFTC advisory opens by describing prediction markets as "a financial asset class and a proven source of reliable information for news media, sports leagues, financial institutions, and everyday Americans."⁷

Same contract. Opposite verdicts, thirty months apart. That gap is the whole subject of this piece — because it reveals something the breathless coverage of prediction markets keeps missing. A new financial asset class is not invented. It is recognized. The instrument can sit there fully formed for years; it becomes an asset class only when a regulator decides to own it. In the United States, that decision just happened, in public, as a reversal — and the shape of that reversal is the template every other country's regulator will eventually face.

This is the companion to how a single contract actually gets listed. That piece is about the machinery. This one is about the mind change that made the machinery matter.

The ban that lost

Start where the hostility was sharpest. In 2023, Kalshi — a federally licensed exchange — self-certified contracts on which party would control Congress. The CFTC opened a review and then, in September 2023, issued an order of disapproval: the contracts, it held, involved "gaming" and activity unlawful under state law, and were therefore contrary to the public interest.² The following May, the agency went further, proposing a rule that would have defined "gaming" broadly enough to ban political, sports, and awards contracts wholesale across the whole market. The vote was divided; two commissioners dissented.³ This was a regulator trying, deliberately, to wall the category off.

It lost. Kalshi sued, and in September 2024 a federal court vacated the disapproval, holding that an election is not a "game" in any ordinary sense of the word. The D.C. Circuit declined to freeze that ruling, the contracts went live and traded straight through the November 2024 election, and in May 2025 the CFTC quietly dropped its appeal.⁴ The ban didn't just stall. It was beaten in court, and then abandoned. That defeat is the hinge everything after it turns on.

The turn

Two things changed at once: who held the gavel, and what the agency wanted. The acting chair through 2025, Caroline Pham, had already begun thawing the posture — running a "crypto sprint" and convening a council of exchange and crypto chief executives to talk through market structure.⁵ Then in December 2025 the Senate confirmed Michael Selig as the 16th CFTC chairman. Selig came directly from running the crypto task force at the SEC, and he arrived talking about a "Golden Age of American Financial Markets" and a doctrine of regulating with "the minimum effective dose."⁵ Within weeks, in February 2026, the agency formally withdrew the 2024 proposal to ban event contracts, along with a 2025 staff advisory that had warned firms off sports markets.⁶ The wall the previous regime tried to build was taken back down by the next one.

It's worth being precise about what "embrace" means here, because it isn't a single dramatic act. It's the accumulation of small official gestures, each of which reads, in hindsight, as the regulator deciding the category is its own.

An asset class, in writing

The clearest of those gestures came in March 2026. The CFTC's Division of Market Oversight issued a staff advisory to every licensed exchange, and its first sentence is the one to dwell on: prediction markets, on which event-contract derivatives trade, are "rapidly increasing in popularity with the American public both as a financial asset class and as a proven source of reliable information."⁷ That is the federal government, in writing, filing prediction markets under the same conceptual heading as stocks, bonds, and commodities.

Hold the significance and the caveat together, because both matter. The caveat: this is a staff advisory — the views of one division, which explicitly says it creates no rights and isn't the position of the full Commission.⁷ Nobody passed a law declaring a new asset class; that's not how it works. The significance: words like these, from this office, are exactly the soft infrastructure on which an asset class is built. The same month, the agency opened a formal request for comment on how it should regulate prediction markets at all⁷ — the procedural opposite of trying to ban them. The question had shifted from whether these markets may exist to how the federal government will house them.

Who's in the room

If you want to know whether a regulator has truly adopted a category, don't read its press releases — read its invitation list. In January 2026 Chairman Selig relaunched the agency's main advisory body as the Innovation Advisory Committee, and the following month named its members.⁸ The roster is the tell. Sitting on a federal advisory committee are the chief executives of the exact firms the agency had spent the prior two years fighting or eyeing warily — Kalshi and Polymarket among them, alongside Crypto.com, Coinbase, Robinhood, DraftKings, and FanDuel, with the heads of CME, ICE, Nasdaq, Cboe, and the major crypto venues filling out the table.

Kalshi is the cleanest irony: the firm whose contracts the CFTC moved to disapprove in 2023 now has its founder advising the CFTC in 2026. The committee's formal charter is broad — innovation, digital assets, AI — and doesn't single out prediction markets by name; the agency parked the prediction-markets work in a separate internal task force, where "prediction markets and event contracts" is one of three explicit workstreams.⁸ But the direction is unmistakable. The people the regulator once tried to shut out are now the people it asks for advice. That is what adoption looks like from the inside.

Exclusive jurisdiction

And defend it is exactly what the agency did next — which is the part of this story that turns a posture into a commitment. Having decided event contracts are its asset class, the CFTC began asserting that they are only its asset class. Under the Commodity Exchange Act the agency holds exclusive jurisdiction over derivatives traded on the exchanges it licenses, and through the first half of 2026 it went on the offensive to enforce that against the states.⁹

The chairman set the tone in a February op-ed: the CFTC "will no longer sit idly by while overzealous state governments undermine the agency's exclusive jurisdiction," he wrote, insisting event contracts are swaps under federal law, not bets under state law.⁹ Then the filings came. The agency sued Arizona, Connecticut, and Illinois in a single action, then New York, Wisconsin, and Rhode Island; it won a restraining order against Arizona's criminal enforcement; and it filed brief after brief — in the Third Circuit, the Sixth Circuit, and Massachusetts's high court — arguing that federal law preempts state gambling statutes wherever a CFTC-licensed exchange is involved.⁹

Read the whole arc and the irony lands fully. In 2023 the federal regulator was the prediction-market industry's chief antagonist. By 2026 it is the industry's chief protector — the entity going to court to keep states from shutting these markets down. You do not litigate that hard for something you consider a nuisance. You litigate that hard for something you've decided is yours.

The new rulebook

The capstone, so far, arrived on June 10, 2026: a proposed rule that builds the durable framework for the asset class the agency now owns.¹⁰ Two design choices in it are worth understanding, because together they're the inverse of the 2024 ban.

First, it refuses a categorical ban. Where the 2024 proposal would have prohibited whole categories — all political contracts, all sports — the 2026 proposal says the law doesn't permit a blanket prohibition, and replaces it with a contract-by-contract public-interest test weighing whether a given contract offers hedging or information value, whether it's manipulable, and whether the exchange can actually police it.¹⁰ Second, it draws the line by the structure of the outcome rather than the topic. Contracts that settle on aggregate, hard-to-rig outcomes — a final score, a season-long total — are likely fine; contracts that settle on a single play, an injury, an officiating call, a physical altercation, or youth sports are likely out, because one person can move them.¹⁰

It is, to be clear, a proposal — open for public comment, not yet binding, and the details will move.¹⁰ But the orientation is the point. A regulator that wanted these markets gone does not spend 200 pages designing a careful, permissive framework for which ones may live. It writes one short paragraph banning them all. The 2024 agency wrote the paragraph. The 2026 agency wrote the framework. That is the distance traveled.

The reversal is the template

Step back from the American specifics and the larger pattern is what matters for everyone else. An asset class is a regulatory decision wearing the costume of a financial instrument. The instrument is necessary but never sufficient: prediction markets existed, fully built, while the CFTC was trying to ban them. What turned them into an asset class was a sequence any regulator can run — stop fighting it, name it in writing, bring its builders to the table, defend its turf, and write its rulebook. The United States just ran that sequence in public, in about eighteen months.

Every other country now has that sequence sitting on the shelf, and each will face the same first decision the CFTC faced: which regulator owns this instrument? Route it to a gambling authority and you get the 2023 verdict — a bet, to be banned. Route it to the body that already owns derivatives and you get the 2026 verdict — an asset class, to be supervised. We've argued that in Vietnam that fork points at one door, the commodity-exchange regime, and that the listing machinery a licensed venue would run there is the same machinery Kalshi runs here.

Seeker's work lives in that unmade decision. We're not claiming a licensed exchange is live anywhere we operate; it isn't, and the license is the goal, not a fact. What we believe is narrower and, we think, harder to argue with after watching the American reversal: this category becomes legitimate the moment the right regulator decides to own it, that decision is a reversal more often than a clean start, and the venue standing ready — compliant, cleared, surveilled — when a country's regulator finally makes the turn is the one that inherits the market. The U.S. just showed the whole world how an asset class is made. The only open question left is who runs the play next, and where.