Kalshi's first lifetime ban is its most important trade

In the days before the State of the Union address in February, Kalshi, an exchange for "event contracts", let people bet on whether George Santos would show up. The night before the speech, the market put the former Republican congressman's attendance at close to 75%. Mr Santos — expelled from the House in 2023 and released from prison early by President Donald Trump after a seven-year sentence for fraud and identity theft — had posted on X that he would be "there for the State of Union in the gallery, guys." He had also been quietly betting, on the same platform, that he would not attend. The next day he announced he was "watching from the airport." He did not appear, the contracts resolved in his favour, and he banked a profit of under $18,000.

The response, not the grift, is the story. On August 31st Kalshi imposed the first lifetime ban in its history and fined him $71,356. A month earlier the Commodity Futures Trading Commission (CFTC), which had pursued him for "manipulative trading" of a State-of-the-Union event contract, settled with Mr Santos for about $35,000 and a three-year trading ban. Private exchanges do not usually punish their customers more harshly than the state does. The exception here reveals what kind of business prediction markets have become.

They have become serious money. Kalshi, a CFTC-regulated exchange founded in 2018, went from a valuation of roughly $5bn at the start of 2025 to $11bn by December and $22bn by the spring; it is now reported to be raising at least $750m at a valuation near $40bn. This year it has recorded its first billion-dollar trading days and nearly $9bn of weekly volume. Robinhood, the retail broker, earned $156m from event contracts in the second quarter, more than ten times its revenue a year earlier. Every one of those numbers rests on a single promise: that the prices quoted are honest. Insider trading does not inconvenience that promise. It negates it.

Underneath the carnival is a mechanism worth understanding. A prediction market's product is the price — a probability distilled from many wallets, each vote backed by money rather than opinion. An insider with private knowledge corrupts that product; an insider who can also manufacture the news corrupts it twice. Mr Santos's airport post was not commentary on the market; it was the information the market was pricing. Hence Kalshi's rule, unusual and telling, that nobody may trade who is "capable of influencing the outcome of the underlying event". Equity markets solved this class of problem long ago with insider-trading law. Prediction markets, which price the behaviour of the very people who know that behaviour best, are only now building the equivalent.

They are building it in public, and escalating. In February Kalshi banned a California politician who had wagered about $200 on his own long-shot run for governor. In April its screening systems caught three congressional candidates, blocked one before he could trade and suspended them all for five years. In June, before the CFTC settled, the platform flagged Mr Santos's February wagers to federal prosecutors. Heads of enforcement describe a policy in which no trade is too small to punish. The escalations are deliberate: every public catch is a demonstration that the house polices itself, which is the only licence a market in other people's honesty can hold.

The limits are as revealing as the machinery. Mr Santos got away with his trades for the better part of a month and was caught only because he is famous, the sums were trivial and a regulator was paying attention; the enforcement head at Kalshi noted that he was the only one of those investigated who refused to cooperate. Detection that works for a cartoonish celebrity is not proof that it will work for a million anonymous bettors, and the surveillance that caught the April candidates is new and costly. The outside pressure, meanwhile, is real: Arizona's attorney-general has charged Kalshi with running an illegal gambling venture; Illinois and New York have barred state employees from trading prediction markets on inside information; Congress has filed bills to restrict how officials use the platform.

That is the trade at the centre of a $40bn valuation: the exchange must make enforcement scale faster than its own growth multiplies the temptation to cheat. For the retail investor, the exposure is indirect — Kalshi is private, its reported round led by Sequoia Capital and Wellington Management, and public reach runs through brokers such as Robinhood. The test is November's midterm elections, when prediction markets expect their largest volumes and their most fertile conditions for abuse. Mr Santos, unrepentant, thanked Kalshi for the lifetime ban and asked how long "you guys" would be around. He has, inadvertently, given the answer his scepticism cannot touch: a public, credible, escalating punishment of an insider is the cheapest insurance a market in prices can buy. The question investors should follow is not the volume chart but the enforcement record — every cheat caught is trust purchased cheaply, and every one that slips through is the valuation spent.