Polymarket Wants Americans to Bet on Events With Borrowed Money

The entire regulatory case for prediction markets rested on one boring fact: every bet was backed dollar for dollar. Polymarket just asked Washington for permission to change that.

On July 3, the company filed an application with the National Futures Association to operate as a futures commission merchant through an affiliate called Coming Home GBA LLC. The goal, per The Block, is regulated margin trading for US customers — the ability to open a position on an election, a rate decision, or a sports final without putting up the full face value in cash.

That sounds like a technical plumbing upgrade. It isn't. It's a change in what the product is.

The safety pitch was the collateral

When Polymarket bought its way back into the US market through QCX LLC, the CFTC-registered exchange it acquired in 2025, every contract on the platform had to be fully collateralized. You could lose exactly what you staked and not a cent more. No liquidation cascades. No margin calls at 3 a.m. No sudden hole in a clearinghouse.

That constraint did real work in Washington. It let prediction markets argue they were nothing like the leveraged casinos regulators spent 2022 cleaning up after. Event contracts were framed as an information tool with a stake attached — a "truth machine" whose prices doubled as the best available forecast of real-world outcomes.

Margin cuts against every part of that pitch. Polymarket will still need the CFTC to approve changes to its rulebook before customers can trade without full collateral, CoinDesk reports. But the direction of travel is unambiguous.

The timing is the strange part

Polymarket is making this ask while under investigation. The CFTC opened a broad probe into the platform in late June — the first high-profile inquiry of Chairman Michael Selig's tenure — reportedly including the company's marketing. The Wall Street Journal found Polymarket had hired dozens of mostly college-aged content creators to film staged trades and fake wins to pull users onto the site. Senators have demanded a federal investigation of their own.

Read those two facts together. A platform accused of manufacturing hype to recruit young bettors is simultaneously asking for the right to lend those bettors money to bet with. A compliance department with normal reflexes would have spaced these out by a year. Polymarket filed anyway.

Why? Because the race with Kalshi leaves no room to wait.

Kalshi already has the license

Kalshi obtained its own FCM license earlier this year, and in May the CFTC approved its BTCPERP bitcoin perpetual contract — the first perpetual futures product on a US-regulated exchange — which went live in early June. Kalshi cleared the margin runway months ago. Every week Polymarket spends fully collateralized is a week its sophisticated flow — market makers, hedge funds, arbitrage desks — has a reason to trade somewhere more capital-efficient.

And the pool they're fighting over is expanding at a rate that explains the urgency. Prediction market volumes hit $51 billion last year and are pacing toward roughly $240 billion in 2026. Nobody wants to be the deleveraged venue in a five-fold growth year.

The capital-efficiency argument is real, and it deserves a fair hearing. Full collateralization is genuinely punitive for market makers, who must lock up enormous sums to quote two-sided prices on hundreds of markets at once. Better-funded market makers mean tighter spreads, deeper books, and — yes — more accurate prices. If you believe prediction markets are forecasting infrastructure, margin for professionals arguably makes the forecasts better.

But leverage doesn't stop at the professionals

Here is the problem with that argument: no venue in history has managed to give leverage to market makers and keep it away from everyone else. The FCM structure Polymarket is applying for is a brokerage license. Brokerages serve customers. And the customers prediction markets have been recruiting — through influencer campaigns, through sports partnerships, through World Cup marketing — are not hedge funds.

A retail bettor with 5x margin on an election contract is not contributing price discovery. He's gambling with borrowed money on an outcome he cannot influence, hedge, or model better than the crowd. When the outcome resolves against him, he doesn't lose his stake — he loses his stake and owes more. Sports betting regulators spent two decades learning what credit does to problem gambling. Prediction markets are about to run the experiment again with a federal futures license.

There's a market-integrity cost too. Fully collateralized books are hard to manipulate cheaply because moving a price requires real capital at risk. Leveraged books lower the price of manipulation by exactly the margin multiple. For markets whose entire social value is the accuracy of their prices — the thing every "truth machine" defender points to — that trade-off lands on the wrong side of the ledger.

The counterargument writes itself: futures markets have run on margin for a century, and wheat prices still work. True. But wheat futures serve hedgers with offsetting physical exposure. The farmer is not a degenerate gambler; he grows the thing. Almost nobody hedging on Polymarket has an offsetting position in "who wins the Champions League." The margin-makes-markets-work logic imports a justification from a market structure that doesn't apply.

What the CFTC actually decides here

The rulebook amendment is the real fight, not the FCM license. Selig's CFTC has so far been friendly to event contracts while pressing hard on marketing conduct. The agency could split the difference: margin for eligible contract participants — the institutional tier — and full collateral for retail. That would capture most of the liquidity benefit and most of the consumer protection at once.

Whether the commission has the appetite to draw that line, with volumes quintupling and two platforms lobbying from opposite directions, is the open question. Kalshi has already shown it will petition the agency against its rival when convenient. Regulation by competitor complaint is not a stable equilibrium.

What's certain is that the innocent phase of prediction markets is over. The fully-collateralized, information-first, nobody-can-blow-up framing was lovely while it lasted, and it lasted about as long as it took the venues to get big enough to want leverage. The next election cycle will be traded on margin, by retail, at scale, on federally regulated exchanges.

When the first margin-call wave hits the morning after a surprise result, we'll find out what the truth machine is worth to the people who built it — and whether Washington remembers that it was warned.