Developing Polymarket Just Shipped 20x Leverage — Everywhere Except the Country It Spent a Year Fighting to Re-Enter

Perpetual futures on bitcoin, gold and SpaceX went live Wednesday. American traders get routed to the CFTC-regulated exchange instead, and that split is the whole story.

September 4, 2026 at 4:42 PM EDT

6 min read

Editor's note: TrueEdge builds odds tools and earns affiliate commissions from licensed sportsbooks — a commercial interest that runs against the exchanges this piece covers. It is disclosed so readers can discount accordingly.

Polymarket turned on perpetual futures on Wednesday. Ten contracts at open — bitcoin, ether, solana, HYPE, gold, silver, WTI crude, the S&P 500, the Nasdaq 100 and a tracker on SpaceX equity — and 67 within hours, spanning 24 crypto pairs, 36 equity-linked markets, three indices and four commodities. Leverage runs to 20x on the majors and 10x on individual stocks. There is no expiry date. A funding rate transfers money between longs and shorts every hour to keep the contract pinned to spot, capped at 4% hourly in either direction.

Americans cannot trade any of it. They are routed to Polymarket US, the separately registered, CFTC-regulated exchange, where the product is still binary event contracts.

That geographic wall is the actual news. Everything else is a competent crypto exchange launch.

The mechanics, because the mechanics are the risk

At maximum leverage, maintenance margin sits at 50% of the leverage cap and roughly a 2.5% adverse move liquidates the position. Bitcoin moves 2.5% on an ordinary Tuesday. That is not a criticism of the design — it is arithmetic that applies to every perpetual futures venue in existence — but it is worth stating in plain numbers, because "up to 20x leverage" is marketing copy and "one bad hour and the position is gone" is the same sentence in the language traders actually experience.

Collateral settles in pUSD, an ERC-20 token on Polygon backed by USDC. Polymarket says it is rebuilding its matching engine toward 200,000 orders per second, about fifteen times its current throughput, with architecture aimed eventually past 400,000, and reports 10-to-20x improvements in p99 latency in testing. It claims the "deepest liquidity, lowest fees" among crypto perp venues without publishing a fee schedule to check that against. We asked the number to speak for itself and it declined.

Why a prediction market builds a leveraged derivatives desk

The uncharitable reading is that a company whose event-contract business is under attack in a dozen states diversified into a product nobody can call gambling. The charitable reading is that this was always the plan and the state fight is a distraction from it.

The charitable reading has better evidence. Intercontinental Exchange — the company that owns the New York Stock Exchange — put $1 billion into Polymarket in October 2025 and another $600 million on March 27 of this year, plus up to $40 million buying out existing holders. That is roughly $1.6 billion from a single strategic backer, at a valuation ICE declined to disclose, saying the terms would follow once Polymarket's fundraising completed. (Several write-ups have rounded this to "$2 billion"; the arithmetic in the releases does not support it.) Exchange operators do not write those cheques for a binary-options app. They write them for order flow, market data and matching infrastructure, and perps are the highest-volume instrument in that category on earth.

Kalshi got there first, incidentally. The CFTC signed off on its bitcoin perpetuals on May 29 of this year. Hyperliquid dominates the decentralised end of the market and has reportedly been in talks with Kraken's parent, Payward, about U.S. access. Polymarket launching perps in September 2026 is not a pivot. It is late.

The split is not evasion, and that is what makes it interesting

It would be easy, and wrong, to write that Polymarket is offering abroad what it cannot offer at home in order to dodge American regulators. The exclusion is the settlement working. Polymarket's 2022 CFTC settlement pushed it out of the United States; the amended order of designation the CFTC granted on November 25, 2025 brought it back on specific terms, permitting intermediated access through futures commission merchants with enhanced surveillance, clearing procedures and Part-16 reporting. Shayne Coplan's line at the time was that the approval let the company "operate in a way that reflects the maturity and transparency that the U.S. regulatory framework demands."

Leveraged perps are simply a different product requiring a different authorisation. CoinGape's write-up makes the point that matters: self-certification of a U.S. derivative is not approval, and the CFTC retains the ability to review or pause a contract before it trades. Polymarket did not try. That is a company reading its own paperwork correctly.

So the criticism is not that the wall is illegitimate. It is that the wall exists at all, and what it implies. A firm arguing in federal court that its sports contracts are financial instruments deserving federal preemption has now demonstrated, with a product launch, that it knows exactly where the line between a regulated derivative and an unregulated one sits, and that it can build on either side of it at will. Every state attorney general litigating against prediction markets will screenshot this.

The best argument against everything above

Here is the strongest version of the other side, and it is strong.

Perpetual futures are boring. They have traded on offshore venues for a decade, they are the most liquid instrument in crypto, and a CFTC-registered exchange listing gold and S&P 500 perps for non-U.S. customers is doing nothing that CME does not do in a more expensive wrapper. Treating this as a moral event because the same company also lists election and sports contracts is guilt by adjacency. Polymarket kept the leveraged product out of the U.S. precisely because it respects the boundary it is accused of gaming. And if the concern is retail harm, a 20x bitcoin perp sold to a professional trader in Singapore is not the American consumer-protection question — the American question is whether an 18-year-old in Texas should be able to trade a Chiefs contract, and Wednesday's launch has nothing to do with it.

Most of that lands. Two things survive it.

First, the funding-rate cap. Four percent per hour, compounding, is a structurally expensive way to hold a position, and it is expensive in a way that binary event contracts are not. An event contract has a terminal value of one or zero on a date; a perp bleeds continuously until it is closed. Those are different risk objects being sold by the same brand to overlapping populations who learned to trade on the first one.

Second, the sports business is not a sideline. RotoWire projects $36.8 billion of NFL trading volume on prediction markets this season. The exchanges are not derivatives venues that happen to list football. Football is the volume.

What we do not know

Polymarket has not published fees, its actual realised throughput, or the size of the book on day two. We do not know the valuation ICE invested at, because ICE says so itself. We do not know whether Polymarket will seek U.S. approval for perps, and we are not going to guess — the honest answer is that the company has an obvious commercial reason to and an obvious legal reason to wait until the Supreme Court decides whether states can regulate its other business at all.

What we can say is that on Wednesday a company at the centre of the most consequential gambling-law fight in a generation shipped a product that makes its legal argument easier to make and its consumer-protection story harder to tell, and put an ocean between the two halves of itself. That is a strategy. It is also a tell.

Watch

Kalshi's chief executive on CNBC making the demand case for perpetual futures on a prediction-market venue.