The filing is nineteen pages and one word long: leverage.
Kalshi — the only CFTC-regulated prediction market operating in the United States — has formally asked the Commodity Futures Trading Commission to approve margin trading on its event contracts. The public document does not contain a stress-testing framework. It does not name a default fund. It does not specify a leverage ceiling. It asks the Commission to authorize a capability the platform has not publicly demonstrated it can survive.
That absence is the finding.
A single line of logic can unravel a thousand lies: if you are asking a regulator to let retail traders borrow against binary outcomes, you must show them what happens when those outcomes resolve against a leveraged book. Kalshi showed the ask. It did not show the answer.
I went looking for the answer anyway. What follows is what a margin-enabled event contract system actually requires, measured against what a fully-collateralized platform has ever had to build. The gap is not an implementation detail. It is the entire risk surface of the product.
Kalshi already holds a Designated Contract Market license. It settles in dollars. It routes through ACH and wire, not chains. Every position on the platform today is fully prepaid — a trader who buys a "yes" contract at $0.42 has posted $0.42, and the maximum loss is that $0.42. This is not a feature. It is the reason the platform is legal.
The 2024 election cycle changed Kalshi's trajectory. When the CFTC tried to block its election contracts, a federal court overruled the agency, and Kalshi listed them. Volume followed. The company grew from a niche curiosity into the compliance-native counterweight to Polymarket, which runs the same product class on-chain, globally, without a license.
Margin trading is the next move in that rivalry. Polymarket users effectively obtain leverage through token mechanics and offshore structuring. Kalshi cannot do that. So it is asking the CFTC for leverage the legal way — which means asking the CFTC to certify the exact machinery that makes leverage dangerous.
The timing is not accidental. The application lands during a leadership transition at the Commission, in the shadow of a court ruling that validated event contracts as financial products rather than gambling. Kalshi is testing how far the framework bends. The company spent years arguing that its contracts deserve a derivatives license instead of a gambling statute. Having won that argument, it is now asking for the privileges that come with the license — and leverage is the first and largest of them.
Here is what margin trading requires that Kalshi has never needed.
A central counterparty clearing module. Today, Kalshi is a matching engine bolted to a full-collateral ledger. Money enters, positions are held against posted cash, money exits. When a trade fills, both sides are already funded. There is no credit. There is no counterparty.
Leverage deletes that property. If a trader posts $0.10 of margin to hold a position with $1.00 of nominal exposure, the platform becomes the guarantor of the ninety cents in between. That is a central counterparty function. It requires mark-to-market valuation, intraday margin calls, and — critically — an automatic liquidation engine. None of that exists in a prepaid model. Building it is not a feature update. It is a second company stapled to the first.
The gap problem. This is where event contracts diverge from every instrument the CFTC has ever regulated.
Futures and equities have continuous pricing. An index moves in ticks. When a leveraged account breaches its maintenance margin, the liquidation engine sells into an order book that has a next price — lower, but present. Losses are bounded by the depth of the market.
Event contracts are binary. They resolve to $0.00 or $1.00. Between the last trade and resolution, there is frequently no intermediate price at all. A contract trading at $0.03 does not walk to $0.40 before it settles. It either expires worthless or pays a dollar. When it pays, the move is not a tick — it is a step function.
A liquidation engine cannot liquidate a step function. By the time the outcome is known, the loss has already happened. There is no book to sell into because the event is over. A leveraged short on the losing side of a sudden resolution is not "underwater." It is dead — and so is any margin posted behind it.
Even the call mechanism is ill-fitting. Futures margin calls are intraday and iterative — the account is topped up as the loss accumulates. Event contract losses do not accumulate. They appear whole at the moment of resolution. There is no interval in which to call for more margin, because there is no interval in which the loss is partial.
This is the crux the filing does not address. In equities, gap risk is an occasional hazard. In event contracts, gap risk is the defining property of the asset. You cannot build a margin product on a binary payoff using linear risk models and expect it to hold.
When I audited the UST de-peg in 2022, I traced $40 billion draining out of Anchor in real time and documented the exact block where the incentive mechanism inverted. The lesson was mechanical: when the design assumes a price path that the market can skip, the design is fiction. A $0.03 contract does not pass through $0.30 on its way to zero. Any risk model that draws the line between them is drawing a line that does not exist.
The model risk nobody has priced. Standard portfolio margin systems — SPAN, SPAN 2, TIMS — were built on decades of continuous price history for liquid instruments. Event contracts have none of that. They have no continuous volatility surface. Each contract is its own idiosyncratic binary with a unique resolution date and a payoff that cannot be partially hedged, because there is no underlying spot to hedge with.
Ask a risk modeler to set a margin requirement for a leveraged position on "will X happen by date Y," and they will reach for a volatility number. Where does the number come from? There is no historical distribution — the event has never resolved before. The model must be invented. An invented model can be wrong in ways no backtest will reveal, because there is no backtest to run.
The regulatory precedent is not on Kalshi's side. The CFTC has spent two decades tightening retail leverage, not loosening it. Retail forex leverage was capped at 50:1 in 2010, then cut to 20:1 on major pairs. Retail precious metals caps followed the same trajectory. The agency's operating assumption is that retail traders lose money faster at higher leverage, and its protection framework is built to slow that. A margin filing for a product class the agency only recently stopped trying to ban is asking the CFTC to reverse a twenty-year instinct.
The wallet anatomy problem — inverted. My usual method is to trace fund flows through wallet clusters until the marketing narrative collapses. I cannot do that here. Kalshi runs on fiat rails; there are no on-chain footprints to cluster. But the absence cuts both ways.
Polymarket positions are traceable. When I want to know how much leveraged directional exposure exists on a given event, I can map it. I mapped wash-trading rings in NFT markets by following circular ETH flows through five linked wallets across 10,000 transactions. The same discipline works on any transparent ledger.
On Kalshi, none of that is visible. There is no public tape, no wallet graph, no way for an outside analyst to measure concentration or model the book's tail exposure. When the platform adds leverage, that opacity migrates from an analytic inconvenience to a systemic blind spot. Regulators will be asked to certify risk limits they cannot independently verify.
The concentration tail. Prediction market liquidity is thin and maker-dominated. A handful of market makers typically carry the bulk of volume. Under a prepaid model, a market maker who is wrong loses only what it posted. Under leverage, that market maker's notional exposure can run many multiples of its capital — and if a major event resolves against it, the loss does not stop at the market maker. It reaches the counterparty that guaranteed the position.
That counterparty is Kalshi.
Operational risk compounds the rest. A startup building a clearing system under regulatory deadline will make mistakes. In traditional futures, mis-timed liquidations and settlement errors generate a steady stream of customer complaints and fines. One publicized failure at Kalshi would not merely cost money. It would cost the credibility of leveraged event contracts as a category, and hand opponents of the model their first real exhibit.
Cold eyes see what warm hearts ignore — but cold eyes also see what the pessimists miss.
The bears on this filing are right about the mechanics and wrong about the alternative. Every argument that Kalshi's clearing stack cannot support leveraged event contracts is correct. What that argument ignores is that leveraged event exposure already exists. It lives offshore. It lives in token-wrapped positions on platforms no American regulator can subpoena, no auditor can inspect, and no court can reach.
Refusing Kalshi a regulated margin product does not eliminate event leverage. It relocates it to the jurisdiction with the fewest guards. The honest question is not "should event contracts have leverage." It is "where should that leverage be surveilled."
Code does not lie, but whitepapers do — and so do objections that pretend a product does not exist because they dislike it.
The bulls also see something real: the moat. If the CFTC approves this, Kalshi's DCM license stops being a badge and becomes a wall. Any competitor that wants to offer leveraged event contracts inside the United States must replicate a multi-year approval process, a clearing system, and a regulatory relationship. That wall is built out of exactly the machinery this filing is asking to be certified. The company is not just requesting a product. It is asking the government to hand it the fence.
Watch the number, not the approval. Whether the CFTC says yes matters far less than the leverage cap it attaches — and it will attach one, because the agency's entire history with retail leverage is a history of tightening ceilings, not lifting them. If the cap lands at 2:1, Kalshi gets a bigger product and the industry gets a precedent. If it lands at 20:1, the same filing that looks like a compliance victory becomes a stress test for a clearing stack nobody has seen.
The real question is not whether event contracts can be levered. It is whether anyone can build a liquidation engine for a payoff that skips every price in between.