Developing Kalshi Asked for Leverage on Everything Except the Three Categories It Is Being Sued Over

Kalshi Klear, the clearing house Kalshi owns, filed with the Commodity Futures Trading Commission on Monday for permission to stop demanding full collateral on some of its contracts. If the regulator lets it through, a qualified member would be able to hold an event-contract position while posting only part of the money it could lose.

That is an ordinary thing in derivatives. It is a new thing here. Every US prediction market today makes you put up the whole maximum loss in cash, which is the single most important reason the products have behaved themselves through two years of extraordinary growth — nobody can lose more than they funded.

The filing does not cover the whole exchange. Election contracts, economic contracts and commodity-linked contracts would become marginable. Sports contracts would not. Culture contracts would not. Mention contracts — the markets on whether a named person says a particular word — would not.

Read that list backwards and it stops looking like a risk model.

The three excluded categories are the three fights

Sports is the category 44 state attorneys general told the CFTC in July it has no authority over. It is the category behind the state cease-and-desists, the Arizona criminal referral, the New York action that pushed the CFTC into emergency relief, and the circuit split that now has the Third Circuit and the Ninth Circuit saying opposite things about federal preemption.

Mention markets are the category the CFTC's own Division of Market Oversight warned about on Monday — the same day as this filing. Release 9302-26 holds that contracts settling on a person's discrete conduct carry "a heightened risk of manipulation because their settlement turns on the discrete conduct of a person that may be neither independently generated nor externally verifiable." It is also the category with a completed enforcement action attached: on August 28 the CFTC ordered former White House teleprompter operator Gabriel Perez to disgorge $107,539.02 and pay $65,000 for trading presidential mention contracts against speeches he had read in advance.

Culture is the thinnest book on the exchange and the one that most resembles a novelty bet.

So the marginable set is: the contracts institutions already understand, priced off public data, with no live litigation. The excluded set is: everything a state attorney general or a federal enforcement lawyer has put in writing this year.

A company building a purely prudential margin model would not produce that list by coincidence.

What Kalshi is actually doing

The commercial logic is real and worth stating plainly, because it is the strongest thing in the filing. Kalshi says institutional volume rose more than 800 per cent in six months. It reports over 1,000 institutional clients and more than 1,000 live markets, and it did its first block trade in April in California carbon allowances. Its annualised volume is around $178 billion, and it accounts for north of 90 per cent of US prediction-market activity.

Full collateralisation is brutal for that customer. If you want to hold a hedge on a Fed decision six months out, posting 100 per cent of notional for six months is a capital cost most desks will simply refuse to pay. Margin fixes that, and the structure Kalshi proposed is conservative on its face: access limited to self-clearing members who "satisfy specified capital thresholds given their direct clearing relationships with Kalshi Klear," and a tiered model where the required collateral rises as a contract nears settlement — tightest exactly when the outcome is least uncertain and the incentive to walk away is highest.

Kalshi's own framing is that margin brings "some prediction contracts closer to the capital model already used throughout traditional futures and derivatives markets." That is true. It is also the point.

We do not know some things that matter here, and we are not going to pretend otherwise. We could not confirm from a primary document which rule subsection the submission was made under, what the review clock actually is, or the precise coverage standard and liquidation window the clearing house proposed. Several secondary accounts carry specific figures for these. We opened enough of them to know they do not all agree, and a clearing-house margin model is exactly the kind of procedural detail this newsroom does not take from a search summary.

The strongest version of the other side

The honest counterargument is that we are reading strategy into plumbing.

Sports event contracts mostly settle in hours. An NFL game contract opens Sunday morning and resolves Sunday evening. Margin exists to solve a duration problem — capital trapped for months — and a contract with a four-hour life has no duration problem to solve. Excluding sports may simply be an acknowledgement that margin would do nothing for it.

Mention markets were just publicly flagged by the regulator as presumptively manipulation-prone; no clearing house is going to volunteer leverage on those in the same week. And culture is too illiquid to margin safely — thin books and forced liquidation are a bad combination.

That argument is good. It is not quite good enough, for two reasons.

First, duration does not actually explain the carve-out. Kalshi lists season-long and futures-style sports markets that run for months, which is the same duration profile as the election contracts it does want to make marginable. If duration were the test, some sports contracts would pass it.

Second, and more telling: the short-duration defence makes the integrity problem worse, not better. A contract that resolves in hours, on the discrete conduct of identifiable people, is precisely the shape the CFTC just described as dangerous. Kalshi knows this. It excluded mention markets for that reason. Sports contracts on individual player props share more of that shape than the exchange's filing admits.

Why this is the more important story than the volume record

Kalshi had its biggest week ever last week and the number that got written about was $15.27 billion. This filing matters more, because volume records are an output and clearing rules are an input.

What Kalshi has proposed is a two-tier exchange: an institutional derivatives venue governed by the capital conventions of the futures industry, sitting on top of a fully collateralised retail venue that looks, to every state regulator now suing it, like a sportsbook. The company has been arguing for two years that these are one market, uniformly federal, uniformly not gambling. Its own clearing house has now drawn the line in a different place.

That line will be quoted back to it. If a state's lawyers are any good, it will be quoted back within the month.

We think the states are wrong on the law and we have said so repeatedly: criminalising a federally licensed exchange is the wrong instrument, and the Third Circuit got closer to right than the Ninth did. But a company that wants preemption to mean one market cannot simultaneously ask its regulator to treat half that market as a different animal and expect nobody to notice.

Editor's note: TrueEdge builds odds and line-shopping tools and earns affiliate commissions from licensed sportsbooks. That is a commercial interest in how this fight resolves, and readers should weigh this piece accordingly.