CFTC Reviews Kalshi's Ether Perpetuals Over $5B Trading Pattern

Kalshi launched ether perpetual futures in June 2026, extending its event-contract exchange into a crypto-derivatives product built to track ether's price indefinitely rather than settle on a single event. An independent research analyst publishing as Beni, of Stealth Neolab, flagged the trading pattern first by mining Kalshi's own public API data, not through any regulatory disclosure.

The Wall Street Journal followed with its own review, reported September 23, 2026, finding nearly one million trades on the ether perpetual market clustered around a single order size of roughly $5,500. More than a third of the market's trades over several recent weeks fell into that pattern, together accounting for over $5 billion in volume across the prior month.

On four specific days between September 16 and 20, the Journal found those uniform-sized trades made up between 48 and 58 percent of the market's total notional volume. The CFTC is now examining the underlying data before deciding whether to open a formal enforcement investigation, according to a person familiar with the matter.

The scrutiny is not just about repeated trade sizes. Kalshi's ether perpetual market showed roughly $539 million in 24-hour trading volume against only about $3.1 million in open interest, a ratio implying most of that volume closed out almost as fast as it opened rather than building any lasting position.

That pattern matters under CFTC rules because wash trading, buying and selling the same or economically equivalent position to generate volume without a genuine change in market exposure, is prohibited outright under Section 4c(a) of the Commodity Exchange Act. A market trading billions in notional volume while carrying almost no open interest is exactly the shape regulators look at when testing for it.

Kalshi has denied any coordinated wash trading and said self-trading is mechanically blocked on its platform, with a surveillance team monitoring for coordinated manipulation. The company said hundreds of distinct traders took part in the pattern, not a small group trading with itself.

Its explanation centers on its own liquidity-incentive program, which it said pays a market maker to hold fixed-size resting quotes at a set size and spread rather than rewarding raw trading volume. Kalshi said the repeated $5,500 sizing reflects faster traders repeatedly filling that market maker's standing quotes, netting the maker roughly $98,000 in taker profit over the period reviewed, and said the CFTC has not contacted the company about the activity.

The Journal's analysis named Jump Trading and Wintermute among the firms placing trades within the flagged pattern. Jump said it trades for its own account, uses self-match prevention tools, and does not coordinate its activity with other traders.

Wintermute has not been reported offering its own on-record response to the specific allegation. Neither firm has been named in any CFTC filing or enforcement document, and the review to date rests on the Journal's data analysis and Kalshi's own market-maker program disclosures rather than a completed agency finding.

This is the CFTC's first reported look at possible manipulation within Kalshi's crypto-perpetuals line specifically, distinct from the mention-markets manipulation advisory the agency's Division of Market Oversight issued on September 22, 2026, which addressed a different product category under a different theory of manipulation risk. The two reviews arrived within a day of each other but rest on separate legal questions, one about a contract category's underlying manipulability and one about whether specific trades reflect genuine market activity.

As of this writing, the CFTC has taken no formal action and made no public statement on the ether-perpetual review. Whether it proceeds to a full investigation will likely turn on whether Kalshi's market-maker explanation, and the identities and trading rationale of the firms involved, hold up under closer examination of the underlying order data.