Kalshi is facing a growing dispute over the way it reports trading activity, first in crypto perpetuals and then in its core event-contract business. The argument intensified on Sept. 21, one day after Kalshi applied to launch perpetual futures tied to U.S. single stocks.
What began as a debate over turnover in crypto perpetuals soon widened into a broader fight over how the company counts and displays volume. Critics cited open interest, trading data, repeated lot sizes and Commodity Futures Trading Commission filings, saying the numbers do not line up with natural market activity. Kalshi’s crypto lead answered that prediction markets and perpetuals are separate products, that incentive programs were publicly filed with regulators, and that nominal volume calculations should not be labeled fake simply because they follow an industry convention.
A market-share post set off the latest round
The immediate trigger was a Sept. 20 post from Kalshi crypto lead IcoBeast.eth. He shared a chart showing Kalshi crypto volume at $363.9 million and a 96.7% market share, well ahead of Polymarket’s $12.3 million.
Users in the replies quickly challenged the claim, with some saying most of the activity was wash trading. IcoBears.eth responded by asking why anyone would wash trade on a venue that charges fees.
Hours later, Beni, who describes himself as a former quant and now co-founder of Stealth Neolab, published a long thread with a blunt claim: Kalshi was faking crypto trading volume, and he said he could prove it. The post later drew more than 1.2 million views.
Later the same day, Beni wrote that if Kalshi itself participated in, directed, or knowingly tolerated such behavior, it could amount to fraud or market manipulation. He added that he had "very interesting material" that could be shown to the CFTC.
A CFTC filing on fee rebates moved to the center of the debate
Beni’s case was not limited to order-book ratios. He also pointed to the structure of Kalshi’s rules. KalshiEX LLC filed an update to its Temporary Perpetual Fee Rebate Program with the CFTC on Sept. 2, 2026. The document said the program would take effect no earlier than Sept. 16, Eastern Time, and remain in place until Dec. 31, 2026, unless revised or terminated earlier.
For crypto perpetuals, eligible taker fees could be rebated down to 0.3 basis points, while makers could receive a net 0.3 basis points. The program applied to all Self-Clearing Members, or SCMs. In Beni’s framing, if both sides of a trade were coordinated, the visible trading cost could be pushed close to zero.
The same filing also included exclusions. Trades suspected of wash trading, self-trading, prearrangement, or other abusive conduct would not qualify for rebates. If stacked incentives produced a net negative fee, payments would be reduced. The chief regulatory officer could revoke eligibility and begin disciplinary action.
That means the filing confirms the existence of a low-friction fee structure, but it does not by itself prove that rebates were paid on wash trades.
ETH-PERP figures became the main point of attack
Beni then used a specific market to explain why he was suspicious. He said Kalshi’s ETH-PERP had roughly $3.1 million in open interest, while 24-hour volume reached $538.6 million. That implies the full open position turned over about 174 times in a day, or once every 8 minutes and 18 seconds on average.
He also said Kalshi’s position leaderboard showed the largest single ETH-PERP position at only about $17,600, a figure he argued looked hard to reconcile with the reported trading scale.
Beyond the market data, Beni said Kalshi and Jump Trading had previously reached an arrangement that exchanged liquidity for equity. On that basis, he questioned whether there could be an incentive to boost platform volume metrics.
He also challenged Kalshi’s volume methodology, saying the platform uses the number of contracts traded as volume while displaying the figure with a dollar sign, which could lead users to misread the actual amount of money traded.
Kalshi’s answer: prediction markets and perpetuals are different products
IcoBeast.eth responded quickly, though he did not directly walk through the 174x turnover figure. Instead, he shifted the discussion to product definitions. He said the Artemis chart Beni first cited was about prediction-market share, not perpetual volume. Kalshi does not offer rebates on crypto prediction markets, he said, and the way contract counts and notional amounts are calculated is consistent with the approach used by prediction-market venues such as Polymarket.
He also rejected the claim that SCMs are simply market makers handpicked by Kalshi. On a CFTC-regulated designated contract market, he said, anyone who meets regulatory requirements can become a self-clearing member, and fair access is itself a regulatory requirement.
IcoBeast.eth acknowledged that perpetuals are still in an early stage. He also argued that CME Group, Hyperliquid and Binance all rely on rebates and incentives to build liquidity, with some venues even offering negative maker fees. The difference, he said, is that Kalshi must publicly file its incentive programs because it operates under U.S. regulation.
The response had two effects. One was procedural: prediction markets and perpetuals are not the same product set. In Kalshi’s glossary, volume for event contracts is defined as the number of contracts traded, while perpetuals involve margin, leverage and funding. The other effect was that it left the core market-structure questions unanswered for critics, especially around ETH-PERP: a few million dollars of open interest, hundreds of millions in daily volume, a top position below $20,000, and later claims that fixed trade sizes kept repeating.
Beni rejected the explanation and said 99% of the perpetual volume was fake. He also said he would keep digging into the prediction-market side.
The dispute then spread to Kalshi’s main business
Before the perpetuals argument had cooled, prediction-market trader @retardmode pushed the debate into Kalshi’s core event-contract operation. He said about 61% of the volume Kalshi discloses comes from parlay bets, or multi-event combinations.
His example was simple: if a user spends $1 on a parlay ticket that pays $14.1 if every leg wins, the platform records $14.1 in volume based on contract face value. He added that nearly half of those parlays contain 11 legs or more, making the chance of winning extremely low.
Using that framework, he said actual trading the previous day was about $136 million, while the public figure was $1.91 billion. He called that a way of misleading investors.
Trader @CarOnPolymarket followed with a sharper version of the same criticism: "Kalshi counts $1 wagered on a 1000x Parlay as $1000 in volume. Yesterday, their real volume was $136M and their reported volume was $1.91B. Over a month, that would imply $57B in fake volume versus $4B in real volume. No wonder they’re ‘breaking records’ every day."
That issue is not the same as the perpetuals allegation
The event-contract dispute is different from the perpetuals allegation. In prediction markets, counting nominal volume based on the $1 face value of each contract has long been common practice, and Polymarket also speaks in contract counts.
The real argument is whether Kalshi’s parlay share is unusually high, whether long parlays inflate headline volume to the point that it loses reference value, and whether the use of dollar signs in the interface makes contract counts look like cash turnover.
Beni said Kalshi’s glossary defines volume as contracts traded, yet the interface places a dollar sign next to the number. In his example, 100 "Yes" contracts priced at $0.30 would cost the user $30,000, while the interface could show $100,000. That is a dispute over presentation and accounting treatment, not the same thing as repeated lot sizes in a perpetual order book.
Combining both lines into a single claim that Kalshi is comprehensively fabricating data would tie together a volume-accounting dispute that can be checked against disclosed rules and a wash-trading allegation that has not been established by regulators.
What public data shows so far
According to defirate data, politics, sports and crypto were still the three largest categories on Kalshi by trading volume over the past 30 days.
The fight over Kalshi’s numbers now spans turnover, rebates and display conventions. Public materials confirm that Kalshi filed a temporary perpetual fee rebate program with the CFTC, and that event-contract volume is based on contract counts. Whether those mechanisms create a misleading picture of actual trading activity remains the central point of dispute.