Kalshi Moves to End Volume Rebate Program, Filing Cites No Reason

The rebate Kalshi wants to close has run since March 2023 ^[3]. It paid from a fixed pool split by each trader's share of qualifying volume ^[2], only on trades priced between $0.03 and $0.97, and never on perpetual futures ^[4]. Payouts on event contracts were capped at half a cent per contract per participant ^[4], or 1% of the premium on a contract trading at 50 cents ^[9]. Whatever produced more than $5 billion of repeated ether-perpetual sizes over roughly a month ^[7] was paid for some other way, if it was paid for at all. The filing does not give a reason for the shutdown ^[5]. It went to the CFTC on September 28 ^[1], six days after CoinDesk published its review of the trade tape and Kalshi published its rebuttal ^[1]. CoinDesk found that trades within $2 of $5,499 made up $7.7 million of the $13.5 million in ether perpetuals it sampled from September 17 to 20 ^[8]. Recurring $2,500 and $5,000 trades made up 54% of the bitcoin sample ^[9]. The pattern, which CoinDesk said may come from bots repeating one order size, appeared in 43 of 46 hourly samples going back to June 19 ^[10]. Beni, a co-founder of Stealth Neolab who flagged it first, cited about $539 million of 24-hour ether-perpetual volume against roughly $3.1 million of open interest ^[11]. That works out to about 174 dollars traded for each dollar of open positions ^[2]. Kalshi's answer concedes more than it denies. In a September 22 blog post the company wrote that "wash trading does not occur on Kalshi" ^[12]. It said self-matching is mechanically blocked and that the repeated sizes come from market makers paid a flat fee to keep resting bids and asks on the book, quotes that faster traders then hit over and over ^[13]. Elisabeth Diana, a spokesperson, said as of September 23 that the CFTC had not contacted Kalshi and that the pattern is normal in programs that pay for liquidity ^[14]. By the company's own account, then, the perpetual liquidity is bought ^[13]. The CFTC review ^[7] would have to settle whether the takers hitting those paid quotes are independent of the firms collecting the fee. The flat fees are a separate arrangement from the volume rebate now being closed ^[2][13]. Nor is Kalshi stepping away from incentives: a Deposit and Trading Reward Incentive Program it submitted on September 25 is in a 10-day review ^[6]. September's record holds even if every flagged perp trade is thrown out. The Block put volume at $52.98 billion through the 29th against $38.67 billion for all of August ^[15], about 37% higher ^[3]. Take out the full $5 billion under review (the reported window is roughly a month and may not match September) and the month is still about 24% above August ^[4]. The flagged prints would be about 9% of September ^[5]. The CFTC could find ordinary paid market making, the account Kalshi gives. It could instead find takers tied to the quoting firms and force a restatement of perpetual volume. Or the end of the rebate after October 13 ^[1] could pull down event-contract volume, showing how much of the larger book the program bought. I think the third outcome matters more to the price than the second, because the rebate touched event contracts and not perpetuals ^[4], and the flagged perps are a single-digit share of September ^[5]. The counter-case is that a finding of coordinated trading would cost Kalshi with the regulator it registers its programs with ^[3], in ways a volume share does not capture. If event-contract volume holds up once the rebate stops, the view that Kalshi bought a meaningful slice of its event liquidity is wrong. Kalshi is in advanced talks to raise about $1 billion at a valuation near $40 billion, with Sequoia Capital and Wellington Management weighing the lead ^[17].