Kalshi customers buying multi-leg sports contracts may be losing substantially more than widely cited headline figures suggest, with new analysis indicating that profitable trades placed against parlays are masking the scale of losses suffered by conventional parlay buyers.
Bloomberg previously reported that traders had recorded roughly $294 million in net losses on Kalshi’s “combo” markets during 2026. Gambling Insider subsequently reconstructed the calculation using public trading records and found a similar overall loss rate of about nine cents for every dollar wagered. However, that figure combines two fundamentally different types of activity: customers buying parlays and traders taking positions that those parlays will fail.
Kalshi describes combos as custom combinations of individual event contracts grouped into one position. Each combination has its own order book and pays a maximum of $1 per contract if every required outcome succeeds. In practice, the product closely resembles a sportsbook parlay, although Kalshi operates it through an exchange structure where traders can take either side.
Parlay Buyers Recorded Much Heavier Losses
An examination of trading between July 14 and July 27 found a sharp difference between the two sides of the market. Customers taking the YES side — effectively buying the parlay — staked about $226 million and lost roughly $79 million. That represents approximately 35 cents lost for every dollar committed.
The period included the FIFA World Cup final, which produced an unusually severe result for parlay buyers. Removing that day reduced the estimated loss rate considerably, but buyers still lost approximately 16.6 cents per dollar during the remaining 13 days.
Meanwhile, traders taking the NO side staked about $204 million and generated approximately $43 million in profits, equivalent to around 21 cents per dollar. When the two groups are combined, the resulting net loss falls to about 8.4 cents per dollar — close to the broader year-to-date figure.
That blended number therefore does not represent the experience of a typical customer simply purchasing a long-shot multi-leg contract. It includes profitable activity from traders taking the opposite position and, in some cases, users closing existing positions before settlement.
Exchange Structure Complicates the Numbers
Unlike a conventional sportsbook, Kalshi allows users to buy contracts predicting that a combo will fail. It also records certain position exits in a way that can appear as NO-side activity.
For example, a trader selling a YES contract before settlement is effectively taking the opposite side in the transaction record. Analysis of one sample day found that around 40% of NO-side volume occurred within an hour of settlement, compared with only 4% of YES-side buyer volume, suggesting that a significant portion reflected customers cashing out existing positions rather than opening new bets against parlays.
Gambling Insider based its study on roughly 11,000 combo markets, including all high-volume markets and a systematic sample of smaller contracts. The analysis also acknowledged that public trading data cannot identify whether individual accounts belong to recreational players or professionals, meaning the YES/NO classification is an approximation of trading behaviour rather than a direct classification of users.
Research Points to Broader Parlay Pricing Concerns
Separate academic research has also identified disadvantages for customers trading multi-leg contracts on Kalshi. A July 2026 study examining approximately 23 million sports moneyline trades found that Kalshi parlays were systematically priced above the combined probability implied by their individual legs, with the pricing gap increasing as more selections were added.
The latest findings also feed into a wider debate about who benefits most from prediction markets. The Roosevelt Institute estimated in July that ordinary Kalshi users had lost hundreds of millions of dollars since the platform’s launch, including significant losses from parlay-style products. Kalshi disputed the methodology behind those claims, arguing that some sophisticated trading activity had been incorrectly classified as retail participation.
The central issue is therefore less whether the reported $294 million figure is mathematically correct and more what it represents. Once profitable NO-side transactions are separated from conventional parlay purchases, the available data suggests that customers buying Kalshi combos can face loss rates considerably higher than the headline average implies.