Retail Traders Lose as Prediction Markets Favor Professionals, Report Finds

Prediction markets are drawing attention as a new venue for event-driven speculation, but fresh data suggests they may be a poor fit for retail users looking for steady wealth-building. As more individuals join and liquidity improves, a growing share of returns appears to accrue to professional traders and market makers—the very participants best positioned to exploit less-informed order flow.

The assessment comes from research by Citizens, which compared user performance in prediction markets with that of legal sportsbook customers. Jordan Bender, an analyst at Citizens JMP Securities, cited transaction data from analytics firm Juice Reel showing that, from July 2025 through mid-March 2026, the median return on investment (ROI) for prediction-market users was -8%, versus -5% for sportsbook users over the same period.

Performance diverges even more sharply when broken down by trading volume. Users who traded at least $500,000 recorded a median ROI of +2.6% on prediction markets, flipping into positive territory. The report said this band aligns with benchmarks associated with ‘sharp bettors’—professional or information-advantaged participants whose edge tends to persist over time.

Below that threshold, every cohort was unprofitable. The smallest participants—those trading less than $100—posted a median ROI of -26.8%, underscoring how “trying it with a small amount” can become structurally unfavorable once fees, spreads, and adverse selection are factored in. The implication is not simply that novices lose, but that the market architecture increasingly routes value from casual liquidity providers to specialists as participation broadens.

Sports betting showed losses as well, but with a different gradient. In the sportsbook data, the $500,000-plus cohort delivered a median ROI of -0.6%, while the smallest accounts came in at -29.3%. While the smallest bettors fared poorly across both formats, Citizens JMP said prediction markets exhibited faster deterioration in profitability for mid-sized and smaller users.

The report argues the key structural difference lies in who sits on the other side of the trade—the ‘counterparty’. Regulated sportsbooks actively manage risk by limiting successful customers through reduced betting limits or, in extreme cases, bans. This traditional “house-managed” model tends to push consistent winners out of the ecosystem, dampening the long-run concentration of elite counterparties.

Prediction markets, by contrast, typically impose fewer restrictions on profitable participants, allowing ‘information-rich capital’ to remain and compound within the venue. That dynamic increases the likelihood that retail traders end up trading directly against professional bettors, market makers, and high-frequency or high-volume participants. As informed counterparties repeatedly absorb less-informed orders, average retail performance can worsen even as the overall market becomes more liquid and efficient.

Two professional bettors who joined a recent Citizens JMP call reportedly framed the opportunity bluntly: prediction markets can offer a more attractive positive-return pathway for professionals precisely because retail users supply liquidity. In other words, growth in participation may simultaneously strengthen the mechanism that transfers economic value from casual users to expert operators.

Industry executives in online gambling, however, appear to view prediction markets as a limited near-term threat to sportsbook revenue. Citizens JMP’s review of comments from 2025 fourth-quarter earnings calls found DraftKings ($DKNG) CEO Jason Robins describing prediction markets as not meaningfully ‘incremental’ to the company’s existing customer base. Flutter ($FLUT) CEO Peter Jackson said he had not seen evidence of material ‘cannibalization’, while BetMGM CEO Adam Greenblatt estimated the impact on betting revenue could land in the “mid-single digits.” Citizens JMP’s own estimate was approximately 5%.

Where the competitive risk may be more significant is customer acquisition—especially among younger users. Citing Sensor Tower data, the report said roughly 24% of Kalshi users are under 25, with a median age of 31. By contrast, DraftKings and FanDuel had under-25 shares around 7%, with a median age closer to 35. Citizens JMP added that around 90% of DraftKings revenue is generated by users aged 30 and above, highlighting how incumbents skew older.

Download trends reinforce the generational split. From September 2025 through February 2026, FanDuel and DraftKings downloads fell 18% and 13% year-over-year, respectively, while Kalshi logged 6.3 million downloads during the same period, according to the report’s cited figures.

In that context, prediction markets may not immediately pull large numbers of existing sportsbook customers away, but could capture a “next generation” that has never installed a traditional sportsbook app. The broader implication is a two-speed shift: prediction markets may thrive on retail engagement and novelty, yet their underlying return dynamics appear increasingly tilted toward sophisticated participants—shaping how capital, liquidity, and competition evolve across event-driven trading platforms.