A recent analysis of prediction markets, including platforms like Polymarket and Kalshi, reveals that a significant majority of retail traders consistently lose money, a phenomenon that financial expert Aaron Brown asserts is an inherent characteristic of such markets. This finding challenges the perception of prediction markets as a readily accessible path to profit, instead aligning them with other high-risk retail derivatives and betting venues.
"The vast majority of traders on prediction markets lose money — and that's exactly how markets like these are supposed to work," stated Aaron Brown, a former head of financial market research at AQR Capital Management, as highlighted by Bloomberg Opinion. He suggests that statistics showing widespread losses, while accurate, are "meaningless" in terms of market design, as they reflect the nature of speculative trading.
Data from various sources underscores this trend. A Bloomberg analysis found that over 100,000 accounts on Polymarket lost at least $1,000 since early 2025, nearly double the number that made similar gains. Further research co-authored by Martineau indicated that approximately 69% of Polymarket traders have lost money since 2022, with the top 1% of accounts capturing roughly three-quarters of the overall profits. Similarly, a Wall Street Journal analysis reported that more than 70% of Polymarket users are unprofitable, and Kalshi data shows around 74% of its traders post losses.
The concentration of profits often stems from automated, high-frequency trading bots. These bots, which average significantly more trades per day than human users, collectively generated $131 million in profits, largely concentrated among a small number of users. Experts suggest these bots succeed not through superior prediction, but by entering markets earlier and securing better prices. This dynamic creates an uneven playing field, where retail traders often lack the sophisticated tools or information to compete effectively.
The high loss rates in prediction markets are not unique, mirroring patterns seen in other speculative areas. For instance, 97% of Brazilian retail futures traders who persisted for over 300 days lost money, and retail equity options traders typically lose 5% to 9% per earnings-announcement trade. Unlike traditional stock markets where few assets become entirely worthless, a losing bet in a prediction market can result in a 100% loss of capital. These findings suggest that while prediction markets offer a novel way to bet on future events, they carry substantial risks for the average participant.
