Prediction markets are reaching new trading milestones, but a Wall Street Journal analysis suggests the gains are flowing overwhelmingly to a small group of sophisticated participants. After examining roughly 1.6 million Polymarket wallets, the WSJ found that more than 70% of accounts finished on the losing side of trades.
The imbalance points to a market where volume growth does not necessarily mean broad user success. While platforms such as Polymarket and Kalshi have promoted prediction markets as tools for aggregating public knowledge, the trading results show a widening gap between professional operators and ordinary retail users.
Profits Are Concentrated at the Top
The WSJ analysis found that profit on Polymarket was highly concentrated. Just 0.1% of accounts captured about 67% of all profits, while the typical losing account gave up between $1 and $100. The bottom 10% of traders averaged losses near $4,000.
Kalshi showed a similar pattern. According to the report, the platform had about 2.9 unprofitable users for every profitable one, reinforcing the idea that prediction markets may reward market structure advantages as much as forecasting skill.
The data undercuts the simple “wisdom of crowds” narrative. When a small group captures most of the upside, market outcomes can reflect speed, capital and execution quality more than collective intelligence.
Algorithms and Capital Shape the Playing Field
The WSJ traced much of the imbalance to professional traders and algorithmic operators. These participants use fast data feeds, trading models and significant server capacity to enter markets earlier than casual users and secure better prices.
Joshua Della Vedova, a professor at the University of San Diego’s business school, said the success of bots comes “not from superior predictive acumen, but from their ability to enter markets earlier and secure more favorable prices.”
The report also described firms executing tens of thousands of trades per day and spending more than $200,000 annually on live data, AI agents and servers. One trading group reportedly turned a $1,000 stake into seven figures, showing how infrastructure and scale can compound returns.
Researchers cited by the WSJ emphasized that persistent profitability depends on discipline and capital, not occasional correct calls. Charles Martineau’s students were reportedly surprised to learn that the top 1% of traders captured roughly three-quarters of profits. Pat Akey, a co-author of related research, summarized the issue directly: “consistently profitable traders possess clearly defined strategies and readily available capital.”
The findings create a product and governance challenge. If casual users consistently lose to faster, better-capitalized participants, liquidity growth may come at the expense of long-term retail confidence.
Prediction platforms will face growing pressure to show whether their rules, fee structures and data access policies adequately address advantages tied to automation, latency and infrastructure. Without that, rising volume may continue to mask a market where most users participate, but few reliably profit.