Most traders end up in the red. Bloomberg analysis showed that more than 100,000 accounts on Polymarket have lost at least $1,000 since the start of 2025. This is almost twice as many as the number of users with comparable profits.
The pattern is systemic. Losses are spread across a broad user base, while profits are concentrated among a narrow group of active participants.
Most Accounts Show a Negative Result
Since the start of 2025, about 2 million wallets have been active on the platform. At the same time, almost half of users recorded results in the range from −$10 to +$10, indicating a test-like nature of participation.
However, even in this group, the outcome was weak. A significant portion of users closed trades with a loss, despite low activity.
Profit Is Concentrated Among a Small Group
The main income is generated by a limited number of accounts. According to the study, the top 1% of traders received 76.5% of the platform’s total profit.
The concentration is even higher at the top level. The top 0.1% of participants control more than half of all income, indicating strong inequality in results.
Bots Control Most of the Turnover
An additional factor is automated trading. About 5% of wallets, which appear to be bots, account for 75% of total trading volume.
This changes the market structure. The high share of algorithmic activity increases competition and makes it harder for retail participants to enter.
There are leaders within this group. About 823 accounts earned more than $100,000 each, and the total profit of active traders reached $131 million.
Frequent Trades at Extreme Prices Lead to Losses
The study revealed an important detail. Losing users are more likely to open positions at extreme prices—below $0.10 or above $0.90.
This strategy increases risk. At the same time, the most successful participants enter trades at these levels less often, which is reflected in the final result.
Retail Traders Guess More Often but Still Lose Money
An interesting finding concerns forecast accuracy. Retail participants more often choose the correct outcome of events, unlike algorithms.
The problem is the entry price. They enter trades later and at less favorable prices, so the final result remains negative.
What This Means for the Prediction Market
The data show a stable pattern. The market distributes income unevenly, and most of the profit goes to active and technically equipped participants.
This brings such platforms closer to traditional markets. User behavior and income structure are beginning to mirror classic trading models.
What's Next?
In the short term, the situation is unlikely to change. The high share of algorithmic trading and competition for liquidity will maintain the current structure.
In the long term, the question is different. Will the prediction market be able to offer mechanisms that reduce the gap between retail and professional participants, or will the model remain just as concentrated?
Read More: Bitcoin Growth Slowed After Fed Decision to Keep Rates Amid Uncertainty in the Middle East
