Polymarket International Edition: About 70% of Individual Accounts Report Losses, According to Galaxy Analysis

Key Points of This Article

• Withdrawal rate after losses is 15.2%, approximately 2.5 times higher than after wins
• The ratio of profitable accounts in tech and science is 41.2%, while in sports it is 25.1%

Distribution of Profits and Losses Across 2.9 Million Accounts

On the 1st, the research department of U.S. Galaxy published a report analyzing all on-chain payment records of the prediction market Polymarket International Edition. Among approximately 2.9 million "individual" accounts that trade at a human pace, 69.2% fell below the breakeven point, with total losses amounting to $338.9 million.

Polymarket is a prediction market where shares are bought and sold based on outcomes such as elections and sports results, settled according to the results. Since all trading counterparties are recorded as public addresses, third parties can verify externally. Will Owens, the author, stated, "Anyone who disagrees with the analysis can verify it themselves," emphasizing the difference from competitors whose unique order books cannot be verified externally.

The analysis covers all history since 2020, aggregating approximately 1.27 billion transactions, 3.07 million wallets, and an assumed principal of $82.8 billion. Accounts that placed more than 50 orders per active day were considered automated trading and excluded. The 125,429 accounts that fell into this category represented only 4.1% of the total but accounted for 80.8% of the order volume and 41% of the assumed principal, generating a total profit of $24.68 million.

The median profit and loss for individual accounts was approximately -$3. Half of the accounts fell within the range of -$36.64 to +$0.40, with most losses being small amounts. In terms of the ratio to invested funds, the median account lost about 0.5%.

The analysis focused only on the international edition, which is a separate exchange from the U.S. version operated by a CFTC-approved subsidiary.

Withdrawal After Losses, Winners Remain

The most significant difference in behavior was the continuation rate of trading. Among accounts that ended with losses, 15.2% did not trade for more than 30 days afterward. This is about 2.5 times the 6.1% of accounts that did not trade after winning. Winners tend to remain on the platform longer, while losers tend to leave quickly.

There is also a difference in risk-taking. After winning, 48.4% took on larger risk positions than before, while 44.7% did so after losing. When comparing entry prices, the tendency for winners to increase their stakes concentrated in price ranges exceeding a probability of 0.50. The median position amount for profitable accounts was $13.96, while for losing accounts it was $10. In almost all layers segmented by trading frequency, winners placed larger bets.

The median holding period was about 20 hours for profitable accounts and about 25 hours for losing accounts. However, there was no consistent trend when viewed by trading frequency layers, and no advantage of short-term trading seen in meme coin transactions was confirmed.

Lowest Profit Ratio in Sports-Specific Accounts

Over 60% of the markets traded were concentrated in a single field, termed "specialized," at 44.1%. This refers to accounts that traded in five or more classified markets. The profit ratio for specialized accounts was 28.1%, slightly below the 30.4% for diversified accounts spanning multiple fields.

Sports dragged down the overall performance, accounting for 47% of specialized accounts but having the lowest ratio of profitable accounts at 25.1%. Specialized accounts in politics and culture also fell below diversified accounts, while finance was at 36.8% and tech and science at 41.2%, exceeding diversified accounts. However, the report notes that the sample size for tech and science was small.

The report points out that the underperformance in sports is likely due to the fact that "it is hard to believe that someone trading only NFL games on Sundays is conducting actuarial analysis."

Polymarket introduced taker fees for the first time in January 2026 in the cryptocurrency price direction market, expanding it to almost all fields by the end of March. The fee for trading at a price of $0.50 is 3.5% for cryptocurrencies, 2.0% for politics, finance, and tech, and 2.5% for sports. While the median loss across all history was approximately 0.5%, now, just one taker trade in the even price range incurs several times that cost.

Regarding the structure where most individuals incur losses, the report positions it as a prerequisite that supports the accuracy of prediction markets. Funds with information require flows without information to serve as trading counterparts. Owens concluded, "Noise trading is what pays the price for predictions."

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.