Crypto: 69% of Polymarket Traders are Losing

Polymarket attracts millions of crypto traders, but the majority do not come out ahead. Galaxy analyzed 2.9 million accounts considered to be retail on the international platform. The result: 69.2% finish below their breakeven point. Together, they accumulate $338.9 million in losses. Automated accounts, on the other hand, show gains of $246.8 million.

In Brief

• 69.2% of the retail accounts studied by Galaxy end up in the red.
• Their cumulative losses amount to approximately $338.9 million.
• Automated accounts, conversely, gained nearly $246.8 million.

The Majority of Crypto Traders End Up in the Red

Galaxy relied on Polymarket's public history: orders, positions, entry prices, and payments are recorded on-chain, allowing the study to track activity over several years. The analysis retains about 2.9 million accounts whose behavior resembles that of human users.

Of this set, 69.2% are losers. This result extends a finding already observed earlier in the year. In April, another study dedicated to Polymarket and Kalshi showed that the majority of users were losing money.

However, Galaxy's new research brings an important nuance. The median loss remains low: about $3 per account. Half of the traders are between a loss of $36.64 and a gain of only $0.40.

Thus, the $338.9 million is not evenly distributed among users. A small portion of accounts concentrates the significant losses. At the first percentile, losses reach about $4,804. At the other end, the 99th percentile shows about $3,381 in profit. Many lose little. A few lose a lot.

Bots Earn $246.8 Million

The contrast becomes much more visible when Galaxy isolates automated accounts. The study identifies 125,429 of them. Together, they show approximately $246.8 million in profits. This category notably includes accounts used for market making, arbitrage, or certain automated strategies.

Not all bots are profitable, however. Galaxy notes that some of them primarily perform numerous trades to take advantage of trading rewards.

The most sophisticated profiles, however, have tools different from those of the retail trader who simply chooses "Yes" or "No" on an event.

The market has become enormous. In the second quarter, the notional volume of predictive markets reached $113.8 billion, up 48.7% over three months. Sports already represented 81% of Polymarket's volume in June.

Galaxy finds this dominance reflected in behaviors. Traders specializing in sports are the least profitable among the major categories studied. Specialists in technology and science achieve the best results.

Knowledge of the subject seems to matter. About 44.1% of traders concentrate more than 60% of their activity in a single category. Another detail: profitable traders tend to take slightly larger positions. Their median position reaches $13.96, compared to $10 for losing accounts.

Polymarket Continues to Attract More New Users

These figures come as predictive markets are no longer a marginal crypto product. Since its launch in 2020, Polymarket has recorded about 1.27 billion orders across 3.07 million wallets, for a notional value of $82.8 billion according to Galaxy. However, the study focuses on the international platform, distinct from the U.S. application of Polymarket.

New users continue to arrive through major sporting events. During the World Cup, 60% of the Polymarket users studied had never used crypto before. For many, the predictive market thus becomes a first entry point into the ecosystem. However, Galaxy observes that losing quickly pushes some users out. After a losing position, 15.2% of accounts made no new trades in the following 30 days. After a win, only 6.1% also stopped for that long.

Winning encourages continuation. Losing much less so. The final observation remains quite simple. Polymarket can aggregate information and produce useful probabilities. This does not mean that its users make money. Of the 2.9 million retail accounts studied by Galaxy, nearly seven out of ten end up in the red. And while they accumulate $338.9 million in losses, automated accounts show $246.8 million in profits.