The 1% of Prediction Markets: How Liquidity Providers Dominate Polymarket

A study of $67 billion in Polymarket trades shows that 76.5% of profits are captured by just 1% of users. These winners typically act as liquidity providers using limit orders, while losers tend to use market orders. The research suggests that market-making behavior, rather than insider trading, is the primary driver of these outsized gains.

Key Points

• Profits are highly concentrated, with the top 1% of traders earning 76.5% of the total gains.
• Success is driven by market microstructure behavior, specifically providing liquidity through limit orders.
• Unsuccessful traders are characterized by taking liquidity through market orders.
• There is no significant evidence that insider trading explains the performance of the most successful accounts.
• Performance persistence is weak and may be influenced by sample selection bias.

Sentiment

The community is mildly favorable toward the paper and its empirical framing, with many commenters finding the liquidity-provider explanation credible and appreciating the author's participation. Sentiment toward Polymarket and prediction markets is more skeptical: commenters repeatedly frame the platform as closer to gambling, bookmaker economics, or a venue where sophisticated participants exploit less informed users. The article itself generates more constructive debate than rejection, but the broader topic is polarizing because it touches insider risk, consumer protection, market manipulation, and wealth concentration.

In Agreement

• Wide spreads and low patience make it plausible that liquidity providers transfer value from users who take bad prices or trade impulsively.
• The sportsbook analogy resonated: someone setting or posting better odds can profit over time when counterparties accept worse prices.
• The author’s clarification reinforced that many top earners behave like market makers, while poor market making can still lose money.
• Several commenters agreed that prediction markets differ from equities because bounded event payoffs make trading gains look more like direct redistribution.
• Sports markets were seen as a natural place for skilled bettors or model-driven traders to profit because user-to-user platforms do not need to ban winners the way bookmakers do.
• Some readers thought the findings support stronger consumer warnings because many users likely enter with unrealistic expectations about profitability.

Opposed

• Some commenters argued that insider trading may be underdetected because private knowledge can be used in isolated event-specific trades rather than persistent account patterns.
• Others suggested cross-venue arbitrage could make local Polymarket market making look like forecasting skill, so the study may need matching across venues to separate those mechanisms.
• Several users pushed back on claims that sports markets are obviously mispriced, arguing that professional bettors should arbitrage those lines if the edge is real.
• Commenters questioned whether profit concentration is meaningful without a baseline simulation showing what random or power-law trading outcomes would look like.
• Some users shifted the critique from trader profits to social harm, arguing that the important question is whether prediction markets incentivize corruption or manipulation outside the platform.
• There was skepticism about platform resolution processes, with users noting that ambiguous contracts and final rulings can make the product less clean than market-price narratives imply.