Columbia Study Finds Up to 25% of Polymarket Trades May Be Artificially Inflated by Wash Activity

A new study from Columbia University has found that a significant portion of trading activity on Polymarket—one of the largest blockchain-based prediction platforms—may be artificially boosted by wash trading.

According to the research, around 25% of Polymarket’s total trading volume over the past three years stemmed from wash trades, where users buy and sell the same contracts among their own accounts to create the illusion of market activity. The study, published on the open-access platform SSRN and pending peer review, suggests that certain design features of Polymarket, including minimal transaction costs and the ease of creating multiple pseudonymous wallets, make this type of manipulation possible.

Researchers analyzed on-chain data from the Polygon blockchain and identified suspicious patterns among roughly 14% of all user wallets. The proportion of wash trades fluctuated widely—reaching as high as 60% in December last year, dipping to 5% in May, and climbing again to about 20% by October.

While the authors did not allege that Polymarket itself was involved in the activity, they warned that artificial volume could distort perceptions of the platform’s popularity and reliability as a measure of public sentiment. Wash trading, often viewed by regulators as a manipulative practice, can undermine confidence in markets that claim to reflect collective intelligence.

The study’s release comes amid growing global interest in prediction markets. Polymarket has seen a surge in use over the past year, though competitors like Kalshi—operating without blockchain transparency—are also expanding, particularly in event-based and sports markets. Researchers noted that part of the inflated activity could be linked to speculation over future token airdrops, which sometimes reward frequent traders.

Polymarket, which settled with the U.S. Commodity Futures Trading Commission in 2022 and remains closed to U.S. customers, has indicated plans to re-enter the American market once it secures a regulated exchange license.

The Columbia researchers emphasized that their estimates are preliminary but highlight the need for stronger safeguards against manipulation in decentralized financial systems. As blockchain prediction markets continue to grow in influence, experts say distinguishing genuine participation from artificial trading will be vital to maintaining trust and transparency in the industry.