Columbia study finds 25% of Polymarket trading volume may be artificial

A recent Columbia University study has found that about 25% of Polymarket’s trading activity over the past three years is artificial and arranged through wash trading. 

A new study by researchers at Columbia University suggests that trading volume on Polymarket, one of the world’s most popular prediction platforms, may be significantly exaggerated due to wash trading.

The report comes as Polymarket plans to re-enter the U.S. market with the acquisition of a sports prediction platform, QCX.

Columbia researchers find traces of inflated trading on Polymarket

According to the research paper, which was published Thursday on the open-access platform SSRN, around 25% of all buying and selling on Polymarket over the past three years appears to be a result of artificial trading. 

Wash trading occurs when the same users rapidly buy and sell identical contracts to one another. It creates the illusion of heavy market activity without taking on real risk.

The study was co-authored by Yash Kanoria and Hongyao Ma from Columbia Business School, Rajiv Sethi of Barnard College, and Allen Sirolly, a Columbia doctoral student. The research was conducted on blockchain data from Polygon, where Polymarket transactions are recorded.

The four authors clarified that their results are estimates rather than definitive proof, but they believe the evidence shows a clear pattern of inflated activity. They also suggested that much of the wash trading may have been conducted by users of the platform rather than the company itself. 

Some traders, they note, may have inflated their activity in hopes of qualifying for a potential token airdrop. Polymarket’s founder hinted at a possible token launch on social media in early October. According to Sirolly, “peaks in organic trading are tied to real events, while peaks in wash trading tend to align with rumors about token issuance.”

“Wash trading doesn’t add liquidity or information to the market,” said Kanoria in an email. “It would be valuable to distinguish authentic from inauthentic volume.”

The study does not directly accuse Polymarket of anything, but it does highlight several design features on the platform that make wash trading possible, like the lack of transaction fees, the use of stablecoins, and the ability for users to create multiple anonymous blockchain wallets.

Wash trading is not new to crypto markets. A 2022 report found that up to 70% of activity on unregulated exchanges was artificial. 

Wash trading and its effect on the market 

The researchers identified signs of wash trading in 14% of the 1.26 million wallets active on Polymarket. These accounts often traded among themselves and rarely interacted with other sides of the market.

Wash trades made up roughly 60% of total volume last December, fell to about 5% in May 2025, and then rose again to around 20% in early October, according to the study.

In regulated markets, wash trading practices are generally considered a form of manipulation because they mislead investors about genuine demand. 

Interest in prediction markets has been surging due to platforms like Polymarket and its U.S. rival Kalshi Inc., allowing users to bet on real-world events. 

The researchers warn that inflated figures may affect how investors and analysts view Polymarket’s strength in relation to its competitors. There’s also the question of whether its odds truly reflect public sentiment.

Polymarket is currently preparing to re-enter the U.S. market after settling a $1.4 million case with the Commodity Futures Trading Commission (CFTC) from 2022 for operating an unregistered exchange. In July, the CFTC and the U.S. Department of Justice concluded separate investigations into whether the firm continued allowing U.S. traders afterward.

Polymarket intends to come back through its acquisition of QCX, a CFTC-regulated platform focused on sports prediction markets.

If you're reading this, you’re already ahead. Stay there with our newsletter.