Polymarket Exposed – 25% of Volume Artificial, Columbia Study Claims

Polymarket, one of crypto’s hottest prediction platforms, is under fire after a new Columbia University study claimed that nearly 25% of its total trading volume may be fake. The research suggests widespread “wash trading”, traders allegedly buying and selling to themselves to inflate activity. The revelation has sent shockwaves through the on-chain prediction market world, raising sharp questions about how much of Polymarket’s booming volume is real, and how much is smoke and mirrors.

Polymarket and the study: what researchers found?

Columbia’s team published an 80-page paper outlining a network approach to detect wash trading on public blockchains. They studied millions of transactions and identified clusters of wallets that traded mostly with each other. The algorithm flags patterns consistent with self-dealing. Overall, the researchers estimate that about 25 percent of Polymarket’s volume over the past three years fits their wash-trading definition.

The paper also shows that the practice varied over time. At its peak, wash trades made up as much as 60 percent of volume in a single month. The prevalence fell and rose again, according to the timeline in the study. Those swings line up with event cycles on the platform, the authors say.

How did the researchers detect artificial trades?

Instead of relying on private data, the team used open on-chain records and a graph algorithm. The method searches for “closed clusters”, groups of addresses that transact heavily among themselves but rarely with outside wallets. The approach isolates suspicious cycles and repetitive, low-risk trades that suggest no genuine market exposure. The technique is described in detail in the SSRN paper.

The authors stress that their method is conservative. They note that some legitimate strategies can resemble wash trading. Still, the scale and persistence of the patterns on Polymarket led them to their headline estimate.

Why does this matter for prediction markets?

Prediction markets like Polymarket claim to reflect public sentiment. Inflated volume undermines that claim. If a large share of trades is artificial, the markets may mislead users about liquidity and consensus. That can distort prices and weaken confidence in using these platforms for forecasting.

For traders navigating unpredictable markets, understanding how to protect capital during volatility is just as important. Here’s a quick guide on how to manage risk during a sudden market crash.

Moreover, wash trading can influence platform valuations and fundraising narratives. The study notes that sudden volume spikes sometimes coincided with token or project rumours, suggesting incentives to create appearance of activity. That connection raises regulatory and investor concerns.

Polymarket response and regulatory context

Polymarket has not issued a detailed public rebuttal to the study at the time of writing. Past regulatory scrutiny has shaped prediction markets’ US footprint, and wash-trading allegations are taken seriously by authorities. Wash trading is illegal in traditional markets and could draw attention from regulators if proven.

Industry observers say the study will likely prompt platforms, auditors and regulators to look more closely at on-chain market integrity. Independent verification and better on-chain monitoring tools may become standard practice.

Takeaways and next steps

The Columbia paper does not claim to prove wrongdoing by named individuals. Instead, it offers a scalable method to detect suspicious trading structures. Still, the headline number, 25 percent, demands attention. Platforms must improve transparency. Users and institutional partners should re-examine how they interpret volume metrics.

Expect more research and debate. Independent replication, platform commentaries and possible regulatory follow-ups will shape the story in the weeks ahead.

Disclaimer

This article summarises publicly available research and news reporting. It does not allege criminal conduct by specific individuals. The Columbia paper is available on SSRN and has not yet undergone peer review. Consult sources and professional advice before drawing firm conclusions. 

Get the news in a Jist. Follow Cryptojist on X and Telegram for real-time updates!