Polymarket volume inflated by ‘artificial’ activity, study finds

Article content

The volume of activity on Polymarket, one of the most popular prediction markets, has been significantly inflated by so-called wash trading in which users rapidly buy and sell the same contracts, according to a new study by Columbia University researchers.

Article content

The “artificial trading,” as the authors call it, varied over time but accounted for an average of 25 per cent of all buying and selling on Polymarket over the past three years, the researchers concluded.

Article content

Article content

The paper, which has not undergone peer review, was posted Thursday on the open-access research platform SSRN. A representative for Polymarket said the company didn’t have an immediate comment and was reviewing the study.

Article content

Article content

The authors do not suggest that Polymarket itself was responsible for the wash trading, but they point to elements of the exchange’s crypto-based structure that make it possible.

Article content

Article content

The findings land as market participants and investors are closely watching the rising trading activity on Polymarket and its closest competitors, which allow customers to bet on the outcome of everything from sports games to elections.

Article content

Advocates promote these markets as an efficient, crowd-sourced barometer of the truth, with odds derived from the traders betting on different outcomes. If some of that volume is “fictitious,” the study says, it could alter the understanding Polymarket’s relative strength in the industry and also undermine the notion that prediction markets reflect the “wisdom of a larger crowd.”

Article content

“I’m hopeful that Polymarket will welcome the analysis in our paper,” Yash Kanoria, a professor at Columbia University’s business school, and one of the paper’s four co-authors, said in an email. “Wash trading doesn’t add liquidity or information to the market, so it would seem valuable to distinguish authentic from inauthentic volume.”

Article content

Article content

Kanoria wrote the paper with another business school professor, Hongyao Ma, as well as Rajiv Sethi, a professor of economics at Barnard College at Columbia and Allen Sirolly a doctoral student at the business school.

Article content

The authors emphasize that their attempts to identify wash trading are not definitive and amount to estimates. But their findings coincide with a heated race to identify and invest in the most successful prediction market platforms. Polymarket recently announced an investment of as much as $2 billion investment from Intercontinental Exchange Inc.

Article content

Wash trading refers to the practice of deceptively entering trades without shouldering any real market risk. The same trader or group of traders typically trade among related accounts. The practice can give the appearance of generating real volume and may impact pricing, but it generally doesn’t reflect real market sentiment. United States regulators and exchanges typically view wash trades as a form of market manipulation.

Article content

The Columbia researchers said they were able to create algorithms to analyze Polymarket activity because the trading takes place on the publicly visible Polygon blockchain ledger.