Researchers from Stanford University have found that Polymarket five-minute Bitcoin prediction markets create incentives to manipulate spot prices around contract settlement, with 821 traders extracting a combined $8.2 million from the market. The study proposes longer settlement windows as a potential fix.
The research analyzed trading patterns on Polymarket short-term Bitcoin price prediction contracts, which resolve based on the exact price of Bitcoin at a specific five-minute window. The study found that a small group of traders was able to profit by manipulating the spot price of Bitcoin on exchanges during the settlement period, effectively turning the prediction market into a wealth transfer mechanism from retail traders to manipulators.
The findings highlight a fundamental design challenge for prediction markets that settle based on continuously traded reference assets. When the settlement window is short enough, traders with sufficient capital can influence the reference price through coordinated spot market activity, making the prediction market outcome partially dependent on their actions rather than on the underlying probability of the event.
The Stanford team estimated that the manipulation distorted Bitcoin spot price during settlement windows, creating additional costs for other market participants who were not involved in the prediction market trades. The study recommends extending settlement windows to at least one hour or using time-weighted average prices rather than instantaneous snapshots.
Polymarket has faced increasing regulatory scrutiny globally. Czech authorities recently banned the platform, and the CFTC has been asserting jurisdiction over event contracts. Despite these challenges, prediction markets continue to grow rapidly, with Q2 2026 notional volume reaching a record $113.8 billion across all platforms.
This article was adapted from Cointelegraph. Read the original here.
