Vanderbilt University researchers have flagged a set of trades on prediction markets that were placed shortly before two recent New York Times/Siena political polls became public. The timing is unusual enough that the researchers say it deserves more scrutiny.
Large Bet on Nebraska Senate Race
According to a Friday New York Times report, a Polymarket user placed almost $200,000 on Oct. 31 that Nebraska Senator Pete Ricketts would be re-elected. About an hour later, a Times/Siena poll was released showing Ricketts in the lead. The user won more than $77,000, the report said. The same user also placed similar large bets on other markets, including Kalshi, according to the report.
Researchers Call for Further Investigation
Brett Goldstein of the Vanderbilt Institute of National Security said the sequence is hard to dismiss as a random occurrence and requires further investigation. Polymarket said it spotted the trades and is investigating. Kalshi declined to comment. It is not clear whether the user had access to inside information. That uncertainty is central to the concern: prediction markets depend on participants believing prices reflect public information, not leaks.
Earlier Case Involved US Army Soldier
The report follows an earlier case that raised similar questions. Cointelegraph reported in April that a US Army soldier had been charged after making more than $400,000 by betting on a prediction market with alleged nonpublic information tied to a US operation involving then Venezuelan President Nicolás Maduro. The Department of Justice alleged that Master Sergeant Gannon Ken Van Dyke was involved in planning and execution of Operation Absolute Resolve, which led to Maduro’s capture at a residence in Caracas, Venezuela, in January. Van Dyke allegedly used that information to buy “yes” shares in related Polymarket contracts, including “Maduro out by January 31” and “Trump invokes War Powers against Venezuela by January 31,” according to the DOJ.
What Comes Next
For now, the questions are straightforward. Did the trader simply make a lucky call? Or did the person know something before the poll went public? Regulators and market operators may need to answer that, perhaps with more detailed trade records and timing analysis. Prediction markets are meant to aggregate views, but cases like this test how well they guard against traders who may hold an information edge. The New York Times report did not establish wrongdoing by the Polymarket user. Still, the researchers’ warning is likely to keep attention on how these markets monitor large trades ahead of public news events.
