On-chain records and trading history weaken suspicions of insider information in U.S. election betting, but do not establish exactly what information the bettor relied on.
The Block’s review of on-chain transfers and trading records from prediction markets has complicated the suspicious bets brought to light by The New York Times. The findings suggest the bets may have been large positions on election outcomes rather than wagers based on learning poll results in advance.
The Times reported that researchers at Vanderbilt University had flagged large election bets on Polymarket and Kalshi as suspicious. According to the researchers, one user wagered a total of $191,277 on Pete Ricketts winning the Senate race in Nebraska and earned more than $77,000. The same user wagered $622,608 on Democrats winning the Senate race in Michigan and made more than $180,000 on the bet.
The Block matched Polymarket trades to a publicly available account named “dubdubdub2.” On Kalshi, it compared the timing and amounts of deposits with trading records. The matching suggests that large bets on Ricketts winning may have been placed about 10 minutes after a Nebraska poll was published, but Kalshi records do not link the trades to specific accounts. The Block also noted that the bettor held the Michigan positions until Election Day rather than seeking to profit from short-term price movements after polls were published.
How does the bettor’s history affect the suspicions?
The review found that the account had also made large trades before the election bets. The user took positions worth hundreds of thousands of dollars in various prediction markets, including the New York mayoral race and the Super Bowl, and held them until the outcomes were determined. In the roughly two-year trading history of the linked wallets, the busiest hour was 3 a.m. U.S. Eastern Time. This analysis excluded the period when the Senate bets were placed.
The wallets also recorded tens of millions of dollars in transfer volume on decentralized finance platforms such as Aave and Pendle. In trades made on Hyperliquid just before the election bets, the user lost $783,000 within 26 hours after depositing about $2.99 million. This history shows that large, risky positions were not unusual behavior for the bettor.
Rajiv Sethi, an economics professor at Barnard College who reviewed The Block’s findings, said the trades looked like bets on election outcomes, rather than attempts to profit from short-term market movements following the publication of polls. Vincent Grégoire, a finance professor at HEC Montréal, said the large bets could have been based on forecasting models, private polls, or other sources of information. Both experts stressed that the possibility of insider information could not be definitively ruled out.
Polymarket told the Times that it had noticed the trades and was investigating the matter. Spokespeople for the Times and the Siena Research Institute also said they would review the research.
Not investment advice.