The Polymarket Trader Who Knew Too Much About Iran

 

Prediction markets exist on the premise that crowds, when given financial incentives, will collectively arrive at accurate forecasts. What they are not supposed to do is reward individuals who appear to already know the answer before the question is even asked. A newly surfaced pattern of trades on Polymarket has raised exactly that concern — and the numbers involved are difficult to explain away.

An analysis conducted by Bubblemaps, a company that tracks blockchain transactions, identified a single trader who accumulated nearly $967,000 in winnings through a series of bets focused almost entirely on US and Israeli military actions against Iran. The trader’s overall win rate across these wagers sat at 83%. For bets exceeding $10,000 in size, that figure climbed to 93%. Most professional high-frequency traders, operating with sophisticated algorithms and vast resources, achieve win rates marginally above 50%.

The timing of the trades is what drew the sharpest scrutiny. Winning positions were placed hours before Israeli strikes in October 2024, hours before American airstrikes on Iranian nuclear facilities in June 2025, and hours before the joint US-Israeli surprise attack in February of this year that triggered the current war. The operations being predicted were unannounced military actions. There was no public information available at the time the bets were placed.

Suspicious but Anonymous

Blockchain technology makes transaction patterns visible while keeping the identities behind wallets anonymous. The accounts linked to this trader cannot currently be traced to a named individual, and it remains unproven whether the person behind the bets had access to classified or otherwise non-public information. What the on-chain data does show is a pattern of connected accounts, extraordinary timing, and a success rate that finance experts describe as statistically implausible under normal circumstances.

A finance professor and former member of a federal regulatory advisory board reviewed the Bubblemaps findings and said the win rates alone were enough to suggest something beyond skill or luck. The Iran trades, he noted, were concentrated in the specific events where advance knowledge would have been most valuable, rather than distributed across a broader range of prediction market categories.

Polymarket, which operates its primary platform outside US jurisdiction and had not responded to requests for comment prior to publication, announced new insider trading rules on Monday. The updated policy prohibits trades based on confidential information, bans tips from individuals with legal obligations of secrecy, and bars anyone with the ability to influence an event’s outcome from betting on related markets. Critics noted the announcement came suspiciously close to the publication of the Bubblemaps findings.

A Regulatory Gap With Real Consequences

The broader issue extends well beyond one anonymous trader. Prediction markets have grown rapidly in visibility and trading volume, particularly around geopolitical events. Concerns about so-called death markets — where users place wagers on the fates of heads of state or military figures — have drawn criticism from lawmakers on both sides of the political aisle. A bipartisan group of legislators has introduced proposals that would specifically prohibit federal officials from using non-public government information to participate in prediction market platforms.

In Israel, prosecutors have already moved forward with charges against two individuals, one of them a military reservist, for allegedly exploiting classified information to place Polymarket bets during an earlier phase of the Iran conflict.

The federal agency responsible for regulating prediction markets recently issued a reminder to approved operators that insider trading laws apply to their platforms and that enforcement action remains a possibility. Whether that reminder translates into active investigation of the pattern identified by Bubblemaps remains an open question.