Prediction markets may have a new insider-trading headache: weather manipulation.
On April 15, an anonymous Polymarket trader made a long shot bet that 18 degrees Celsius would not be the highest temperature in Paris on that day.
The market was based on temperature readings from a monitoring station at Charles de Gaulle International Airport.
It was a cheap bet. The prediction market showed a more than 90% chance that the maximum temperature that day would be 18 degrees Celsius. The trader put in just $119, buying more than 21,000 "No" shares for six cents apiece.
Then, at about 9 p.m., the temperature at the station unexpectedly spiked from about 16 to 22 degrees Celsius, before dropping back down.
The user, who has since deleted his username, made more than $21,000 in profit.
The trade garnered suspicion online and has since led to an official inquiry.
Elenie Alfera, a spokesperson for Météo France, the country's official meteorological agency, told Business Insider it filed a complaint with the police over suspected tampering with its automated data system.
Polymarket did not immediately respond to a request for comment. The individual Polymarket trader could not be reached.
Polymarket and Kalshi allow users to bet on the weather in major cities, including maximum temperatures and the likelihood of specific events such as rain or snowfall.
Both prediction markets have been the subject of insider trading concerns, particularly around bets based on government actions, such as the US capture of Venezuelan President Nicolás Maduro.
In response, both platforms have taken steps to beef up enforcement against insider trading. On Wednesday, Kalshi announced that it had suspended and fined three congressional candidates who had bet on their own elections.