Analysis of Polymarket found roughly $600 million in trading volume on election markets after those elections were already decided, according to reporting by WIRED.
The pattern does not look like trading. Looking at the Ethiopian election market, MSR Decode analysed more than 38,000 "trade loops" — individual wallets buying from and selling to each other in quick succession — and found over 98% resulted in neither side gaining or losing money. The median time an account held a contract before selling was nine seconds.
That is the signature of wash trading: volume generated without economic purpose, which inflates the appearance of activity on a market.
It is not isolated. Earlier this year Barron's reported that much of the activity on markets predicting the 2028 US presidential election centred on unlikely candidates such as Elon Musk and New York mayor Zohran Mamdani, neither of whom is eligible to run. Kalshi has also faced wash trading allegations on its perpetual futures market, which it denies.
The CFTC, which regulates prediction markets in the US, declined to say whether it is investigating.
Why it matters
- Volume is the metric prediction markets use to demonstrate growth. If it can be manufactured at near-zero cost, it measures very little.
- Nine-second holding periods are not a market finding a price. They are a market producing a number.
- Regulators have fined wash trading heavily in other sectors, so the exposure here is not theoretical.