Trading volume on two Polymarket election markets exceeded $600 million — Ars Technica
On the prediction market platform Polymarket, two markets related to elections in Ethiopia and Romania generated more than $600 million in trading volume combined. As Ars Technica reports, a significant portion of the activity involved contracts on unlikely candidates who did not win.
Activity on the Ethiopia market
In Ethiopia’s federal election, Prime Minister Abiy Ahmed’s Prosperity Party won 438 seats in parliament, while its closest competitor secured 13 mandates. At the same time, trading volume in contracts on Ahmed’s victory amounted to only $173,000, while contracts on the appointment of his chief of staff, Shimelis Abdisa, as the country’s next leader exceeded $41 million. Total trading volume on this market surpassed $285 million.
Former Polymarket trader Chris Park, who now heads MSR Decode, said that approximately $284 million in volume was generated by pairs of accounts that repeatedly bought and sold contracts to one another. Barnard College economics professor Rajiv Sethi called such activity a form of trading-volume manipulation.
More current news is available on the UA.News Telegram channel Telegram.
Under Polymarket’s rules, the Ethiopia market was to be settled in October, after the prime minister’s formal appointment and swearing-in. For 123 days after the election, series of unusual transactions continued on it. Polymarket representative Shana Bautista told WIRED that the platform’s team identified the activity, investigated it, and blocked the accounts involved.
Suspicions of reciprocal trades
Anomalies were also recorded on the market concerning Romania’s 2024 election. After the first vote was canceled over concerns about foreign interference, Polymarket did not close the previous market and left it available alongside a new one. An algorithm developed by Sethi and his colleagues flagged nearly all trading volume on the old market as suspicious; according to the researcher, in certain weeks the share of transactions potentially linked to wash trading reached 99.6%.
MSR Decode analyzed more than 38,000 trade cycles on the Ethiopian market after the election. More than 98% of them produced neither profit nor loss for either side, and the median contract holding time was nine seconds. The company suggested that the most likely explanation was activity farming — artificially inflating trading volumes; its participants may have expected a possible Polymarket token distribution. Sethi stressed, however, that external signs of wash trading are not always conclusive proof of this practice.
