WIRED
Prediction markets are often celebrated by proponents as indicators of public sentiment, or even as "truth machines" that may provide clearer insights than traditional polling data. A notable instance was in 2024 when Kalshi successfully outperformed many pollsters in predicting the outcome of the U.S. presidential race. However, the trading activity on these platforms can also yield more complex narratives.
In a striking development, two global election markets on Polymarket's offshore platform accounted for over $600 million in trading volume, a figure significantly higher than what is typically observed in political markets. Predominantly, the transactions involved contracts associated with long-shot candidates who ultimately did not succeed.
During the summer, Ethiopian Prime Minister Abiy Ahmed's Prosperity Party achieved a decisive victory in the federal elections, securing 438 parliamentary seats, thus allowing him to embark on another five-year term. His nearest contender managed to capture a mere 13 seats. Typically, trading volumes in prediction markets align closely with the probability of certain outcomes; leading candidates attract substantial trading activity, while long shots receive much less. In this situation, however, the trading patterns starkly deviated from the norm. A mere $173,000 was wagered on Ahmed's victory, whereas more than $41 million was traded on his chief of staff, Shimelis Abdisa, becoming the next leader of Ethiopia. Cumulatively, the volume traded exceeded $285 million, predominantly involving losing candidates.
Further compounding the peculiarity are the trading dynamics. "Nearly all of it—about $284 million—came from pairs of accounts engaged in repeated transactions with one another," notes former Polymarket trader Chris Park, who now heads the prediction-market forensic analysis firm MSR Decode.
The considerable volume of these seemingly futile transactions has attracted scrutiny from experts. Rajiv Sethi, an economics professor at Barnard College who specializes in prediction markets, remarked, "It's obviously some form of volume manipulation."
In response to inquiries regarding these markets, Shana Bautista, Polymarket's head of global intelligence and investigations, informed WIRED that her team "immediately flagged this activity, investigated, and blocked the accounts from trading." However, Polymarket did not provide any follow-up details about how these accounts managed to engage in such high trading volumes prior to intervention.
The chronology of these trades added an additional layer of oddity. Polymarket's regulations indicated that the Ethiopia market would not be closed until October, following the official appointment and swearing-in of the prime minister, which meant that it remained accessible for several months even after Ahmed's victory was tacitly recognized. For 123 days post-election, there were sporadic bursts of seemingly irrational trading, prompting user commentary such as "THIS NOT RESOLVED?" and "fuckkkk."
Sethi has termed this phenomenon a "zombie market," where participants can buy and trade on events that have already concluded or that are no longer taking place.
A similar market concerning a 2024 election in Romania exhibited what Sethi characterized as "even more dramatic" trading irregularities. An initial election was annulled due to concerns regarding foreign interference, leading to another race that took place several months later. Instead of terminating the original market during this interval, Polymarket opted to keep it operational alongside the new election option.
It was during this timeframe that Sethi first discerned unusual patterns in the earlier market. An algorithm developed by Sethi and his colleagues to detect potential wash trading on Polymarket flagged nearly all of the volume in this market as suspicious. "99.6 percent of volume in this market for certain weeks has been wash trading," Sethi reported.
Wash trading refers to a form of market manipulation whereby accounts engage in repetitive buying and selling of assets to fabricate the appearance of substantial trading activity, with funds simply being exchanged back and forth. This practice is forbidden on regulated financial exchanges in the United States and is expressly prohibited by Polymarket's terms of service. Previous research by Sethi and his team uncovered extensive evidence of wash trading on Polymarket's offshore platform.
While wash trading might elucidate the occurrences in the Ethiopian and Romanian political markets, the anonymity of the traders sparks uncertainty. "What might seem like wash trading isn't always clear-cut," warns Sethi.
MSR Decode conducted a comprehensive analysis of the Ethiopian market for WIRED, concluding that the most plausible explanation for the anomalous trading activity is "activity farming," where traders aim to artificially inflate their trading volumes. The firm posits that traders may have sought to boost their volume in anticipation of a potential cryptocurrency token launch by Polymarket. In 2025, Polymarket's chief marketing officer, Matthew Modabber, indicated to a crypto-themed podcast that "there would be a token" but provided no further details. Sethi also considers it plausible that traders could be motivated by expectations of favorable "token allocation" for higher-volume accounts.
Such behavior has been observed before in the cryptocurrency sector. "These are textbook examples of how market participants engage in farming a potential airdrop," stated Andre Guettler, director of Ulm University's Institute of Strategic Management and Finance, who studies the architecture of prediction markets.
MSR Decode's analysis of the Ethiopian election market indicated it was highly improbable that the trading volume could be attributed to sophisticated traders employing algorithms for profit maximization. The review of over 38,000 "trade loops" occurring post-election (instances where individual wallets transacted with one another in rapid succession) revealed that over 98 percent resulted in no net loss or gain for either side involved. The median duration that accounts held the contracts before reselling them was a mere nine seconds.
These atypical trading patterns are not exclusive to Polymarket. Earlier this year, Barron's reported that much of the trading activity on markets forecasting the 2028 U.S. presidential election was centered around improbable candidates such as Elon Musk and New York City Mayor Zohran Mamdani, both of whom are ineligible to run due to their foreign-born status.
Both Polymarket and its main competitor, Kalshi, have faced public scrutiny over unusual market activities in recent months. Concerns surrounding suspicious transactions in its newly launched perpetual futures market for cryptocurrency intensified on social media, prompting Kalshi to issue a blog post refuting allegations of wash trading activity. Sethi concurs with Kalshi's stance, arguing that unusual trading volumes could instead be attributed to market makers' incentives, providing opportunities for other traders to profit from minor price discrepancies.
Polymarket recently secured over $1 billion in funding, spearheaded by 1789 Capital, where Donald Trump Jr. serves as a partner. The current valuation of the company is approximately $21 billion, and earlier this summer, it announced that its annualized revenue had surpassed $1 billion. The substantial growth of Polymarket is often quantified through trading volumes, as this metric reflects user engagement.
The Commodity Futures Trading Commission (CFTC), which oversees prediction markets within the United States, has refrained from commenting on whether it is conducting an investigation into these election trades. The CFTC has previously looked into various suspected incidents of market manipulation on Polymarket, including cases reported exclusively by WIRED, but these generally involved potential insider trading rather than wash trading or other dubious activities.
Historically, the agency has imposed significant penalties for wash trading in different sectors. A former CFTC employee noted, "It seems plausible that the CFTC would be interested in investigating unusual activity within such a considerable market."
Share this story