Wash Trading, Zombie Markets, and Hidden Manipulation: The Unseen Chaos Roiling Top Prediction Markets

Proponents of prediction markets have long framed these platforms as far more reliable than traditional public opinion polling, branding them as near-unbiased societal barometers or even "truth machines" that deliver sharper, data-backed previews of future outcomes. The 2024 U.S. presidential race reinforced that reputation for many observers, with regulated platform Kalshi outperforming the vast majority of mainstream pollsters in calling the final election result. But below that polished public image of accuracy, trading patterns on leading offshore prediction market Polymarket are revealing deeply illogical, potentially manipulative activity that defies every unwritten rule of standard market behavior.

The $285 Million "Zombie" Ethiopian Election Market

Two 2024 global election markets hosted on Polymarket’s unregulated offshore platform blew past typical political market trading volume benchmarks, notching a combined more than $600 million in total wagers – almost all of it concentrated on long-shot candidates who held no real path to victory.

The most stark example comes from Ethiopia’s 2024 federal general election, where Prime Minister Abiy Ahmed’s Prosperity Party secured a landslide victory, taking 438 of 547 total parliamentary seats to lock in another five-year term. His closest rival won a mere 13 seats, a result that was effectively a foregone conclusion for months before the official swearing-in. Normally, prediction market trading volume aligns closely with perceived odds: clear frontrunners draw the vast majority of wagers, while little-known, low-chance candidates attract only trivial sums. That standard pattern was completely flipped for the Ethiopian race. Barely $173,000 in total was traded on Ahmed to win the premiership, while over $41 million was wagered on his own chief of staff Shimelis Abdisa – a figure who never even ran for the top office – to take the presidency. In all, total trading volume for the market topped $285 million, nearly all of it placed on impossible, losing outcomes.

What made the pattern even more eyebrow-raising? Former Polymarket trader Chris Park, who now runs prediction market forensic analysis firm MSR Decode, confirmed to WIRED that almost the entire $284 million of that sum came from pairs of accounts repeatedly buying and selling identical contracts back and forth to each other, with no actual external trading activity.

Unambiguous Manipulation, Experts Confirm

Rajiv Sethi, a Barnard College economics professor who has spent years researching prediction market structure, says the pattern is clearly deliberate: "This is obviously a form of volume manipulation." When reached for comment by WIRED, Polymarket’s Global Intelligence and Investigations head Shana Bautista said the platform "immediately flagged this activity, conducted a full investigation, and blocked the associated accounts from trading." The company did not respond to follow-up questions asking why the accounts were permitted to generate such enormous volumes of trades before the platform intervened.

Adding to the strangeness, the market remained fully open for 123 full days after Abiy Ahmed’s landslide win was confirmed, because Polymarket’s resolution rules stipulated the outcome would only be settled once the prime minister was formally sworn in that October. That months-long stretch of post-election activity turned the event into what Sethi dubs a "zombie market": a marketplace where users could keep trading contracts tied to an outcome that had already long since been settled in the real world, spurring dozens of frustrated user comments on the platform asking why the market had not yet been closed out.

A separate 2024 Romanian election market hosted on Polymarket saw even more extreme anomalies, per Sethi’s analysis. The initial planned Romanian national election was canceled entirely over documented fears of foreign interference, and a new rescheduled race was announced for several months later. Instead of shutting down the old original election market and letting users settle their positions, Polymarket left the defunct old market running alongside a new market for the rescheduled vote. An algorithm built by Sethi and his research team designed to flag suspicious wash trading activity on Polymarket found that 99.6% of all trading volume on the old, canceled Romanian election market for multi-week stretches was suspicious, qualifying as textbook wash trading.

Wash trading is a banned manipulative practice where linked accounts repeatedly pass the same assets back and forth between each other to create a false illusion of massive, widespread market activity, without any net change in the traders’ actual holdings or capital. It is strictly prohibited on all regulated U.S. financial exchanges, and explicitly banned under Polymarket’s own published terms of service. Earlier research published by Sethi and his colleagues found widespread evidence of similar wash trading patterns across multiple other markets on Polymarket’s offshore platform.

The Leading Motive: Crypto Airdrop Farming

Even with the clear wash trading patterns, the exact motivations behind the massive anomalous volumes cannot be 100% confirmed since Polymarket does not publicly disclose the identities of platform traders. But analysis of the Ethiopian market conducted by MSR Decode for WIRED points to the overwhelmingly most likely explanation: activity farming, where users artificially inflate their total reported trading volume to qualify for future unannounced platform perks.

The firm’s analysis of more than 38,000 sequential "trade loops" on the Ethiopian market after election day found that over 98% of the paired back-and-forth trades resulted in zero net profit or loss for either party, with the median time a wallet held a contract before reselling it to its paired counterparty hitting just nine seconds. The hypothesis ties back to public comments Polymarket leadership made in 2025: then-CMO Matthew Modabber told the hosts of a popular crypto podcast that the platform would be launching its own native cryptocurrency token, though he did not share additional details about distribution plans. Traders widely speculated that users with higher total historical trading volumes would receive larger free token airdrops, a common incentive structure seen across the crypto industry.

"These are textbook examples of market participants farming for a potential unannounced airdrop," noted Andre Guettler, director of Ulm University’s Institute of Strategic Management and Finance and a leading prediction markets researcher.

Regulatory Scrutiny Looms For The Fast-Growing Sector

These odd, manipulative trading patterns are far from isolated incidents on Polymarket. Earlier this year, Barron’s reported that a huge share of activity on Polymarket’s 2028 U.S. presidential election market was concentrated on ineligible long-shot candidates including Elon Musk and New York City mayor Zohran Mamdani, both of whom are foreign-born and constitutionally barred from running for the U.S. presidency.

Both Polymarket and its main U.S.-regulated rival Kalshi have faced growing public scrutiny for unusual trading activity in recent months, to the point that Kalshi was forced to publish an official blog post denying widespread wash trading allegations on its newly launched perpetual crypto futures market. Sethi has publicly pushed back against those Kalshi allegations, noting the unusual volumes there are easily explained by the platform’s existing market maker incentive programs, not deliberate manipulation.

Polymarket’s valuation has skyrocketed in recent months: it recently closed a $1 billion funding round led by 1789 Capital, a venture firm where Donald Trump Jr. is a partner, pushing its total estimated valuation to $21 billion. The platform announced earlier this summer that its annualized revenue has crossed $1 billion, a milestone it largely credits to fast-growing total trading volume, the core metric the platform uses to showcase its user activity and popularity.

The U.S. Commodity Futures Trading Commission, the federal body tasked with regulating U.S.-based prediction markets, declined to comment on whether it has opened a formal probe into the anomalous election-related trading on Polymarket. The agency has previously opened investigations into instances of potential insider trading on Polymarket, but has not publicly targeted large-scale wash trading on the platform before – a practice that typically triggers steep regulatory fines in other more established U.S. financial markets. A former CFTC staffer told WIRED, "It stands to reason that the CFTC would want full visibility into this kind of deeply irregular activity on a platform that’s now operating at multi-billion dollar scale."