London Business School and Yale Study Finds 3.5% of Polymarket Traders Capture Over 30% Profits

A working paper from London Business School and Yale University argues that Polymarket’s forecast accuracy reflects “the wisdom of an informed minority, not the wisdom of the crowd,” rather than collective intelligence from a broad user base.

“A new academic study finds that prediction market accuracy on Polymarket comes from a small group of informed traders, not the broad crowd of participants the industry typically credits”

Bitcoin News

The study says about 3.5% of informed traders, including market makers and skilled takers, capture over 30% of profits on prediction platforms, with the remainder funding that accuracy.

Bitcoin News

Cointelegraph reports that the authors—Roberto Gomez-Cram, Yunhan Guo, Theis Ingerslev Jensen and Howard Kung—revised the paper on April 25 and based the findings on Polymarket trades between 2023 and 2025.

Cointelegraph adds that the authors used a sign-randomization test that repeated each account’s past trades 10,000 times to simulate profit-and-loss distributions.

Cointelegraph also states that prediction markets “now consistently record more than $15 billion in monthly trading volume across markets covering everything from sports and elections to financial results and cultural events.”

The Block similarly describes the paper’s conclusion that “just 3% of traders account for most price discovery,” while the rest “provide liquidity and generate volume” but are “on the losing side” against the informed minority.

CoinDesk frames the same core result as “Only 3% of traders drive prediction markets' accuracy, not the crowd,” and says the researchers analyzed 1.72 million accounts and $13.76 billion in trading volume.

To distinguish skill from luck, the paper’s authors ran a simulation that repeatedly flipped trading directions and recalculated outcomes.

Cointelegraph says the researchers used a sign-randomization test that repeated each account’s past trades 10,000 times to simulate profit-and-loss distributions, and it reports that “The remaining majority does not produce accuracy; rather, it funds it.”

CoinDesk

Coinpaper likewise describes the same approach, saying the authors “ran a large-scale simulation, replicating each trader’s historical activity 10,000 times,” to estimate how much profit came from skill versus randomness.

CoinDesk adds that the test kept “everything the same except the direction,” using a coin-flip benchmark to determine whether actual results consistently beat chance.

CoinDesk reports that among the biggest winners by raw profit, “only 12% of the biggest winners by raw profit cleared that benchmark,” and it says “roughly 60% of 'lucky winners' become losers when their performance is checked against a separate sample of events.”

The Block provides additional detail on persistence, saying “44% of accounts classified as skilled in a training sample stayed skilled in a held-out sample,” and it contrasts that with “about 10%” in a parallel test on active mutual funds.

Bitget’s write-up similarly states that “The test found that only 12% of top earners overlapped with the skilled group,” and it reports that “About 60% of 'lucky winners' later moved into losses when tested on another sample of events.”

The paper also examines suspected insider trading and links its timing to isolated events, while arguing it does not explain overall accuracy.

“Only 3% of traders drive prediction markets' accuracy, not the crowd, study finds Researchers show market accuracy comes from a tiny group of informed traders, not broad participation”

CoinDesk

Cointelegraph says the authors acknowledged that insider trading is a “particular concern” in prediction markets and notes that platforms face less regulatory oversight than securities markets because many users are pseudonymous and contracts are narrowly defined around specific events.

Cointelegraph quotes the authors: “These features make prediction markets an attractive venue for trading on private information.”

ForkLog reports that the researchers identified 1,950 accounts that executed trades just before significant events and ceased activity after their conclusion, and it cites a case involving bets on the resignation of former Venezuelan President Nicolás Maduro.

CoinDesk adds that the paper’s case study concerns “the U.S. removal of Nicolás Maduro from power in Venezuela in January,” describing three newly created Polymarket accounts that placed unusually large bets before the price moved and collectively made more than $630,000 when the raid happened.

CoinDesk also states that “Insider trades, when they occur, move prices even more aggressively per dollar, about seven-to-12 times more than typical skilled trades,” while emphasizing that they are “rare and concentrated in a handful of events.”

Bitget’s coverage similarly says the researchers flagged “1,950 accounts” and that these accounts moved prices “7 to 12 times more per dollar than skilled traders,” but it adds that the activity was “too focused on separate events to explain overall market accuracy.”

The study’s conclusions collide with how prediction-market executives describe their platforms, and multiple outlets quote those positions directly.

The Block says the paper “takes direct aim at a marketing line that has become standard across the industry,” quoting Kalshi CEO Tarek Mansour describing prediction markets as harnessing “the power of the wisdom of the crowds.”

Cointelegraph

The Block then quotes Polymarket CEO Shayne Coplan saying, “It's the most accurate thing we have as mankind right now, until someone else creates some sort of a super crystal ball,” and it places those remarks alongside the paper’s findings.

Cointelegraph similarly describes the regulatory scrutiny around prediction markets and says the rise of these platforms has been accompanied by increased regulatory scrutiny, driven by concerns that insider traders may be using prediction markets to turn confidential information into profits.

Cointelegraph also notes that the authors acknowledged insider trading is a “particular concern” and that the platforms face less regulatory oversight because many users are pseudonymous and contracts are narrowly defined around specific events.

Coinpaper emphasizes that the study’s conclusion is direct that “most participants are not improving market accuracy, they are financing it,” and it reiterates that their losses transfer to a small, better-informed minority.

Bitget’s write-up also includes a direct quote from Polymarket CEO Shayne Coplan, stating “the most accurate thing we have as mankind right now.”

The findings arrive as regulators and lawmakers focus on prediction markets, and the coverage points to both enforcement actions and ongoing policy debate.

“Prediction markets reflect 'wisdom of an informed minority,’ not crowd: Study About 3”

Cointelegraph

Cointelegraph says the rise of prediction markets “has been accompanied by increased regulatory scrutiny,” driven by concerns that insider traders may be using prediction markets to turn confidential information into profits, and it highlights that pseudonymity and narrowly defined contracts make enforcement harder.

ForkLog

CoinDesk reports that the CFTC filed an insider trading complaint on April 23, 2026 tied to a Polymarket contract on Nicolas Maduro’s removal from power, and it frames the case as part of a broader challenge to the industry’s core claim about crowd-driven accuracy.

The Block says the paper lands “at a sensitive moment for the sector,” noting that Polymarket is “reportedly in talks to raise $400 million at a $15 billion valuation,” and it adds that “lawmakers in Washington, New York, and California have introduced bills or executive orders aimed at insider participation.”

Cointelegraph also reports that prediction markets “now consistently record more than $15 billion in monthly trading volume,” reinforcing the scale that regulators may be weighing.

Bitget adds that Polymarket is “also reportedly in talks to raise $400 million at a $15 billion valuation,” and it repeats the authors’ conclusion that Polymarket’s accuracy reflects “the wisdom of an informed minority, not the wisdom of the crowd.”

Cointelegraph cites earlier research by crypto analyst Andrey Sergeenkov, saying that “just 0.015% of traders make profits large and consistent enough to entertain walking away from their day job,” based on Polymarket users sustaining $5,000 or more over four consecutive months between April 2024 and April 1, 2026.