A Haaretz investigation has found that a handful of high-value anonymous accounts, led by one trader controlling a significant share of bets on Benjamin Netanyahu, have repeatedly moved prices on Polymarket’s Israeli election market. The findings raise broader questions about whether prediction markets reflect genuine collective judgement or can be influenced by concentrated capital, particularly in relatively small and lightly populated markets where betting odds increasingly attract public and media attention.
Prediction markets have increasingly become part of the political conversation, with many viewing them as an alternative to traditional opinion polls by reflecting not what people say they believe, but what they are prepared to risk financially. In Israel, the cryptocurrency-based betting platform Polymarket has emerged as a closely watched gauge of the country’s political future, with movements in the odds on Prime Minister Benjamin Netanyahu and his potential challengers regularly attracting news coverage. Yet an investigation by Haaretz suggests that, at least in Israel’s relatively small election market, the appearance of collective market wisdom may in some instances mask the outsized influence of a handful of anonymous traders.
The investigation, based on an analysis of thousands of transactions in the market predicting Israel’s next prime minister, found that a single anonymous account known as “25xp” now controls roughly a quarter of all bets backing Netanyahu. According to Haaretz, the account repeatedly increased its investment immediately after Netanyahu’s odds declined, effectively helping to lift or stabilise his standing on the platform during periods of weakening market sentiment.
The findings challenge one of the central assumptions surrounding prediction markets—that prices emerge from the collective judgement of thousands of independent participants. Instead, Haaretz’s analysis suggests that in certain circumstances a relatively small number of high-value traders, commonly referred to as “whales”, can materially influence prices and create the impression of stronger market confidence than may actually exist.
The investigation traces the activity of the account “25xp” back to its first appearance on Polymarket on 7 April. At 4:59 p.m. Israel time, the account invested approximately US$18,000 in a low-probability market predicting that the United States and Iran would reach a ceasefire before the end of that day. At the time, the market assigned only a four per cent probability to such an outcome. Hours later, CNN reported that “a Middle East source” had signalled an approaching agreement, after which the account reportedly sold its position for a profit of approximately US$200,000.
Following that successful trade, the account disappeared from the platform for roughly two and a half months before re-emerging to invest almost exclusively in Netanyahu’s prospects of remaining prime minister. Haaretz found that the account has invested about US$40,000 in Netanyahu’s favour and has repeatedly entered the market after declines in both Polymarket odds and conventional polling. On 21 June, substantial purchases helped push Netanyahu’s “yes” shares from 33 to 35 cents, followed by another increase from 36 to 37 cents the following day. On 26 June, shortly after another major account betting against Netanyahu sharply reduced his market value, “25xp” placed its largest single purchase, worth US$20,500, helping to offset the decline.
Haaretz raises—but does not answer—the question of whether someone close to Netanyahu, potentially with access to privileged information, could be using profits from the earlier Iran trade to support the prime minister’s standing on the platform. The newspaper presents the possibility as one explanation among others, noting that it remains impossible to determine the identity behind the anonymous account.
Polymarket itself has grown rapidly since its founding in 2020 by Shayne Coplan, who launched the platform from his New York apartment during the COVID-19 lockdowns. The platform allows users to buy and sell shares representing the probability of future events, with prices ranging from zero to one US dollar corresponding directly to the market’s implied probability. Traders may either hold their positions until an event is resolved or sell them earlier as prices fluctuate.
Within Israel’s prime ministerial market, users can wager on a range of political figures including Benjamin Netanyahu, Gadi Eisenkot, Naftali Bennett, Avigdor Lieberman, Ayelet Shaked and Moshe Feiglin. The attraction lies in the simplicity of the mechanism, allowing traders to speculate on outcomes without requiring detailed knowledge of electoral mathematics or coalition negotiations.
Yet Haaretz argues that the apparent scale of the market may itself be misleading. Although Polymarket displays trading volume approaching US$30 million for the question of Israel’s next prime minister, the newspaper found that this figure represents cumulative trading volume rather than the amount of capital currently committed. According to its analysis, the actual money sitting in the market at the beginning of the week amounted to about US$760,000—approximately 39 times lower than the displayed volume.
The newspaper’s examination also found that most participants place only very small bets. Around 80 per cent of those backing Netanyahu had invested less than US$10, while approximately three whales controlled nearly one-third of the market supporting him. Alongside “25xp”, another newly created account named “OnlyBibi2026” has become the second-largest supporter of Netanyahu, investing roughly US$13,000 exclusively in his favour.
Other prominent traders have taken the opposite position. Accounts such as “turo” and “tetrose” have wagered substantial sums against Netanyahu, although Haaretz notes important differences in their trading behaviour. Unlike “25xp”, “tetrose” has a broad portfolio spanning interest rates, cryptocurrencies and multiple political markets, and has also incurred significant losses on other Iran-related trades, making allegations of insider knowledge more difficult to sustain based on its overall activity.
The investigation also highlights substantial movements in betting on Gadi Eisenkot. Two accounts, “zzzhghnh” and “fuxfux007”, together invested nearly US$48,000 over separate periods, pushing Eisenkot’s implied probability sharply higher. Combined, the two accounts now account for roughly 36 per cent of the market supporting his candidacy. Because Eisenkot’s market has significantly fewer participants than Netanyahu’s—about 1,200 compared with approximately 4,300 bettors—large investments exert proportionally greater influence over prices.
Haaretz further identified unusual trading surrounding Naftali Bennett. Two accounts, “jingsh” and “douya-150”, simultaneously placed matching bets totalling US$164,000 against one another over a period of seven minutes. According to the newspaper, the transactions created the appearance of heavy trading activity without introducing new capital into the market, inflating reported volume while leaving Bennett’s implied probability unchanged.
Academic researchers consulted by Haaretz said the behaviour observed in the Netanyahu market combined characteristics associated with both potential insider trading and attempts to influence market prices. Professors Moran Ofir of Haifa University and Joshua Mitts of Columbia University analysed every Polymarket transaction exceeding US$10,000 between February 2024 and February 2026. Among tens of millions of transactions, they identified 211,000 that displayed characteristics commonly associated with suspicious trading, generating combined profits of US$143 million. After reviewing the activity of “25xp”, they concluded that only a handful of accounts generated most of the trading activity in the Netanyahu market during roughly half of recent trading hours, describing them as participants who “move the market, order after order”.
Psychologist Prof. Michael Gilead of Tel Aviv University told Haaretz that prediction markets generally possess strong forecasting ability, citing broader academic research showing they frequently outperform conventional opinion polls. However, he argued that Israel’s relatively small political betting market represents an important exception because modest amounts of capital can move prices significantly. In his assessment, influencing such markets could be achieved relatively inexpensively and may not even violate existing rules.
While Polymarket’s defenders argue that market manipulation is ultimately self-correcting as larger numbers of participants restore prices to their fair value, the Haaretz investigation concludes that such assumptions may not always hold in smaller markets. With trading expected to intensify as Israel’s next election approaches, the newspaper argues that questions surrounding transparency, concentrated influence and anonymous participation are likely to become increasingly significant as prediction markets continue to shape public discussion alongside traditional political polling.
