Israeli Election Betting Markets Vulnerable to Manipulation, Study Finds
Translated & summarized from Davar by baba
A new study of over 1.24 million transactions on the Polymarket betting platform found that Israeli political prediction markets are highly concentrated and easily manipulated by small sums of money. Researchers identified instances where as little as $152 USD significantly altered candidate rankings, and noted that 93% of election market traders are also active in conflict betting. The study suggests these markets are susceptible to perception engineering rather than reflecting genuine public opinion, especially as election day approaches and poll publication is restricted.
The story in 6 lines · by baba
- Israeli political betting markets are highly concentrated and easily manipulated with small sums, a new study found.
- A $152 USD purchase reportedly placed candidate Gadi Eisenkot at the top of election prediction markets.
- 93% of election market traders are also active in betting on Middle East conflict events, the study revealed.
- Researchers identified suspicious profits from bets placed before military actions and ceasefire agreements.
- The study suggests these markets are vulnerable to perception engineering, not just the 'wisdom of the crowd'.
- The findings raise questions about using betting odds as a political information source, especially before elections.
A new study examining over 1.24 million transactions on the Polymarket betting platform has revealed that political prediction markets in Israel suffer from low liquidity and high concentration, allowing individual traders to significantly influence candidate rankings. Researchers identified unusual transactions that rapidly altered candidate standings and activities suggesting attempts to manipulate the market. The study was conducted by Professor Moran Ophyr from the Haifa University Faculty of Law, Professor Joshua Mitts from Columbia University Law School, and Dan Zimmerman from BETINT.
The findings are being released ahead of the October 27th Knesset elections, amidst concerns that prediction markets could become an alternative information source to opinion polls, the publication of which will be prohibited from Friday, October 23rd, until polls close. Polymarket allows users to trade contracts based on future events, including elections, with contract prices reflecting the probability of an event occurring. Unlike opinion polls, which rely on sample responses, Polymarket probabilities are determined by user investments.
However, the study found that in some cases, small sums were sufficient to alter displayed prices and influence public and media interpretations of the race. In one instance, a purchase of just $152 USD reportedly placed Gadi Eisenkot at the top of the candidate rankings on the prediction market. In total, 11 cases were found where purchases under $10,000 USD caused sharp price changes. In another case, a focused purchase of contracts worth $100,000 USD increased Benjamin Netanyahu's projected chance from 29.5% to 36%, a jump of about 6.5 percentage points. The researchers noted that the transaction cost approximately $28,000 USD above the pre-existing market price, and the impact dissipated within two hours, suggesting a potential attempt to influence the displayed price rather than merely bet on the outcome.
An additional finding concerns market participants, with 93% of trading volume in the election market originating from traders also active in markets for Middle East conflict events. The researchers also identified 16 accounts that made unusual profits from bets placed before military strikes and ceasefire agreements. Four accounts, for example, collectively profited about $2.97 million USD from bets related to an agreement between the United States and Iran, raising suspicions of insider trading, though not proving it. The study also noted 19 instances of transactions occurring late at night or on weekends, coinciding with sharp market fluctuations, which the researchers suggest could be an attempt to influence prices during periods of low trading activity.
Professor Ophyr stated, "We see a thin and fragile market that is very easy to move with very little money, alongside heavy and reciprocal influence from the media, small market moves gain their power and can lead to substantial influence due to media reports, and reports can be born out of artificial market fluctuations. When anonymous users open accounts seconds before placing a bet in the election context, and traders who bet on the war also manage the election market, this is not 'wisdom of the crowd' but a particularly sensitive market exposed to perception engineering."
The study's findings raise questions about using betting odds as a source of information on the political system, especially in the final days before an election. While new election poll results will be banned from publication, the legal status of prediction markets and the applicability of similar restrictions remain unclear. The researchers emphasize that market movements do not necessarily reflect shifts in public opinion and that reports attributing political significance to them should be examined cautiously. The findings indicate a potential vulnerability of the pricing mechanism to individual traders but do not prove that all unusual fluctuations resulted from deliberate or coordinated manipulation.
Mentioned
