Original article by Omera Goldberg, Founder of Chaos Labs
Original translation: Pzai, Foresight News
In just a few days, Trumps odds on Polymarket soared to 52.6%, while Harriss dropped to 46.7%. However, mainstream polls show Harris leading by a narrow margin, and this 6% difference makes the two inconsistent. Is this a feedback from a real shift in the polls, or is the market manipulated?
As of September 23, it would take an estimated $40 million to flip the Trump/Harris odds from 47%/51% to 51%/47%. This shows how easily a wealthy campaign or actor can change voter perceptions, especially in the critical weeks leading up to an election.
A FiveThirtyEight poll shows Kamala Harris maintaining a narrow lead, in stark contrast to Polymarkets odds.
Election Prediction Markets
Prediction markets like Polymarket allow people to bet on the outcomes of events like elections, sports, and even cryptocurrency prices. Prediction markets have received a lot of attention recently, especially during the U.S. presidential campaign. The principle is simple: the more money bet on a candidate, the more likely they are to win. These platforms are often seen as a real-time barometer of public sentiment, with hundreds of millions of dollars in bets coming into play during major elections. But unlike traditional polls, which are designed to be a fair reflection of voters opinions, prediction markets are easily swayed by economic influences — raising serious concerns about manipulation, especially by political candidates looking to tip the odds in their favor.
How Polymarket works
In prediction markets, users buy shares based on the likelihood of an event happening, such as a “yes” or “no” on whether a candidate will win. The prices of these shares reflect the collective judgment of the market at any given moment. For example, if a “Trump will win” share is trading at $0.52, the market is showing a 52% chance of Trump winning.
Driven by supply and demand, prediction markets can react quickly to news events, making them more dynamic than traditional polls. Because of this speed, platforms like Polymarket are often seen as real-time indicators of voter sentiment, with their large trading volumes showing how users assess candidates’ chances in a volatile political environment.
Manipulation Potential
While prediction markets provide valuable insight into political sentiment, their structure makes them highly susceptible to financial manipulation. Wealthy individuals or organizations (such as political campaigns) can pour large amounts of money into the market to distort the odds in favor of a desired outcome. For example, if Trump or Harriss team wanted to boost his approval rating, they could buy a large amount of supportive shares to artificially inflate his stock price and create the illusion of rising approval ratings.
This manipulation is particularly worrisome because prediction markets are often viewed as an accurate reflection of public opinion. Skewed odds can mislead not only the media and voters, but also political insiders who rely on these markets to judge campaign momentum, thereby fueling false narratives.
U.S. political ad spending is expected to reach a record $15.9 billion in 2024, while trading volumes on Polymarket are orders of magnitude lower — highlighting the huge gap between traditional political influence budgets and prediction market positions and trading volumes.
Why is it easier to manipulate markets than opinion polls?
Traditional polls are based on scientific sampling and use rigorous methods to reduce bias. In contrast, prediction markets are driven by the circulation of capital, which means that money can be used to influence the results. Directly injecting money into a poll will not change its results, but injecting money into a prediction market can.
This makes prediction markets more susceptible to manipulation. Well-funded political campaigns can shift the odds in their favor simply by betting large sums on a candidate. This creates a misleading narrative that can influence public perception or even trigger a feedback loop where undecided voters support a candidate because they believe that candidate is likely to be the winner.
While traditional polls are not without flaws — often with errors — the key differentiator for a platform like Polymarket is its permissionless nature. Here, we can easily quantify the costs of market manipulation, making it more transparent, but also more susceptible to financial interference.
What is the cost of guiding the market?
The cost of manipulating the prediction market depends on factors such as market size, liquidity, and the stability of the current odds. In highly liquid markets, large sums of money have already been bet, and a large amount of money is required to significantly change the odds. However, in less liquid markets, even relatively small amounts of money can significantly change the odds.
Polymarket 2024 presidential election prediction market goes live on September 23
To estimate the cost of manipulation, let’s look at Polymarket’s 2024 U.S. presidential election market, where hundreds of millions of dollars are at stake. As of September 23, it would cost an estimated $40 million to flip the odds of Trump/Harris from 47%/51% to 51%/47%. This shows how easily a wealthy campaign or actor can change voter perceptions, especially in the critical weeks leading up to an election.
The U.S. presidential campaign is expensive, and the financial incentives to manipulate these markets are obvious. While such behavior can be characterized as illegal manipulation, the anonymity attached to blockchain wallets makes it difficult to track the manipulators, further complicating law enforcement.
Are the current results manipulated?
Polymarket’s election market, its largest single prediction market to date, has seen a noticeable surge in volume recently. Over the past five days, its daily volume has surged to $54.5 million, more than four times the $12.5 million daily average over the past three months. The surge coincides with a growing divergence between Polymarket and traditional polls. Before October 2, both showed Harris with a slight lead, while Polymarket now shows Trump leading with 52.8% to Harris’ 46.6%.
While this may highlight the markets ability to respond to new information, the sharp divergence from polling data also raises concerns about potential manipulation. The timing and size of the increase in trading volume suggest that financial influences may be at work, shifting market sentiment in ways that are inconsistent with broader public opinion.
Note the dramatic spike in volume on October 3rd, which coincides with a sudden change in Polymarket odds.
Election poll results source: 270 towin.com
Risk of market manipulation
Manipulating prediction markets poses several risks:
Creating False Possibilities: By skewing market odds, campaigns can create the impression that a candidate is more popular than he or she actually is, thereby influencing media coverage and voter sentiment.
Feedback loops: Voters who see a candidate with a better chance of winning may be more inclined to support them, believing they are more likely to win. This could distort the democratic process.
Reduced market credibility: If manipulation becomes more prevalent, prediction markets will lose their credibility to accurately predict events. This will weaken the role of prediction markets in predicting political outcomes and providing real-time public opinion data.
Interpret the results carefully
Prediction markets like Polymarket are a valuable and exciting way to gauge public sentiment. They provide real-time insights and allow more people to participate in discussions about major events. As prediction markets grow, we have the opportunity to improve their integrity and make them less susceptible to manipulation.
It’s important to note that, like any financial instrument, prediction markets are subject to risk. At Chaos Labs, we build products to protect on-chain applications from manipulation, and prediction markets face similar challenges. Vigilance is key to maintaining the credibility and usefulness of these markets.
