The Financial Times
JEMIMA KELLY
Published NOVEMBER 15 2025
Any New Yorker wanting a steer on how things might have panned out in the city’s recent mayoral election could have done so without so much as glancing at a newspaper, turning on the TV or scrolling through social media. All they had to do was to look upwards — at one of several giant digital billboards that had suddenly popped up all over the city centre, featuring some pretty attention-grabbing numbers.
“94% MAMDANI 6% CUOMO”, the billboards screamed, along with the phrase “OKAY Trade” and then a company name: Polymarket.
What was this? Was it some kind of an advert for the Democratic candidate (and eventual winner) Zohran Mamdani, or for a company called Polymarket? Did it reflect polling numbers, betting odds or something else? Were they numbers to be trusted?
What these billboards in fact showed was the implied probability of a Mamdani win, based on the price that traders — some might call them gamblers, it’s a blurred line — were paying for a binary option on one of the biggest prediction markets platforms, Polymarket. According to the way in which these markets function, if Mamdani won, a wager of 94 cents would pay out $1; if he lost, nothing.
Mamdani himself might have been quite pleased with the free publicity but these were adverts for Polymarket. And they were not the only such site advertising themselves in this way: rival site Kalshi had also been plastering live prices for prediction contracts on both the Democratic candidate and independent rival Andrew Cuomo across digital billboards in the run-up to the election.
Prediction markets are booming, with political contests playing the main role in their explosion (though sports events are rapidly gaining in popularity, too). Trading volumes on both Polymarket — which is currently in talks to raise money at a valuation of $12 to $15bn — and Kalshi hit record highs in the run-up to the New York election, recording volumes of over $3bn and over $4bn respectively in October. Last week, Google announced that it would be integrating data from the sites into its search results, “so you can ask questions about future market events and harness the wisdom of the crowds”.
And if such faith in the wisdom of financially motivated, largely young and male crowds doesn’t make you raise an eyebrow, perhaps hearing about the newest major player in the market will. Last month, Truth Social, the social media platform majority-owned by Donald Trump and his family, announced the launch of Truth Predict, a new prediction market platform, in affiliation with crypto exchange Crypto.com. Donald Trump Jr, meanwhile, is not only on the board of Truth Social’s parent company, he is also an investor in Polymarket, a member of its advisory board — and a “strategic adviser” to Kalshi.
But such conflicts of interest are not the only thing to consider. What about the possible use of prediction markets to influence the outcomes of elections? To encourage or discourage certain politicians from entering the race or delegitimise election results that have gone in a way not predicted? And who is to stop foreign actors from getting involved?
Last year, more than $3.6bn was staked on the presidential race on Polymarket, despite the fact that US users were not officially allowed to use the platform at the time. They may have circumvented this using VPNs, but users elsewhere also took a keen interest — four non-American accounts from outside the US together placed more than $30mn in wagers on Trump winning, creating a decisive swing in his favour on the platform in the weeks before the election.
The research on whether voters are more or less likely to vote for a candidate they believe will win is mixed: some studies suggest it could discourage voters from going to the ballot box by making them complacent, while others have found a “bandwagon effect” that creates more support for a candidate who appears popular.
Either way, the opportunity these platforms afford to manipulate the perception of the outcome of political events is troubling, particularly at a time in which trust in institutions — including polling companies — is low and it can no longer be assumed that the legitimate election results will be accepted. How much worse might the backlash from Trump’s supporters have been after the 2020 presidential election if billboards reading “94% TRUMP 6% BIDEN” had been erected all over the US?
Prediction markets, like regular betting odds, might provide a useful gauge on how events may turn out. But wariness about their findings — and about the wisdom and benevolence of the particular crowds they attract — is needed.
jemima.kelly@ft.com
Published NOVEMBER 15 2025
Any New Yorker wanting a steer on how things might have panned out in the city’s recent mayoral election could have done so without so much as glancing at a newspaper, turning on the TV or scrolling through social media. All they had to do was to look upwards — at one of several giant digital billboards that had suddenly popped up all over the city centre, featuring some pretty attention-grabbing numbers.
“94% MAMDANI 6% CUOMO”, the billboards screamed, along with the phrase “OKAY Trade” and then a company name: Polymarket.
What was this? Was it some kind of an advert for the Democratic candidate (and eventual winner) Zohran Mamdani, or for a company called Polymarket? Did it reflect polling numbers, betting odds or something else? Were they numbers to be trusted?
What these billboards in fact showed was the implied probability of a Mamdani win, based on the price that traders — some might call them gamblers, it’s a blurred line — were paying for a binary option on one of the biggest prediction markets platforms, Polymarket. According to the way in which these markets function, if Mamdani won, a wager of 94 cents would pay out $1; if he lost, nothing.
Mamdani himself might have been quite pleased with the free publicity but these were adverts for Polymarket. And they were not the only such site advertising themselves in this way: rival site Kalshi had also been plastering live prices for prediction contracts on both the Democratic candidate and independent rival Andrew Cuomo across digital billboards in the run-up to the election.
Prediction markets are booming, with political contests playing the main role in their explosion (though sports events are rapidly gaining in popularity, too). Trading volumes on both Polymarket — which is currently in talks to raise money at a valuation of $12 to $15bn — and Kalshi hit record highs in the run-up to the New York election, recording volumes of over $3bn and over $4bn respectively in October. Last week, Google announced that it would be integrating data from the sites into its search results, “so you can ask questions about future market events and harness the wisdom of the crowds”.
And if such faith in the wisdom of financially motivated, largely young and male crowds doesn’t make you raise an eyebrow, perhaps hearing about the newest major player in the market will. Last month, Truth Social, the social media platform majority-owned by Donald Trump and his family, announced the launch of Truth Predict, a new prediction market platform, in affiliation with crypto exchange Crypto.com. Donald Trump Jr, meanwhile, is not only on the board of Truth Social’s parent company, he is also an investor in Polymarket, a member of its advisory board — and a “strategic adviser” to Kalshi.
But such conflicts of interest are not the only thing to consider. What about the possible use of prediction markets to influence the outcomes of elections? To encourage or discourage certain politicians from entering the race or delegitimise election results that have gone in a way not predicted? And who is to stop foreign actors from getting involved?
Last year, more than $3.6bn was staked on the presidential race on Polymarket, despite the fact that US users were not officially allowed to use the platform at the time. They may have circumvented this using VPNs, but users elsewhere also took a keen interest — four non-American accounts from outside the US together placed more than $30mn in wagers on Trump winning, creating a decisive swing in his favour on the platform in the weeks before the election.
The research on whether voters are more or less likely to vote for a candidate they believe will win is mixed: some studies suggest it could discourage voters from going to the ballot box by making them complacent, while others have found a “bandwagon effect” that creates more support for a candidate who appears popular.
Either way, the opportunity these platforms afford to manipulate the perception of the outcome of political events is troubling, particularly at a time in which trust in institutions — including polling companies — is low and it can no longer be assumed that the legitimate election results will be accepted. How much worse might the backlash from Trump’s supporters have been after the 2020 presidential election if billboards reading “94% TRUMP 6% BIDEN” had been erected all over the US?
Prediction markets, like regular betting odds, might provide a useful gauge on how events may turn out. But wariness about their findings — and about the wisdom and benevolence of the particular crowds they attract — is needed.
jemima.kelly@ft.com
No comments:
Post a Comment