The Kalshi office erupted in screams when its cofounder and CEO, Tarek Mansour, got the call from his lawyer two months before the U.S. election. Kalshi had won. Against all odds, the small startup had beat its regulator in court, forcing the government to legalize election betting. Now, for the first time in almost a century, U.S. investors could make nearly unlimited bets on the presidential election -- and soon after, money came rolling in. Money that favored Donald Trump to prevail in November.
When Fortune visited Kalshi's SoHo headquarters in Manhattan last Friday, its newly legal presidential market was just a couple of weeks old. The Sequoia-backed company had prepared for the moment since its 2021 launch, and now Mansour boasted the platform could handle single trades of up to $100 million. The office buzzed as employees scrambled to implement the company's latest gambit -- an integration with X. Elon Musk himself had become a booster, posting a Kalshi chart putting Trump's odds at 58% with the caption, "Making progress."
Mansour wouldn't make his own prediction on whether Trump would win. "You don't need to ask that question anymore," he said. "That's the whole point." The answer was right there, on Kalshi's website, in the cold logic of cash.
The case for prediction markets goes like this: Pollsters have skin in the game but for all the wrong reasons. They're politically biased, they can be manipulated, and there are conflicts of interest. On top of that, the methodology is outdated and inaccurate.
By contrast, markets are both pure and efficient. They aggregate information persuading people to put real money on the line. They're incentivized to be truthful. "People don't lie with money," Mansour said.
It's a view increasingly shared by academics, investors, and even analysts. The prominent election prognosticator Nate Silver now advises Polymarket, Kalshi's offshore and unregulated competitor. For a growing segment of the political class, the wisdom of the gambling masses trumps the experts.
Whether you agree with the thesis or not, this is the first U.S. election to test it on a large scale. Thanks to Kalshi's victory over the Commodity Futures Trading Commission in federal court, users can now bet on an array of politically-themed contracts, from Senate races to margins of victory.
Most of all, though, people are watching -- and betting on -- the presidential result, with a majority of Kalshi users blowing off the polls' prediction of a November toss-up and betting on a Trump victory. But as electoral markets face their first real trial in modern history, critics warn that Kalshi-style betting will place further strains on democracy.
Just 28 years old, Mansour looks every bit the whiz kid tech founder -- messy hair, frenetic conversation, and a disdain for convention. He came to the interview wearing a t-shirt, and rushed out of the room to get a hoodie so that he would appear "more formal."
He founded Kalshi in 2018 with Luana Lopes Lara, who he met while studying computer science and math at MIT. They bonded as part of the campus international community -- Lopes Lara is from Brazil, while the U.S.-born Mansour lived in Lebanon until he was 17.
In the summer of 2016, while still in college, Mansour worked for Goldman Sachs' exotics desk -- a unit at the bank that manages complex financial instruments like options and swaps. There, he regularly encountered investors seeking to hedge against disruptive events like Trump getting elected or Brexit. These were large institutions -- hedge funds, family offices, and corporations -- but they still could not find a financial product to help them offset the fallout from such events.
This gave Mansour an idea. He thought of how futures contracts on commodities like grain and oil allow investors to hedge against price swings. Why not build instruments where the underlying asset is the event itself -- effectively letting investors bet on elections like they do on the price of pork bellies?
After a stint at the trading firm Citadel, Mansour and Lopes Lara attended the prestigious Y Combinator startup incubator with the idea for Kalshi. Their betting marketplace launched two years later. In late 2020, the CFTC gave them permission to launch event contracts with binary options on a limited array of questions, like whether it would rain tomorrow in New York City. Election-based betting, however, remained off-limits.
Kalshi was far from the first company to venture into the field. The U.S. has a rich history of presidential betting markets, dating back to the late 19th century when trading moved from poolrooms to financial exchanges and eventually Wall Street firms. In presidential races from 1896 through 1924, all the major New York newspapers provided near-daily price quotes from October through election day, according to a 2004 paper from the economic historians Paul Rhode and Koleman Strumpf, who argued that the rise of polling firms like Gallup led to the dwindling popularity of the betting markets, which had nonetheless done a "remarkable job" forecasting elections.
The CFTC has sought to limit electoral betting markets as they have popped up, claiming they are closer to gaming than to the other types of financial derivatives that Congress charged the agency with overseeing in 1974. The CFTC allowed limited academic use cases, like the Kalshi competitor PredictIt. The agency withdrew its tacit support in 2022 and is currently enmeshed in litigation with the project, which is a collaboration between the Victoria University of Wellington and the political tech and data provider Aristotle.
The CFTC also went after Polymarket, which launched its own prediction market in 2020 built on crypto infrastructure, forcing the Peter Thiel-backed company to move offshore and pay a $1.4 million fine.
In June 2023, Kalshi made its move, launching betting contracts over which party would control the House of Representatives and the Senate. As expected, the agency blocked the product. Kalshi filed its lawsuit just a few months later.
Kalshi's legal strategy focused on legislation that grants the CFTC power to prohibit event contracts if they involve illegal activity, terrorism, assassination, war, gaming, or a "similar activity" that it deems to be "contrary to the public interest." Kalshi argued that electoral contracts did not fit under any of these categories and served an economic interest for investors who wanted to hedge against the volatility of political outcomes.
Among its lengthy arguments in opposition, the CFTC replied that it did not want to take on the role of election cop. What would happen if investors cried fraud? Was the CFTC, a small and chronically underfunded agency, supposed to investigate?
While a new crypto company seems to sue a government agency every week, Kalshi took a gamble when it decided to challenge the CFTC in court last November. "From a pattern perspective, it never really works," Mansour said. "Now it's cool to do it, but it takes the first one with some balls to basically be like, we're going to fucking do it."
Kalshi prevailed less than a year after filing suit, with Judge Jia Cobb releasing a one-page order in early September ruling that the CFTC had to allow the platform's congressional contracts -- and more. "Kalshi's contracts do not involve unlawful activity or gaming," she wrote in her full opinion, released a week later. "They involve elections, which are neither." The agency has appealed, and experts say the case could make its way to the Supreme Court.
John Phillips, the CEO of PredictIt-backer Aristotle, said that electoral markets emerging to hedge political risk were inevitable. "What is surprising is that it continues to be this waste of taxpayer dollars trying to stop Americans from doing something that is now, thanks to the litigation, demonstrably legal," he told Fortune.
Mansour may have started Kalshi with the intent of upending U.S. financial regulation, but he insists that he's a conservative person by nature. "I became more risk on over time, but I'm pretty calculated," he told Fortune.
After his lawyer, the Jones Day partner Yaakov Roth, called Mansour to tell him that Kalshi had won, Mansour didn't want to celebrate. One of his investors, the VC firm Neo, brought a bottle of champagne to the office. "I was kind of forced into it," he said.
He probably won't have another chance for a long time. The election is in just two weeks, but Kalshi doesn't plan to pay out the bets until the inauguration in January to avoid any ambiguity in the results. Meanwhile, Kalshi is working to onboard new liquidity providers beyond the trading firm Susquehanna. Mansour added that the platform is planning to launch crypto deposits in the stablecoin USDC as soon as this week, which he said will be faster than bank transfers.
While Polymarket remains a bigger force, Mansour insists that Kalshi's compliance-based approach -- even if he had to sue to obtain that supervision -- will make Kalshi the more appealing option in the long term. He predicted that U.S. electoral betting could be a $1 trillion market within a few presidential cycles.
The more existential threat for Kalshi will be its ongoing legal battle, and relationship, with its regulator. With prediction market lines now blasted across social media channels and newspaper covers, scrutiny is also growing, including allegations of manipulation on Polymarket pushing up Trump's support. Some consumer advocacy groups argue that shadowy backers could make large bets to shift the odds and later spread doubts about the results. "We're talking about the integrity of our democracy," said Cantrell Dumas, the director of derivatives policy at Better Markets.
Mansour isn't bothered that people attack financial markets -- he argues they just don't understand them. Markets are where he finds truth.