Prediction markets may seem novel, but in politics and other fields, betting on the outcomes of important events has a long history.
Pre-Modern Prediction Markets
Informal prediction markets can be traced back at least a thousand years, involving countless events, including betting on military battle outcomes, gambling on the next king, and wagering on the results of Chinese imperial examinations that determined an individual's entry into the civil service system.
More formal prediction markets can be traced back at least 500 years, to the early 16th century in Italy. At that time, people used markets to predict the next papal successor and quoted odds in letters. The first formal "legislation" on prediction markets appeared in 1591, when Pope Gregory XIV stated that anyone betting on the outcome of the papal conclave would be excommunicated.
In England, the earliest recorded prediction market began in 18th-century London coffee houses. Jonathan's Coffee House (later becoming the London Stock Exchange) traded news about parliamentary scandals and prime ministerial changes in the early 18th century. Trading on these events became commonplace among the elite, with contemporary newspapers even publishing odds.
The first recorded whale in this environment was British Member of Parliament Charles James Fox. From at least 1771, he began making large bets on political events, including whether the Tea Act would be repealed. In fact, he likely also bet on the outcome of the American Revolutionary War. Ultimately, he went bankrupt, and his father had to bail him out with millions of dollars (adjusted for inflation). Some might notice a similarity to an obscure modern American political figure.
American prediction market betting can be traced back to the early 19th century. James Buchanan, who later became president, wrote that he lost three plots of land due to a bad election bet in 1816. We also have the first recorded American "gambler" from this era: John Van Buren. Then the Attorney General of New York, he recorded over 100 bets in the 1834 midterm elections, totaling $500,000 (adjusted for inflation). His father, Martin Van Buren (himself a recorded election gambler), was then the Vice President.
The more formal American prediction market center was not London's coffee houses, but billiard rooms in New York City. The first major rules dispute (or in modern gamblers' parlance: rules corruption) occurred in a billiard room. The 1876 election was a chaotic struggle. Irregularities were more numerous than Theranos' blood tests, and the final result was delayed by months. As a result, "Smoking Old" Morrissey, who ran New York's largest billiard room, decided to return everyone's bets, but with a slight difference: he kept the commission. He was a famous boxer and opponent of Bill the Butcher, so I'm not sure anyone would have objected to such an arrangement.
Like London, American election odds were frequently cited in journals. At the time, before polling was widespread, betting odds were often the best indicator of public sentiment. In fact, newspapers would sometimes publish the names and bet amounts of bettors. This could be considered the earliest prediction market leaderboard.
It wasn't until 1936 that Gallup polls replaced betting odds as a reliable indicator for journalists. Subsequently, odds reporting drastically decreased, and during and after World War II, any betting markets in New York became taboo, only to be replaced by informal personal bets decades later.
Modern Prediction Markets
Betting company wagers on elections (predecessors of peer-to-peer prediction markets) began in London in the 1960s, when Ladbrokes offered (poor) odds for the Conservative Party leadership contest. The 16/1 odds helped that candidate win. In the UK, the tradition of regularly holding election betting with (almost) no incidents continues to this day, and the UK is home to the world's largest peer-to-peer betting market, Betfair.
Betfair creates prediction markets for the most notable elections and political events (including the famous Brexit). However, the vast majority of its trading volume comes from sporting events. Overall, in the UK, betting on sports event results and political outcomes has become the norm. In any major city's betting shop, you can bet a few pounds on almost any major political event you want. If you consider yourself a great politician, you can even run for office and bet on your own victory (more than one candidate has successfully done this!).
In the United States, the situation has been variable, but mostly still relatively normal.
The Iowa Electronic Market (IEM) was launched in 1988 as an academic experiment associated with the University of Iowa. The government organization regulating futures trading, the Commodity Futures Trading Commission (CFTC), neither explicitly allowed nor prohibited such trades but sent a letter stating they would take no action as long as no one held a position over $500 in the market. This was the first website to use 0-100 pricing. If you won, you'd get $1 per share. If you lost, you'd get $0. IEM initially was very small and deliberately kept small, with only a few markets and few users. It was difficult to bet more than a few dollars on their website (let alone $500). Although it still exists today, it's more of a historical footnote than a serious market.
Intrade / Tradesports went online in 2002/2003, with partial funding from famous American billionaires Paul Tudor Jones and Stan Druckenmiller. It was a website offering peer-to-peer binary contracts, trading based on event outcomes. Similar to IEM, if you won, the contract was worth $10; if you lost, it was worth $0. The website was headquartered in Ireland and had a tacit agreement with the U.S. Commodity Futures Trading Commission (CFTC) since 2005 that if the site prevented Americans from trading traditional futures contracts like gold, oil, and other strictly monitored commodities, the U.S. government would not pursue it aggressively. It became the preferred website for political odds in the presidential elections of 2004, 2008, and 2012.
At the time, public understanding of data, statistics, and numerical analysis was not as advanced as today, so the market size for such content was smaller than for polls. Nevertheless, Intrade's CEO John Delaney, a tireless prediction market advocate, frequently appeared as a guest on the U.S. business TV show CNBC, explaining event prices on his website and more broadly promoting prediction markets. He passed away while climbing Mount Everest in 2011.
A few weeks after the 2012 election, the U.S. government attacked Intrade, claiming it violated the 2005 agreement prohibiting Americans from trading oil/gold/other commodity futures. As a former user residing in the U.S. at the time, I can personally attest that all these markets did not allow any American trading. However, instead of engaging in a lengthy and costly legal battle with the U.S. government (which would be a loss even if won), Intrade expelled all American users and declared bankruptcy a few months later.
In addition to launching the first serious prediction market allowing massive U.S. user participation, Intrade was also famous for two notable whales: the McCain "whale" and the Romney "whale". They each accumulated massive positions in their respective campaigns against Barack Obama. Unlike the French whale that would later appear on Polymarket and actively bet on Trump winning in 2024, these two gamblers lost all their money and handed their winnings to everyone who bet on Obama.
In 2010, a brief movie box office futures market called Cantor Exchange (named after its parent company Cantor Fitzgerald) received full approval from the U.S. Commodity Futures Trading Commission (CFTC). I was one of the earliest joiners. The film industry hated the idea and immediately launched an intense lobbying campaign against it. The approval lasted less than two months before being banned by the U.S. Congress. In the U.S., only two types of futures are banned: onion futures and movie box office (named so because in the 1950s, some greedy merchants bought out almost the entire U.S. onion supply and monopolized the market). An interesting fact: Howard Lutnick, who long served as head of Cantor Fitzgerald, is now the U.S. Treasury Secretary (this article is written in early 2025).
PredictIt was Intrade's successor in the U.S., using exactly the same template as the Iowa model: when they launched in 2014, they received a "no-action" letter from the U.S. Commodity Futures Trading Commission (CFTC). The website was created by a U.S. political consulting firm Aristotle in collaboration with Victoria University in New Zealand. They set a cap of $850 per position (which is the IEM cap, adjusted for inflation). PredictIt became the most frequently cited odds source in the 2016 and 2020 elections. In 2022, the CFTC revoked PredictIt's "no-action" letter because it began offering markets that the CFTC considered too gambling-oriented. These included the number of tweets a political figure posts weekly. One market was "How many tweets will Andrew Yang post this week?", which led to a call threatening Yang to post more tweets, ultimately triggering a law enforcement lawsuit. Ultimately, PredictIt struggled to escape CFTC regulation in its startup phase and paid the price. There are now rumors that if approved by the CFTC, PredictIt will attempt to reorganize, rebrand, and relaunch as a regulated site.
2020 saw the birth of modern prediction markets: Polymarket and Kalshi. Both markets were still small in scale that year.
2020 also saw the emergence of many other cryptocurrency sites with rather rough prediction markets. Several companies shared the underlying market for the presidential election: Augur, Catnip, and FTX. On these sites, you bought crypto tokens stored in your crypto wallet, worth $1 if your candidate won, and $0 if the candidate lost. All these markets were replaced by Polymarket and have almost ceased to exist, reduced to jokes or memories. It's worth noting that FTX broke the law for unrelated reasons. As a supplementary note: a friend of mine is (theoretically) deeply in debt due to a bet on Trump's 2024 victory registered on the Alameda County ledger, which is an item in the bankruptcy filing.
Current Industry and Future
As of 2025, prediction markets are dominated by two giants: Kalshi and Polymarket. Strictly speaking, they have two non-intersecting user groups, which often leads to slightly different event prices in the two markets.
Kalshi is a website/app with hundreds of markets. After a quick start with Y Combinator, Kalshi received full approval from the U.S. Commodity Futures Trading Commission (CFTC) in 2020 to trade event-based markets, excluding elections. In 2022, the CFTC commissioners voted to formally reject Kalshi's election contract application, and Kalshi subsequently sued the CFTC. In 2024, after a Supreme Court case ended "Chevron's deference" and severely weakened the power of agencies like the CFTC, the judge ruled that Kalshi could list election contracts. In 2025, Kalshi further pushed legal boundaries by allowing bets on sporting events from an event contract perspective. Currently, they are facing lawsuits from multiple U.S. states. Kalshi does not allow non-U.S. citizens to participate, and all bets are denominated in U.S. dollars.
Meanwhile, Polymarket is a cryptocurrency-based website with hundreds of markets. All transactions are conducted on-chain, and all bets are processed by smart contracts and verified by a third-party program called UMA before payment. Polymarket paid a fine to the Commodity Futures Trading Commission (CFTC) in January 2022 and agreed to prohibit US users from using its website. Although they also use cryptocurrency, bets are priced in USDC, a stablecoin operated by Circle. Cryptocurrency price fluctuations do not affect users' bets. The website is located on Ethereum's L2 called Polygon. In a rough analogy, Polymarket is like an app, Polygon is like an operating system, and Ethereum is like a phone manufacturer.
Both companies were founded by ambitious, similarly aged founders in New York. Both companies have substantial capital and numerous well-known investors and advisors. Their competition is comparable to Uber and Lyft, or Visa and Mastercard.
With political chaos in the United States (especially an unprecedented collapse and late withdrawal of a party's presidential candidate), media is more focused on betting odds to understand the world than ever before. Therefore, the trading volume of each company and the entire prediction market experienced explosive growth on the eve of the 2024 election.
One story that particularly caught public attention was my discovery and conversation with the "French whale". This mysterious gentleman became the largest prediction market bettor in history within just a few weeks. He bet tens of millions of dollars on Trump winning the 2024 election and significantly raised Trump's odds. Ultimately, his prediction was correct.
The future of these markets will likely depend on how each company competes for dominance. However, the most probable outcome is that by 2028, both companies will become giants, while smaller companies try to get a piece of the pie.
Regulatory issues remain a challenge in the United States, especially considering Kalshi's push of legal boundaries in sports and the Department of Justice's (DOJ) raid on Polymarket's founders at the end of Biden's presidency.
With the significant surge in prediction market trading volume in recent years, the demand for novel and innovative markets has never been higher. This is both an opportunity and a challenge: regulatory issues are now the industry's biggest obstacle. This will inevitably delay the next major growth peak of the market, as new users will be frustrated by the need for complex specialized knowledge to navigate the field.
In conclusion, despite historical suppression, prediction markets continue to emerge whenever there is an opportunity. When two people disagree on an important issue, the best solution is to support your argument with money. And by gathering all these arguments and money into a massive market, wisdom can be unleashed.
Final Thoughts
The future is unknowable and unpredictable.
Do not view prediction markets as a silver bullet for predicting the future, but as a better flashlight that allows us to glimpse the future. Markets are certainly more effective than expert opinions, and even better than polls, because they include polls.
As citizens of Western democracies become increasingly polarized and obtain news from increasingly biased sources, prediction markets can and will pierce through partisan rhetoric and reveal the truth.
