Prediction markets evolve into a global information pricing infrastructure

Article by: DWF Labs

Compiled by: Luffy, Foresight News

As early as 1503, betting on the next pope took place in Rome. By 1916, Americans had wagered approximately $211 million in 2012 U.S. dollar terms on U.S. presidential elections alone, all occurring within New York’s betting markets. On busy trading days, election-related bets surpassed stock trading volumes on Wall Street’s over-the-counter exchanges.

Since the Renaissance, people have been pricing uncertain outcomes with money, and such markets have done a remarkably good job at it. So why have mature, viable prediction markets taken so long to emerge? How did Kalshi and Polymarket finally break through? This article traces the evolution of prediction markets and explores their future direction.

Prediction markets allow users to trade shares on the outcomes of future events, with prices ranging between $0 and $1, reflecting the market’s real-time assessment of the event’s probability. Unlike sports betting, you don’t need to hold your position until the event is resolved. As market conditions and probabilities change, you can open or close your position at any time.

In theory, prediction markets could become comprehensive information markets: allowing trading on the probability of a Fed rate hike, the number of Grammys Taylor Swift will win, the temperature in Paris on February 18, and more.

Kalshi's cultural theme prediction markets

1988 was the foundational year for modern prediction markets. Robin Hanson, widely recognized as the pioneer in this field, published the first academic theories on information markets and idea futures. In the same year, three professors at the University of Iowa developed the Iowa Electronic Markets (IEM). Over five election cycles, the IEM’s probability forecasts proved more accurate than opinion polls in 74% of cases, validating Friedrich Hayek’s 1945 assertion that markets are the most efficient mechanism for aggregating collective wisdom.

Iowa Electronic Markets founders George Neumann, Forrest Nelson, and Robert Forsythe, source: NBC News

Although proof of concept had been established early on, the 2000–2010s were filled with numerous failed projects. In July 2003, the Policy Analysis Market launched by the U.S. Defense Advanced Research Projects Agency (DARPA) was shut down after just one day, as two senators accused the project, designed by Robin Hanson, of effectively creating a betting market for assassinations. The U.S. Congress banned the Hollywood Stock Exchange from transitioning into a genuine movie futures exchange. Intrade operated in Dublin for over a decade but was sued by the U.S. Commodity Futures Trading Commission (CFTC) in 2012 for offering unregistered options to U.S. users; the platform collapsed in March 2013.

The crypto industry was once seen as a solution. The launch of the Ethereum mainnet provided developers with a programmable foundational infrastructure featuring decentralization and censorship resistance—qualities that seemed ideally suited to the needs of prediction markets. But new challenges soon emerged. Augur, which launched in 2018, required users to pay high Ethereum gas fees and offered a poor user experience. Its peak user count reached only 265, plummeting to just 37 within a month.

Before 2024, numerous projects failed, commonly attributed to a harsh regulatory environment and poor product execution. Teams underestimated the intensity of regulatory scrutiny while neglecting product interfaces and user experience. However, these factors alone do not fully explain the deeper structural contradictions within prediction markets. In a widely circulated 2024 article in Works in Progress, Nick Whitaker and J. Zachary Mazlish offered a more insightful analysis.

A market that can operate sustainably requires three core participants:

• Savers: Seek long-term returns to grow their wealth
• Gambler: Seeking thrill and excitement
• Professional traders: Rely on in-depth analysis to arbitrage mispricings.

The basic form of prediction markets lacks appeal to all three of these groups. Prediction markets are zero-sum games, and after fees, they become negative-sum games, making savers completely unwilling to participate, as they require positive-sum markets to grow their wealth. Most real-world events have long settlement periods and niche topics, making it difficult to attract gamblers, who generally prefer assets with quick outcomes.

Without traders seeking to save or gamble providing counterparty liquidity, professional traders cannot find sufficient liquidity to enter the market. In the end, only professional traders remain trading against each other—this is the real-world version of the no-trade theorem: if everyone is sufficiently rational, no one is willing to act as a counterparty.

Setting aside market structure, the vast majority of topics have limited appeal to the general public. Without trading volume, professional traders have no incentive to enter and compete for slim potential profits. Of course, exceptions exist, such as sports and political topics. Whitaker and Mazlish concluded that, without external subsidies, the "everything is predictable" prediction market model cannot be scaled.

Despite the aforementioned practical limitations, prediction markets have evolved into a mature product category. They experienced a surge during the 2024 U.S. presidential election, with their projected probabilities widely cited as a factual reference. The New York Times referenced prediction market data, CNBC reported on it, and Bloomberg Terminal integrated the data directly. Total funding in this sector has exceeded $5 billion, with financing accelerating over the past 18 months.

Even if you haven’t been following this space, you’ve likely heard of the two platforms driving industry breakthroughs: Polymarket and Kalshi. Together, they account for over 90% of the industry’s trading volume, with monthly total trading volume surpassing $58 billion.

Polymarket

Polymarket was founded by Shayne Coplan in 2020. A dropout from New York University, he participated in the Ethereum ICO in 2014 and wrote to Robin Hanson in 2019 expressing his desire to bring prediction markets to life. He launched the platform from his apartment in New York City’s Lower East Side during the COVID-19 pandemic.

Iterations in crypto infrastructure—such as low-cost Layer 2 networks and stablecoins—have helped Polymarket avoid the pitfalls of early crypto prediction markets. The platform operates on Polygon (an Ethereum Layer 2 network), reducing gas fees to just a few cents. Settlements are conducted using the platform’s native stablecoin, PUSD, ensuring that $1 is always worth exactly $1, with no price volatility risk during holding periods. Trading uses a hybrid order book: off-chain matching ensures speed, while on-chain settlement guarantees trustworthiness, combining the smooth experience of centralized exchanges with the non-custodial nature of decentralized settlement.

Shayne Coplan, founder and CEO of Polymarket, source: Forbes

Polymarket adopted a "launch first, address regulatory issues later" approach, gaining greater freedom and faster iteration speeds compared to its competitors at the time. During the 2020 U.S. presidential election, the platform experienced early growth, reaching monthly trading volumes of approximately $26 million, and continued expanding through markets related to the pandemic and pop culture.

The risks of regulatory neglect eventually caught up. In January 2022, the CFTC imposed a $1.4 million fine on Polymarket and required the platform to block U.S. users. Compliance became the top priority: the platform implemented geographic blocking for the U.S., hired a former CFTC chair as an advisor, and continued operations in all other global regions. In 2023, the platform’s trading volume reached approximately $73 million—negligible compared to current volumes, but sufficient to see it through the crypto winter.

Next comes the 2024 U.S. presidential election, a highlight moment for the industry. Although officially prohibited from serving U.S. users, Polymarket became the cultural representative platform for this election. Markets related to the election generated a total trading volume of approximately $3.6 billion, and its probability forecasts for Trump’s victory proved more accurate than opinion polls and expert analyses. This surge in popularity brought prediction markets into the global spotlight.

One week after the election, the FBI raided Coplan’s apartment to investigate whether U.S. users had circumvented the 2022 ban by trading on the international version of the platform. In July 2025, the U.S. Department of Justice and the CFTC concluded their investigation without filing any charges. Shortly thereafter, Polymarket acquired QCEX, a CFTC-licensed exchange, for $1.12 billion. In October of the same year, Intercontinental Exchange (ICE), the parent company of the New York Stock Exchange, agreed to invest up to $2 billion in Polymarket at a pre-investment valuation of $8 billion; ICE also became Polymarket’s global distributor for event data. Leveraging the acquisition of QCEX, Polymarket re-entered the U.S. market in December 2025.

After the election hype subsided, sports and international geopolitical events drove continued platform growth. As of July 2026, Polymarket had completed 707.7 million trades with a total trading volume exceeding $111.9 billion. In 2024, it was unquestionably the industry leader, before being surpassed by Kalshi.

Kalshi

From its inception, Kalshi made a fundamentally different choice. Founders Tarek Mansour and Luana Lopes Lara, both MIT graduates (Lara was formerly a professional ballet dancer who performed in Swan Lake before transitioning into financial markets), founded the company in 2018. They bet that adhering to compliance from the start was more important than speed of growth.

Kalshi co-founders Tarek Mansour and Luana Lopes Lara, source: Forbes

The two founders waited nearly two years to receive approval to launch. In November 2020, the CFTC granted Kalshi designation as a Designated Contract Market, making it the first exchange in the United States to receive federal regulatory authorization to list event contract derivatives. This license forms the legal foundation of Kalshi, as federal law takes precedence over state gambling laws. Kalshi officially launched in July 2021.

However, obtaining a license has also brought a series of challenges. The platform’s process for launching new popular events is slow, and stringent KYC requirements have further hindered growth. The most significant blow came when the CFTC rejected Kalshi’s application to launch markets for the 2024 U.S. presidential election—leaving Kalshi unable to participate in the largest catalyst in industry history.

Kalshi filed a lawsuit and won. In September 2024, a federal judge ruled that the CFTC had overstepped its authority, determining that election-related contracts were neither illegal nor gambling. With only 32 days remaining until the general election, Kalshi reopened election trading. This delay, combined with strict KYC requirements that excluded international users, caused Kalshi to lose ground in user perception and market share to Polymarket. Kalshi’s total election-related trading volume amounted to approximately $500 million, while Polymarket reached $3.6 billion.

After the election, investments in compliance began to pay off. Robinhood partnered with Kalshi to launch its first prediction market product, and Bloomberg Terminal integrated Kalshi’s data directly—both partnerships built on Kalshi’s regulatory credentials.

In 2025–2026, Kalshi significantly expanded beyond political markets, with sports becoming its largest category, driven by breakout marketing campaigns. The slogan “Knicks sweep in four games,” featured in street interviews during the NBA Finals, garnered tens of millions of views. The team leveraged the World Cup with targeted advertising featuring Timothée Chalamet, Lionel Messi, and Luka Dončić. During the World Cup, 3 million users on the platform generated $27 billion in trading volume.

As of July 2026, Kalshi has executed a cumulative total of 9.826 billion trades with a total trading volume of $155.7 billion, surpassing Polymarket to become the new industry leader.

The success of two leading platforms has prompted numerous exchanges and brokerages to launch their own prediction markets. Many companies have adopted a rapid launch strategy, testing market demand by integrating existing platforms into their own applications.

Coinbase and Interactive Brokers took a different path by adopting an aggregation model, consolidating liquidity and markets from multiple platforms. CME Group built its own products from scratch. Robinhood pursued a systematic approach, initially routing orders to Kalshi; after validating demand, it developed its own underlying system and launched Rothera, a CFTC-regulated exchange, in June 2026. This bet quickly paid off: in the second quarter of 2026, Robinhood’s event contracts business generated $1.56 billion in revenue, more than ten times year-over-year growth, surpassing its cryptocurrency trading revenue of $1 billion.

Kalshi and Polymarket have addressed execution-level failures: regulatory strategy, low fees, and consumer-friendly product experiences. However, whether they have solved the deeper demand challenges posed by Whitaker and Mazlish is another question.

• Gambling audience: Partially addressed, but far from making everything predictable. Sports topics are naturally suited to prediction markets; the core issue has simply been whether to capture trading volume from traditional sports betting—and they’ve already done it. So far this year, most trading volume has come from sports. Parlay betting has been a key growth driver: it accounted for only 3% of total trading volume when launched on Kalshi in September 2025, rising to 38% by July 2026. Following closely behind are cryptocurrency price predictions. Political topics are no longer limited to elections—military and geopolitical conflict markets have generated $27.6 billion in trading volume, slightly exceeding the $27.3 billion from U.S. presidential elections; when including overseas elections, the broader political segment continues to grow. Cultural markets (music, film, celebrities) are also expanding steadily. Economic topics such as Federal Reserve interest rate decisions and inflation are seeing significant volume as well. Trading volume is rising across multiple categories. Although there is still a long way to go toward the grand vision of “every topic forming an active market,” substantial progress has been made over the past 12 months.
• Professional traders: Partially addressed through incentives. Professional traders require sufficient trading volume and counterparties beyond other professional traders. The $120 billion sports market and $22 billion crypto market are already large enough to meet this need. Susquehanna joined Kalshi as a market maker in 2024 and later formed a joint venture with Robinhood to launch a prediction market. Jump Trading has made equity investments in both platforms in exchange for providing liquidity; Citadel is also evaluating entry possibilities.
• Savers: The issue remains unresolved. Prediction markets are still zero-sum games, where funds put at risk forego potential returns from investing in government bonds or elsewhere.

In the next phase, prediction markets will evolve from niche platforms into infrastructure that prices global information, giving rise to entirely new mechanisms.

Customized hedging

Businesses can hedge against unique risks that traditional finance and insurance cannot cover. For example, an ice cream shop can hedge against the risk of cooler-than-expected summer temperatures. Such needs were previously unmet, as traditional insurers would not underwrite them due to the difficulty of achieving profitable coverage for niche exposures.

Breakthrough of the $0–$1 binary pricing model

Perpetual market: A continuous trading market for any event. For example, with inflation: a binary market can only bet on a fixed outcome, such as “Will inflation exceed 3.1%?” A perpetual market allows direct long or short positions on the inflation rate itself.

Combined markets and idea governance: Instead of pricing only single events, price the relationship between two events. For example: “What would Tesla’s stock price be if Elon Musk resigns?” or “How high would oil prices go if the U.S. launches an invasion?” By having separate markets price each variable, asset values can be assessed more accurately. Idea governance goes further, using conditional markets to inform governance and policy decisions: promoting policies that markets predict will yield better outcomes.

AI agents act as truth seekers

In the future, AI agents will operate extensively in the market, conducting research and executing trades 24/7. These agents can serve as automated truth detectors, scanning Telegram communities and social media to uncover factual evidence faster than human commentators, and executing trades based on mispricings to further enhance the accuracy of prediction markets.

The data infrastructure behind the media

Deep integration of news: The collaboration between CNN and CNBC signals an even closer symbiotic relationship with traditional media. Media coverage no longer focuses solely on events that have already occurred, but also on what is likely to happen next.

Long-tail topic expansion: Extend the market to long-tail areas by pricing community-local events and niche cultural trends, such as tech industry layoffs or Taylor Swift album performance. However, to stimulate demand, user habits must change—users should not only consume news but also participate in betting on outcomes.

Solving the problem of missing savers

Yield-bearing collateral: Traders no longer need to lock up idle USDC; they can use yield-generating assets such as sUSDe or tokenized U.S. Treasuries as margin. Over the past two years, Ethena’s sUSDe alone has generated annualized yields of 4% to 30% through perpetual funding rates and is already accepted as collateral on platforms like Aave, Pendle, and Morpho. If prediction markets support similar collateral, savers’ capital can continuously earn returns while serving as margin for positions, rather than remaining idle until event settlement.

Financial and DeFi modular combinations: Structured products can be built on top of prediction markets, either bundled with other assets or used as collateral for borrowing. When event contracts can be bundled with interest-bearing assets or used as collateral for loans, zero-sum gaming is no longer the sole return for this capital.