Prediction markets moved $63.5 billion in 2025, up from $15.8 billion the year before. That number is almost quaint compared to where the industry is headed — analysts project $200 billion to $325 billion in 2026 volume. But the mechanism behind that money, letting people bet real cash on real-world outcomes and calling it a market instead of a wager, is nearly 40 years old.
The evolution of prediction markets is a story about regulators chasing innovation, retail traders chasing edge, and a handful of academics who built something in a University of Iowa classroom that eventually became a tool ICE invested $2 billion into. Understanding that arc explains why Polymarket and Kalshi look the way they do today, why the CFTC keeps changing its mind, and why the next 12 months will look nothing like the last decade.
This guide walks through every major era: the academic origin story, the Intrade boom and collapse, the crypto reboot, and the regulatory breakthrough that turned prediction markets into a mainstream financial product. If you want the mechanics of how these markets actually price outcomes, see our companion piece on how prediction markets work.
The Iowa Electronic Markets and the Academic Origins (1988-2000)
The first modern prediction market launched in 1988 at the University of Iowa's business school. Economists built the Iowa Electronic Markets (IEM) to test a simple hypothesis: that a market made of real money traders would forecast presidential election outcomes more accurately than polls. It worked. Across multiple election cycles, IEM's vote-share contracts consistently beat national polling averages, and the CFTC granted it a no-action letter allowing small-stakes trading for academic research.
IEM never scaled beyond a few thousand dollars in contract value per election. Trades were capped at $500 per account, and the entire operation existed to generate research papers, not revenue. But it proved the core thesis that still drives every platform today: aggregated financial incentives produce better forecasts than expert opinion or survey data. That single finding is the intellectual foundation everything else in this article builds on.
Intrade and the First Retail Boom (2001-2013)
Intrade, founded in Ireland in 2001, took the Iowa concept and removed the guardrails. No $500 cap, no academic oversight, and contracts on everything from elections to Saddam Hussein's capture to Oscar winners. By the 2008 and 2012 US presidential elections, Intrade was moving tens of millions of dollars and getting cited on CNN and in the Wall Street Journal as a real-time probability tracker.
Intrade's downfall previewed every regulatory fight the industry still has today. The CFTC sued Intrade in 2012 for offering off-exchange commodity options to US residents without registration. Intrade blocked US customers in 2013, then collapsed entirely in early 2014 after auditors discovered a shortfall in customer funds — a solvency scandal, not a regulatory one. The lesson that survived: unregulated offshore prediction markets can build real liquidity, but they carry real counterparty risk. That tension between innovation speed and regulatory legitimacy still defines the legal status of prediction markets in the US in 2026.
The Crypto Reboot: Augur, PredictIt, and the DeFi Experiment (2014-2019)
Intrade's collapse left a vacuum, and two very different projects filled it. PredictIt launched in 2014 as a nonprofit-adjacent academic platform under a CFTC no-action letter, capping trades at $850 per contract per market and restricting it to political and current-events contracts. It became the go-to political forecasting tool for journalists and academics for the next decade, until the CFTC moved to revoke its no-action letter in 2022 — a fight that dragged into 2023 before PredictIt survived with a smaller scope.
Augur, launched in 2018 on Ethereum, took the opposite approach: fully decentralized, no company to sue, no KYC, resolution handled by token-holder consensus instead of a corporate oracle. It proved that prediction markets could run entirely on-chain, but Augur never solved the liquidity problem. Thin order books and slow, disputable resolution made it a proof of concept rather than a product traders actually used at scale. It matters historically because Polymarket's entire architecture — on-chain settlement, off-chain order matching, a separate oracle for resolution — is a direct response to what Augur got wrong.
Polymarket, Kalshi, and the Regulatory Breakthrough (2020-2024)
Polymarket launched in 2020 on Polygon, settling trades in USDC and using an optimistic oracle (UMA) for dispute resolution instead of pure token voting. It grew slowly through 2020 and 2021, then exploded during the 2024 US presidential election, when its
