From Tavern Bets to Billion-Dollar Platforms
Prediction markets are often described as a modern innovation, but the idea of betting on future events to aggregate information is centuries old. The history of prediction markets reveals a consistent theme: whenever people have been allowed to wager on outcomes, the resulting prices have been remarkably informative.
This is the story of how informal election betting evolved into the sophisticated, blockchain-powered platforms we use today.
The Early Days: Election Betting in America
Wall Street Betting Pools (1868-1940)
The earliest well-documented prediction markets were the Wall Street betting pools on US presidential elections. Starting in 1868, organized betting on elections took place at the Curb Exchange (now the American Stock Exchange) and in major New York hotels and saloons.
These were not small operations. During the 1916 election between Woodrow Wilson and Charles Evans Hughes, election betting volumes were estimated at $165 million in today's dollars — enormous for the era. Newspapers regularly reported betting odds alongside their election coverage, and these odds were considered the most reliable indicator of election outcomes.
The accuracy was impressive. Between 1868 and 1940, Wall Street betting markets correctly predicted the winner of every presidential election except 1916 (the closest race of the era). They also provided surprisingly precise estimates of victory margins.
Why They Disappeared
Election betting markets declined in the 1940s due to several factors:
- The rise of scientific polling — George Gallup's polling methods, introduced in the 1930s, provided a seemingly more rigorous alternative
- Anti-gambling sentiment — Growing moral opposition to gambling led to tighter regulations
- World War II disruption — The war disrupted normal commercial and social activities
For nearly fifty years, prediction markets essentially disappeared from public life.
The Academic Revival: Iowa Electronic Markets
IEM Launch (1988)
The modern history of prediction markets begins in 1988 at the University of Iowa. Professors Robert Forsythe, Forrest Nelson, George Neumann, and Jack Wright created the Iowa Electronic Markets (IEM) — an academic project that allowed participants to trade real-money contracts on presidential election outcomes.
The IEM operated under a no-action letter from the CFTC, which allowed it to function as a research tool with limits on individual account sizes ($500). Despite its small scale, the IEM produced groundbreaking results.
The IEM's Impact
Research using IEM data demonstrated that:
- Market prices predicted election outcomes more accurately than polls in the majority of comparisons
- Markets incorporated new information faster than any other forecasting method
- Even small markets with limited participants could produce accurate probability estimates
These findings laid the intellectual groundwork for commercial prediction markets. The academic literature on prediction market accuracy — much of it based on IEM data — remains influential today.
The First Wave: Online Prediction Markets (2000s)
Intrade and TradeSports
The first major commercial prediction markets emerged in the early 2000s. Intrade, an Ireland-based platform, and its sister site TradeSports allowed real-money trading on political, financial, and sporting events.
Intrade gained significant media attention during the 2004 and 2008 US presidential elections. Its odds were widely cited by journalists, analysts, and even the candidates themselves. At its peak, Intrade had hundreds of thousands of registered users and millions in trading volume.
However, Intrade faced mounting regulatory pressure. The CFTC filed a civil complaint against the platform in 2012 for offering commodity option contracts to US customers without regulatory approval. In March 2013, Intrade suspended trading and eventually shut down, citing "irregular activity" in its accounts.
The Pentagon's Prediction Market Controversy
In 2003, the Defense Advanced Research Projects Agency (DARPA) proposed a prediction market called the Policy Analysis Market that would have allowed trading on geopolitical events, including the likelihood of terrorist attacks and political assassinations.
The proposal created an immediate political firestorm. Senators called it "morally repugnant" and a "terrorism futures market." The program was canceled within 24 hours of its public announcement, and the deputy director responsible for it resigned.
The controversy set back prediction market development by years. Despite academic arguments that such markets could provide valuable intelligence signals, the political toxicity of the concept prevented government-sponsored prediction markets from materializing.
The Regulatory Path: Kalshi and the CFTC
Kalshi's Approval (2020)
A major turning point came in 2020 when the CFTC approved Kalshi as a designated contract market (DCM). This made Kalshi the first fully regulated prediction market in the United States — a development that legitimized the entire industry.
Kalshi's approval required the platform to:
- Hold customer funds in segregated accounts
- Implement robust KYC/AML procedures
- Submit new markets for regulatory review
- Maintain transparent trading records
The regulatory framework was not without friction. Kalshi fought protracted battles with the CFTC over specific market types, particularly election markets. But the overall trajectory was toward greater acceptance and clearer rules.
The Election Markets Battle
For years, the CFTC was reluctant to approve contracts on US election outcomes, viewing them as too close to gambling. This position softened significantly following the 2024 election, when prediction markets' superior accuracy compared to polls became undeniable.
By 2025, Kalshi was offering a full suite of political markets, and regulatory conversations had shifted from "should prediction markets exist?" to "how should they be regulated?"
The Blockchain Revolution: Polymarket
Polymarket's Rise (2020-Present)
Polymarket launched in 2020 on the Ethereum blockchain (later migrating to Polygon for lower transaction costs). It represented a fundamentally different approach to prediction markets: decentralized, global, and crypto-native.
Polymarket's early years were modest, but the platform gained explosive traction during the 2024 US presidential election. Several factors drove this growth:
- Zero trading fees attracted active traders
- No geographic restrictions (except the US) enabled global participation
- Blockchain transparency allowed anyone to verify trading data
- Viral attention as prediction market odds became mainstream news
During the 2024 election, Polymarket processed over $3.5 billion in trading volume, making it the largest prediction market in history by a wide margin. The platform's odds were cited by major media outlets worldwide and proved more accurate than polling averages.
Post-2024 Growth
The 2024 election was a watershed moment. It demonstrated that prediction markets could operate at scale, attract mainstream attention, and produce genuinely useful forecasts. In the aftermath, Polymarket expanded its market offerings, improved its user experience, and attracted institutional market makers.
By 2026, Polymarket processes billions in monthly volume across hundreds of markets. It has become the default reference point for real-time probability estimates on major world events.
The Present and Future
Where We Are Today
In 2026, prediction markets are more legitimate, more liquid, and more widely used than at any point in history. The industry has evolved from academic experiments and regulatory gray areas into a recognized component of the information ecosystem.
Key trends shaping the current landscape:
- Mainstream media integration — Major news outlets routinely report prediction market odds
- Institutional participation — Hedge funds and financial firms actively trade on prediction markets
- Regulatory frameworks — The US, EU, and other jurisdictions are developing specific prediction market regulations
- Platform maturity — Better interfaces, mobile apps, and fiat onramps are reducing barriers to entry
What Comes Next
The history of prediction markets suggests a consistent pattern: periods of growth, regulatory challenges, and eventual acceptance. Each cycle brings more sophisticated platforms, broader participation, and greater accuracy.
The next frontier likely includes:
- Conditional markets — "If X happens, what is the probability of Y?" enabling complex scenario analysis
- Government adoption — Using prediction markets internally for policy forecasting
- AI integration — Algorithmic traders and AI-powered market makers improving liquidity and accuracy
- Global expansion — New platforms serving underserved markets in Asia, Africa, and Latin America
Explore today's prediction markets on PredMarket.io, and see our guide to the best platforms in 2026 to find the right one for you.
