News

What is a prediction market? A plain-English definition for 2026

Prediction markets are peer-to-peer exchanges where users buy and sell binary contracts on real-world outcomes. Here's how they actually work — and why investors are paying attention.

← News

Note, November 15, 2026: this analysis was written in May 2026. Figures in it are estimates that we have not traced to a primary source, and some details have changed since. For current rules and status, see our regulation news and state table.

TL;DR:

A prediction market is an exchange where users trade contracts that pay $1 if a specified event happens and $0 if it does not. The market price between $0 and $1 is interpreted as the crowd's implied probability. Polymarket and Kalshi crossed $13bn in monthly trading volume in December 2025, up from under $100m in early 2024 — but the legal status of the category is still being decided in US federal court.

Key takeaways
  • Prediction markets are exchanges, not sportsbooks. Users trade against each other, not against a house.
  • Each contract resolves to $1 (yes) or $0 (no). Today's price is the market's probability estimate.
  • Monthly volume scaled from <$100m to >$13bn in under two years (Source: NEXT.io, March 2026).
  • In the US, the two dominant venues are Polymarket (offshore, crypto-settled) and Kalshi (CFTC-regulated, USD).
  • Whether prediction markets are 'gambling,' 'derivatives,' or a new category is the central legal question of 2026.

How a prediction-market contract works

A prediction-market contract is a binary financial instrument. It pays the holder $1.00 if a specified event occurs and $0.00 if it does not.

Take a simple example: a market on "Will the Federal Reserve cut rates at its September 2026 meeting?" If you believe the probability is 70%, and the market is trading "YES" at $0.55, you buy YES. If the Fed cuts, your contract settles at $1.00 — you made $0.45 per contract. If the Fed holds, your contract is worth nothing.

Because the two sides (YES and NO) must always sum to $1.00, the price of YES is a direct, real-time readout of the crowd's probability estimate. That property — turning belief into a number you can stake money on — is what makes the asset class distinctive.

Exchange, not sportsbook

The economic structure matters more than the user experience.

A sportsbook is a dealer market. The operator (DraftKings, FanDuel, bet365) sets the odds, takes the opposite side of every wager, and earns its margin from the spread between true probability and posted odds. The house is a counterparty.

A prediction market is an exchange. The operator runs an order book and matches buyers with sellers. Liquidity comes from other users — market makers, hedgers, syndicates, and retail traders all in the same book. The platform earns fees on volume, not on outcomes. It has no rooting interest in whether you win or lose.

That distinction has three consequences:

  1. No house edge in the traditional sense. Pricing is determined by supply and demand, not by an operator's vig.
  2. Limits scale differently. A sportsbook's maximum bet is constrained by its risk appetite. An exchange's depth is constrained only by who else is willing to take the other side.
  3. Regulators treat them differently. Exchanges that list event contracts are typically regulated under derivatives law (CFTC in the US), not gambling law (state regulators). See who regulates prediction markets.

The two platforms that matter

Two venues account for the overwhelming majority of US-facing prediction-market volume.

Polymarket runs on the Polygon blockchain, settles in USDC, and does not accept US-based users on its global platform. Since 2025, US residents use Polymarket US, a separate CFTC-regulated exchange. It dominates politics, geopolitics, and culture markets. Liquidity peaked around the 2024 US presidential election and has stayed structurally elevated since.

Kalshi is a CFTC-registered Designated Contract Market. It settles in USD via traditional rails, accepts US users by default, and has aggressively expanded into sports event contracts — a move that triggered cease-and-desist letters from gambling regulators in more than a dozen states. The litigation is unresolved as of mid-2026.

Smaller venues — PredictIt, Manifold, Insight Prediction, and a wave of crypto-native platforms — exist, but liquidity sits at Polymarket and Kalshi.

Why volume is growing

According to NEXT.io's March 2026 report The Great Divide, monthly prediction-market trading volume rose from under $100m in early 2024 to more than $13bn in December 2025. Four forces explain the curve:

  1. The 2024 US election. Polymarket's election book became a daily news story; financial media (Bloomberg, FT) cited contract prices as a probability gauge during the campaign.
  2. Kalshi's sports launch. Event contracts on NFL and college football games gave the platform a recurring, high-frequency liquidity engine.
  3. Regulatory ambiguity has stayed unresolved. Operators have built and shipped product faster than courts can constrain them.
  4. Crypto rails. Settlement on Polymarket is on-chain, near-instantaneous, and global. That removes friction the legacy gambling stack still imposes.

Is it gambling?

This is the question every investor, regulator, and operator is asking, and there is no settled answer. The honest framing:

  • Functionally, putting money on whether the Chiefs cover a spread looks identical whether the contract lives at Kalshi or at FanDuel.
  • Legally, the CFTC's position is that listed event contracts are derivatives — not gambling — and therefore preempt state gambling law where they conflict.
  • Politically, state regulators and tribal gaming interests dispute that interpretation and have filed suit.

The Supreme Court has not ruled. Until it does, every participant is operating under regulatory uncertainty. For more on the legal landscape, see the CFTC, the states, and prediction markets.

What this means for affiliates and publishers

For anyone working in the affiliate and media layer, three implications matter:

  • Demand is real and growing. The category is past the speculative phase. $13bn/month is not a curiosity.
  • Distribution is unsettled. Sportsbook affiliates own SEO for "sports betting." Nobody owns "prediction markets" yet.
  • Regulatory risk is concentrated at the venue, not the affiliate. A media property that educates users and routes to multiple licensed venues has a more durable risk profile than the venues themselves.

We unpack the affiliate-economics in detail in unit economics of a prediction-market affiliate site.

The European lens

Europe has had peer-to-peer betting exchanges since Betfair launched in 2000 and was regulated under UK gambling law from 2005. The structural lesson: exchanges and sportsbooks coexisted, but exchanges captured a specific user — the price-sensitive, sophisticated bettor — rather than displacing the casual one. US prediction markets are likely to follow the same pattern, with one twist: classification as a CFTC-regulated derivative (rather than as gambling) gives them a wider distribution surface than Betfair ever had.

#category

#explainer

Related