Lesson 01·Foundations·8 min read

What Is a Prediction Market? A Beginner's Guide

A prediction market is an exchange where contracts pay out based on real-world events — elections, sports, economic data, weather.

A prediction market is an exchange where the things being traded are not stocks or commodities but contracts that pay out based on the outcome of a future event. If you buy a contract for "Candidate X wins the 2028 election" at $0.40 and Candidate X wins, the contract pays $1.00. If they lose, it pays $0.00. The price of the contract — a number between $0 and $1 — can be read directly as the market's implied probability of the event happening.

This makes prediction markets unique. A poll asks a thousand people what they think; a prediction market asks anyone with money on the line what they're willing to bet. Because traders lose money for being wrong, prediction-market prices tend to incorporate information faster and more accurately than polls or pundit forecasts.

How it works, step by step

A prediction market lists an event with a clear, verifiable resolution — for example, "Will the Federal Reserve cut rates at its next meeting?" Two outcome tokens are created: YES and NO. Each token pays $1 if its outcome occurs and $0 otherwise. Traders post buy and sell orders in an order book (Kalshi, Polymarket) or against a liquidity pool (Limitless, some Manifold formats).

You deposit funds, buy the side you believe in at the current price, and either hold until the market resolves or sell earlier if the price moves your way. When the event settles, the platform (or an oracle) reads the outcome from a pre-specified source and pays out winners.

A short history

Informal prediction markets have existed for centuries — bookmakers took bets on 19th-century US presidential elections long before Gallup ran his first poll. The Iowa Electronic Markets, launched by the University of Iowa in 1988, is the oldest active academic prediction market and consistently forecast elections better than contemporary polls.

After the rise of crypto, real-money prediction markets exploded in scale. Polymarket, launched in 2020, traded billions of dollars during the 2024 US presidential election. Kalshi, regulated by the CFTC, brought event contracts to the US derivatives market in 2021 and won a landmark 2024 court ruling that let it list election contracts nationwide.

Why they're often more accurate than polls

Markets aggregate information. Anyone who knows something — a tracking pollster, a campaign operative, a sports trainer, a weather modeler — has an incentive to act on it. Their trades move the price. Decades of academic research (Wolfers and Zitzewitz, Hanson, Arrow et al.) have found that prediction markets generally outperform expert forecasts, especially over longer horizons.

They aren't infallible. Markets can be thin, manipulated by concentrated capital, or biased by partisan retail enthusiasm. Brexit and Trump 2016 are the canonical examples of markets that priced a correct-in-hindsight outcome as unlikely. But over many markets and enough time, they remain a remarkably reliable probability machine.

Worked example: reading a market

Suppose Polymarket lists "Will inflation come in below 3% for the next print?" and the YES contract trades at $0.68. That price says the market collectively assigns roughly a 68% probability. If you think the true probability is 80%, you buy YES at $0.68 — your expected value per contract is $0.80 minus the $0.68 you paid, or $0.12.

If a new inflation reading comes in soft, other traders update and the price might rise to $0.85. You can hold to resolution for the remaining $0.15, or sell now and lock in a $0.17 profit. Either way, the mechanic is: buy low, sell high, or hold to $1 if you're right.

Where to trade

Polymarket is the largest crypto-based prediction market — you deposit USDC on Polygon and trade against an on-chain order book. Kalshi is the leading US-regulated platform, where you deposit USD via ACH and trade CFTC-approved event contracts. Manifold uses play-money (Mana) and is a great sandbox. PredictIt is a legacy academic political market with strict position limits.

Each platform has tradeoffs in liquidity, regulation, fees, and contract variety. Status pages — like this one — let you check whether the platform is online before you try to trade. That matters most during peak events: election night, Fed decisions, and major sports finals are exactly when platforms strain.

Common pitfalls for beginners

Do not confuse price with certainty. A $0.90 contract still fails 10% of the time. Do not chase low-liquidity markets — the spread will eat your edge. Do not treat prediction markets as gambling; the ones with a real edge treat them as forecasting instruments with a side of profit.

Watch out for resolution ambiguity. "Will Elon Musk tweet X by Friday?" is unambiguous. "Will there be a recession in 2026?" depends entirely on who defines "recession" — check the resolution source before you trade.

When to use a prediction market

Use markets when you want a live, numeric probability that updates faster than any news cycle. Journalists cite them for elections; researchers use them as ground-truth against forecasts; traders use them to hedge exposure to real-world events. If you just want an opinion, read a pundit. If you want a number, read a market.

Frequently asked

Are prediction markets the same as betting?
In regulated venues like Kalshi they're legally treated as event contracts — derivatives, not gambling. Economically the mechanics are similar to binary options: you pay a premium for a probability-weighted payout. The regulatory framing matters for taxes, consumer protections, and where the platform can operate.
Can I make money on prediction markets?
Yes, but most participants lose money to fees, spread, and the smartest traders. Serious traders treat markets as information aggregators first and profit-seeking instruments second. Consistent winners specialize — one category, one methodology, one edge — rather than trading everything.
How do prediction markets resolve?
Regulated platforms like Kalshi define a specific resolution source (a government release, a wire service) in the contract itself, and pay out automatically. Crypto platforms like Polymarket use oracles such as UMA, where a proposed outcome can be disputed and voted on by token holders before payout.
What's the minimum I need to start?
Kalshi has no minimum deposit; you can trade from $1. Polymarket requires a small amount of USDC and MATIC on Polygon — practically, $20–$50 is a reasonable floor to cover fees. Manifold is free because it uses play-money.
Are prediction markets legal where I live?
In the US, Kalshi is legal nationwide, Polymarket is geoblocked, and PredictIt operates under a limited academic exemption. In the EU and UK, most platforms are accessible but face varying regulatory scrutiny. Always check your local law before depositing.
How do fees work?
Kalshi charges a variable fee on profitable closes (usually 1–7%). Polymarket has 0% trading fees but you pay the bid/ask spread and Polygon gas. Manifold has no fees because it's not real money. Robinhood charges a small per-contract fee on top of the underlying exchange's costs.
Can prediction markets be manipulated?
Thin markets can be moved by a single well-funded trader, but manipulation is expensive and usually leaves an obvious arbitrage for everyone else. Deep markets like Polymarket's flagship election contracts require millions of dollars to move by even one cent.
What's the difference between a prediction market and a sportsbook?
Sportsbooks set odds and take the other side of every bet, profiting from the vig. Prediction markets match users against each other and take only a fee. Prices float continuously with supply and demand rather than being set by a house model.

The .health network

Independent status monitors for the tools and markets we track.