A prediction market is an exchange where you trade contracts that pay out based on whether a real-world event happens. Each contract settles at $1 if you're right and $0 if you're wrong. Prices between $0.01 and $0.99 reflect the market's collective probability estimate. A contract priced at $0.65 means the crowd thinks there's a 65% chance the event occurs.
In 2026, prediction markets moved from fringe experiment to serious financial infrastructure. Kalshi hit $22 billion valuation. Polymarket is raising at $20 billion. Combined monthly volume crossed $44.8 billion in June. The Federal Reserve now cites Kalshi in its own research on market expectations.
This is the complete guide to what prediction markets are, how they work, and how to use them — updated August 2026 with the latest platform data, regulatory status, and research findings.
Table of Contents
- What Is a Prediction Market? (Short Answer)
- Brief History: From Hayek to Polymarket
- How Prediction Markets Actually Work
- How Accurate Are Prediction Markets?
- Prediction Markets vs Sports Betting
- Prediction Markets vs Polls
- The Major Platforms in 2026
- Risks and Limitations
- How to Get Started
- Frequently Asked Questions
What Is a Prediction Market? (Short Answer)
A prediction market is a financial exchange where participants trade binary contracts tied to the outcome of a future event. If you buy a Yes contract at $0.40 and the event happens, you get $1. If it doesn't, you get $0.
Prices reflect the crowd's collective probability estimate. When new information hits, traders react and prices update in real time.
Think of them as a stock exchange where each "stock" is a question with a binary answer. "Will the Fed cut rates in September?" trades like a security. Its price is the market's implied probability.
Brief History: From Hayek to Polymarket
The economic theory behind prediction markets traces to Friedrich Hayek's 1945 essay "The Use of Knowledge in Society," which argued that prices are the most efficient way to aggregate dispersed information.
Political betting markets existed on Wall Street from 1884 onward. Records show average turnover per US presidential election reaching 50% of campaign spending in the pre-scientific-polling era.
Modern academic prediction markets began with the Iowa Electronic Markets in 1988. Studies later found Iowa beat 964 opinion polls at predicting election outcomes 74% of the time.
The commercial era started with Intrade (2004), collapsed after regulatory pressure, and rebooted with Kalshi (2018) and Polymarket (2020). Kalshi became the first CFTC-approved prediction market in the US in 2020.
The tipping point came in 2024. Polymarket priced Trump's US election victory correctly before major polls did, moving $3.6 billion in volume during the race. Media outlets began citing prediction market prices as leading indicators. By June 2026, combined monthly volume across Kalshi and Polymarket hit $44.8 billion.
How Prediction Markets Actually Work
Step 1: A market is created
Someone (usually the platform) creates a market with a clearly defined resolution question. Every market must have:
- A binary outcome (Yes or No)
- A resolution source (official data, government report, referee decision)
- A resolution date
Example: "Will US CPI year-over-year be above 3.2% for August 2026?" — resolves based on BLS data released September 11, 2026.
Step 2: Contracts start trading
The market opens. Traders buy Yes or No contracts at whatever price the order book shows. Prices move as more traders enter and information changes.
Contract prices range from $0.01 to $0.99. The lower the price, the less likely the market thinks the event is. Prices always sum to $1 across Yes and No in a two-sided market with reasonable liquidity.
Step 3: New information moves prices
When news hits, prices react. If Fed governors give hawkish speeches, "Fed will hike in September" contracts move up. If inflation data comes in soft, they move down.
Research by Fed staff in 2026 found Kalshi's day-before FOMC forecasts have outperformed Fed funds futures at predicting the actual decision. This is why hedge fund macro desks now watch both.
Step 4: The event resolves
When the resolution date arrives and the outcome is known, the market closes. Winning contracts settle at $1. Losing contracts settle at $0. Payouts go to the winning side automatically.
On Kalshi, resolution comes from official sources (Fed, BLS, sports leagues). On Polymarket, resolution comes from UMA oracles, where token-holders vote to confirm outcomes.
How Accurate Are Prediction Markets?
Prediction markets typically achieve Brier scores of 0.15 to 0.25 across resolved contracts. Lower is better; a perfect prediction scores 0. Expert forecasters typically score 0.20 to 0.35.
Specific accuracy data:
- Iowa Electronic Markets beat 964 opinion polls 74% of the time from 1988 to 2004
- Polymarket claims 94% accuracy at one month out
- Kalshi day-before FOMC predictions beat Fed funds futures over the last 12 meetings
- Combined Kalshi + Polymarket Q1 2026 accuracy: Brier 0.12 in economics, 0.14 in politics
The most important 2026 finding came from a Yale + London Business School paper analyzing $13.76 billion of Polymarket trades: 3% of skilled traders drive nearly all price discovery. The other 97% funds their profits. Accuracy comes from a persistent minority, not the wisdom of the crowd.
Read our deeper analysis: The 3% Rule: Why "Wisdom of Crowds" Was Always a Myth
Prediction Markets vs Sports Betting
Prediction markets and sportsbooks look similar. Both let you stake money on outcomes. The mechanics are different.
| Feature | Prediction Market | Sportsbook |
|---|---|---|
| Odds set by | Traders (peer-to-peer) | Bookmaker (against house) |
| Regulator | CFTC (federal) | State gaming commissions |
| Instrument type | Event contract (swap) | Wager |
| Exit before event | Yes, sell anytime | Rarely, mostly locked |
| Fees | Small, per-contract | Built into spread |
| Covers non-sports events | Yes — anything | Limited to sports |
| Tax treatment | Capital gains (1099) | Gambling income (W-2G) |
The Pew Research Center analysis in April 2026 showed prediction market monthly volume ($24B) surpassed US sportsbook volume ($14B). The categories are diverging fast.
Prediction Markets vs Polls
Polls ask people what they think. Prediction markets ask people what they'll bet on. Those are different signals.
Advantages of markets:
- Update in real time (polls take days)
- Aggregate global information (polls sample locally)
- Reward accuracy with money (polls have no accuracy incentive)
- Incorporate expert and insider views (polls exclude structural knowledge)
Advantages of polls:
- Better at capturing demographic detail
- Not subject to manipulation attempts (though rare in liquid markets)
- Can measure intensity of preference, not just direction
The best forecasters use both. Nate Silver's 538 model combines polls with prediction market prices. Bloomberg tracks Kalshi alongside Fed funds futures for macro forecasts.
The Major Platforms in 2026
Kalshi
CFTC-regulated US exchange. USD funding. Best for: Fed decisions, US politics, mainstream sports, weather. $31.5B June 2026 volume. $22B valuation.
Polymarket
Blockchain-based, now with US regulated version. USDC funding. Best for: global politics, geopolitics, crypto, niche markets. $10.8B June 2026 volume. $20B valuation.
Manifold Markets
Play-money prediction market. Free to use. Best for: learning mechanics, calibration training, community forecasting. Not for profit.
Metaculus
Long-horizon research-focused platform. Free. Best for: multi-year forecasts, science and technology questions.
PredictIt
Academic prediction market run by Victoria University. Limited to politics. Small position caps ($850).
Compare all of these in depth: Kalshi vs Polymarket: Complete 2026 Comparison.
Risks and Limitations
Total loss
Every contract can go to $0. If you paid $0.65 for a Yes contract and the event doesn't happen, you lose the entire $0.65. Position sizing matters more than being right.
Liquidity risk
Niche markets can have thin order books. You can enter but not exit at fair price. Rule of thumb: don't trade markets with less than $500,000 in cumulative volume.
Resolution disputes
Ambiguously worded contracts can lead to disputes at settlement. Kalshi froze $54 million on a "Khamenei out as Supreme Leader" contract in 2026 over grammatical ambiguity. Read every market's resolution criteria before trading.
Regulatory risk
State-level bans can restrict access with little notice. Minnesota tried to ban all prediction markets in May 2026. A federal judge blocked it in July. This regulatory uncertainty is still active.
Insider trading
Prediction markets have been exploited by insiders — MrBeast video editor (Feb 2026), congressional candidates trading on their own races, teleprompter operator for Trump's State of the Union. Both platforms have expanded surveillance, but the risk exists.
How to Get Started
- Pick a platform. Kalshi if you want the simple regulated route. Polymarket if you want the broader market coverage. Manifold if you want to practice with play money first.
- Verify your account. KYC/AML checks are required by all regulated platforms.
- Start with $50–$100. Individual contracts can cost as little as $0.01. Small stakes let you learn without risking capital.
- Trade what you understand. Fed decisions if you follow macro. Sports if you follow sports. Elections if you follow politics.
- Track your predictions. Keep a spreadsheet. After 50 trades, calculate your calibration score. This separates traders from gamblers.
- Read the rules. Every contract has a resolution page. Read it before you buy. Ambiguity kills.
Frequently Asked Questions
Are prediction markets legal in the US?
Yes, at the federal level. CFTC-regulated platforms (Kalshi, Polymarket US) are legal nationwide. State-level restrictions apply in some states. As of August 2026, prediction markets are technically legal in all 50 states, but Kalshi is unavailable in Washington and Polymarket US is unavailable in AZ, IL, MA, MD, MI, MT, NV, and OH.
Are prediction markets gambling?
Legally in the US, they are event contracts regulated as derivatives under the CFTC — same category as futures contracts, not gambling. The distinction matters for taxes and regulation. Culturally, many people still describe them as gambling.
How much money do I need to start?
Nothing to browse. Around $10 to trade. Contracts cost between $0.01 and $0.99. Kalshi and Polymarket have no deposit minimums.
Do prediction markets pay real money?
Yes on Kalshi, Polymarket, and PredictIt. No on Manifold and Metaculus, which use play money.
Can prediction markets be manipulated?
Short-term price manipulation is possible but usually unprofitable. Research by Hanson et al. shows manipulated prices revert within hours as arbitrageurs correct the mispricing. The larger and more liquid the market, the harder manipulation becomes.
What's the difference between Yes and No contracts?
Buying Yes means you think the event will happen. Buying No means you think it won't. In a two-sided market, Yes price + No price = $1. Buying No at $0.35 is equivalent to selling Yes at $0.65.
How do prediction markets make money?
Small per-contract fees on trades. Kalshi charges $0.07 to $1.75 per 100 contracts. Polymarket charges 0% to $1.75 per 100 contracts and waives fees on geopolitical markets. Neither takes positions against users.
Why are Kalshi and Polymarket priced differently for the same event?
Different trader pools, different funding methods, different fee structures. When they diverge significantly, that gap is usually where professional arbitrage happens. The gap tends to close as resolution approaches.
Can I make a living trading prediction markets?
The Yale/LBS study found roughly 3% of traders are consistently profitable. Two out of three users lose money on average. It's possible but requires the same discipline as professional trading in any market — a repeatable process, risk management, and calibrated forecasting.
What Juno Lets You Do
Juno Predictions is building event contract infrastructure for markets that Kalshi and Polymarket don't currently serve — starting with Thailand. Same probability pricing, same binary contracts, same real-time resolution. Built for a jurisdiction that Kalshi requires US residency and Polymarket got blocked in.
The prediction market category is going global. The infrastructure to serve non-US traders legally is only starting to exist.
The Bottom Line
Prediction markets are the most accurate forecasting tool humans have built at scale. They work because a small group of skilled traders keeps prices honest, funded by a larger group of less-informed participants.
You can use them without trading them. The prices themselves are information — read them the way you'd read a weather forecast. When Kalshi shows 59% for a September Fed hike, that number is one of the sharpest signals available anywhere.
If you do trade, treat it like a skill you're building. Small positions, tracked predictions, patient calibration. That's the difference between the 3% who win and the 97% who fund them.
Article last updated: August 31, 2026. Data current as of Q3 2026.