A prediction market turns a question about the future into a contract you can buy and sell. This guide walks through the whole machine: what an event contract is, how the price becomes a probability, why an exchange has no house, how an order book matches buyers and sellers, how a market settles, the kinds of venue, the regulation, and the risks. Information, not advice.
The direct answer. A prediction market is a place to trade contracts on whether a defined future event happens. Each contract pays a fixed amount, usually 1 dollar, if the event occurs and nothing if it does not, so its price between 1 cent and 99 cents reads as the crowd estimate of the chance.
The key dated figure. These are real, sizeable markets. The CFTC reported that volume across its designated contract markets exceeded 25 billion dollars in 2025, and it issued new proposed rules for event contracts in 2026, so the framework is active and changing.
The one honest thing to know. An exchange has no house, but that does not make it safe. You can still lose, the price can be wrong, fees and thin books take a slice, and legality differs by region and moves quickly.
Last reviewed 26 June 2026 · Refreshed when platform mechanics or the regulatory position change.
This page is general information, not financial, investment, legal, tax, or betting advice. Event contracts carry a real risk of loss. A price is an implied probability, not a prediction of the result, and legal status varies by region and changes often.
A prediction market lets people buy and sell contracts tied to a defined future event, where a yes contract settles at 1 dollar if the event happens and at zero if it does not. Because the payout is fixed at those two ends, the live price sits between 1 cent and 99 cents and reads directly as the market estimate of probability, so a contract at 60 cents means the crowd puts the chance near 60 percent (per the way Kalshi and similar venues describe their contracts, as of June 2026). On an exchange there is no house taking the other side, you trade with other participants through an order book, and the venue earns from fees rather than from your loss.
Start with the smallest piece, because everything else is built from it. An event contract is a simple promise. It states a question with a clear yes or no answer, such as whether a named figure is released above a threshold by a certain date, and it pays a fixed amount if the answer turns out yes. On the regulated venues that amount is almost always 1 dollar, and the matching loss is zero. That is the entire instrument. There is no coupon, no maturity value to model, no counterparty negotiating terms. It is worth a dollar or it is worth nothing, and the only open question is which.
Because the two ends are fixed, the price has nowhere to go but between them. While the question is live, a yes contract trades somewhere from 1 cent to 99 cents. It cannot trade above a dollar, because no one rationally pays more than the most it can ever return, and it does not sit at zero or a dollar until the event is actually decided. Kalshi, a federally regulated exchange, describes this directly: a yes price of p cents corresponds to a market estimate of about p percent that the event happens, with the no side carrying the rest, per Kalshi educational material as of June 2026. The price is the probability wearing a dollar sign.
The reason this matters is that the number is not handed down by a company deciding what looks fair. It is produced by people buying and selling, each acting on their own information and willing to be wrong with their own money. When new information arrives, someone trades on it, and the price moves. The output is a single, continuously updating figure that represents the crowd best guess at the odds. That is the quiet genius of the design, and it is why a prediction market can be a useful instrument for reading uncertainty rather than only a place to take a position.
One more detail completes the picture. Every yes contract has a matching no contract, and the two are simply the same probability described from opposite directions. If yes trades at 60 cents, no trades at about 40 cents, because the holder of no collects the dollar in exactly the cases where yes does not. Add them together and you get close to a dollar, with any small shortfall or overshoot explained by fees and the gap between the best buy and sell prices. That tidy arithmetic is a quick way to sanity check a market: if yes and no are nowhere near a dollar apart, the book is thin and the number is soft. We unpack that reading in full in reading prices as implied probability.
This is the part that most often confuses newcomers, so it is worth saying plainly. On an exchange style prediction market there is no house. When you buy a yes contract, you are not betting against the platform, you are buying from another participant who is selling. The venue is the matchmaker, not the opponent. It earns a fee for bringing the two of you together, and it makes the same fee whether your side wins or loses. Compare that with a traditional sportsbook, which sets the odds itself, takes the other side of your wager, and profits structurally when bettors lose. The two look similar from the outside and are built on opposite incentives underneath.
That structural difference is not a marketing point, it is the reason the category is regulated the way it is. In the United States, federally regulated event contract exchanges are treated as commodity derivatives venues overseen by the Commodity Futures Trading Commission, not as gambling operators, precisely because they match traders rather than book bets. We explain that oversight in detail on the role of the CFTC, and the practical contrast in how regulated exchanges differ from offshore. The short version is that an exchange where participants trade with each other, and where the contract is classified as an event contract, sits inside a derivatives framework rather than a gambling one.
It also changes how to read the price. A bookmaker bakes a margin into its odds, so the implied chances it shows add up to more than 100 percent and the surplus is its built in edge. An exchange price has no such house margin, which is why it is closer to a clean probability. It is not perfectly clean, because the fee and the gap between the best buy and sell prices still take a slice, but the starting point is an honest estimate rather than a number tilted in the operator favour. That is the single biggest reason analysts and journalists quote prediction market prices as probabilities at all.
Most serious venues run a central limit order book. It is a live list of every resting buy order and sell order at each price, sorted so the best prices sit at the top. Buyers post the most they will pay, called bids. Sellers post the least they will accept, called asks. When a bid meets an ask at the same price, a trade fills. Kalshi runs this model under federal oversight, and Polymarket runs a central limit order book where orders are matched and then settled on a blockchain, with every share priced between zero and 1 dollar so the price reads directly as probability, per Polymarket documentation as of June 2026.
Two ideas make the book fair and predictable. The first is price priority: the highest bid and the lowest ask are first in line, because they are offering the best deal to the other side. The second is time priority: among orders at the same price, the one that arrived first fills first. Orders can also fill in part, so a large order may be matched piece by piece against several smaller ones on the other side. A resting order that waits on the book is called a maker order because it makes liquidity, and an order that crosses the spread to trade immediately is called a taker order because it takes liquidity. Many venues charge those two differently, which is why the same trade can cost more or less depending on whether you wait or cross.
The shape of the book is also information. A deep book with size stacked at many price levels means you can trade a meaningful amount without moving the price much. A thin book with only a few small orders means even a modest trade can jump the price, and the quote you see may not be the price you actually get. That hidden cost is called slippage, and it is one of the most underrated risks for a newcomer reading a tidy looking price on a quiet market.
The word prediction market covers several quite different things, and the difference decides who can use a venue, how your money is held, and what happens if a result is disputed. The table below sets out the broad families. The platform profiles on this site cover each named venue one at a time, and the legality pages cover where each is available, because availability depends on where you live and the rules change.
| Venue family | How it matches | Who holds funds | Oversight |
|---|---|---|---|
| Regulated US exchange for example Kalshi, ForecastEx | Central limit order book, traders matched with each other | Held with the regulated exchange and its arrangements | CFTC, as a designated contract market under the Commodity Exchange Act |
| Onchain venue for example Polymarket | Order book matched off chain, settled on a blockchain, peer to peer | You hold funds in a wallet; settlement runs on chain | Outside the US retail framework; availability is restricted and contested |
| Play money or reputation for example Manifold Markets, Metaculus | Trading or forecasting with virtual currency or scored predictions | No real money at stake in the markets themselves | Not a money venue; used for forecasting practice and research |
Methodology: families summarised from each platform published terms and from CFTC material, as of June 2026. Example venues are named for illustration of the model, not as recommendations. Classification, availability, and fund handling vary by venue and region and can change. Check the platform profile and the legality pages, and verify the current terms on each venue, before relying on any of this.
A market is only as good as the way it ends. Before trading opens, every well built market states exactly what counts as yes and what counts as no, and it names the source of truth that will decide the outcome. That source might be an official government data release, a final score, a certified election result, or a recorded price at a stated time. Fixing the criteria up front is what stops the outcome from becoming an argument later, and it is the difference between a clean contract and a vague one. A reader should always check the resolution wording before treating a market as meaningful, because an ambiguous question can resolve in a way that feels wrong even when the rules were followed.
When the event is decided, settlement is mechanical. Winning contracts redeem for their full dollar and losing contracts expire at zero. On a regulated exchange this is handled by the venue against the named source. On an onchain venue the outcome is reported by an oracle, a mechanism that brings the real world result onto the blockchain, and a dispute process allows a contested result to be challenged before it is finalised. That dispute layer is a genuine difference from a regulated exchange, and it is one of the reasons settlement on decentralised venues can occasionally be slower or more contentious. We cover the detail in how event contracts settle and in resolution disputes and how they work.
None of this removes uncertainty about the event itself. It only removes uncertainty about how the event will be judged. That is an important distinction. A market can be perfectly well designed, with crisp criteria and a reliable source, and still leave you holding a contract that loses, because the event simply did not go the way the price implied. Good resolution rules protect you from being cheated by ambiguity. They do not protect you from being wrong, or from the market being wrong.
In the United States the central regulator for event contract exchanges is the Commodity Futures Trading Commission, which oversees these venues as designated contract markets under the Commodity Exchange Act. That status is what lets a venue list event contracts to the public under federal rules rather than under state gambling law. It is also why the classification of a given contract, and whether a particular category of event is permitted, can be a live legal question rather than a settled one.
The position is genuinely in motion, and we flag that plainly rather than pretending it is fixed. The CFTC reported that total volume across its designated contract markets exceeded 25 billion dollars in 2025, a sign of how quickly the category has grown. In 2026 the regulator moved to formalise the framework: it issued an advance notice of proposed rulemaking in March 2026 seeking comment on event contracts, and a notice of proposed rulemaking in June 2026 that would revise the rules governing certain event contracts, including the standards for when a contract may be treated as contrary to the public interest, per CFTC notices as of June 2026. Because these are proposals rather than final rules, the detail may change, and which contract types are clearly permitted remains a contested and developing area. We maintain the current picture on the role of the CFTC and track changes as they land.
For a reader, the practical takeaway is simple. The fact that a venue is regulated tells you about its structure and oversight, not about whether any particular outcome is likely, and not about whether the venue is available to you where you live. Legality and availability are separate questions that depend on your region and on rules that move, so the safe habit is to verify your own eligibility on the legality pages before doing anything, and to treat the regulatory backdrop as a moving picture rather than a fixed one.
It is fair to ask why a price set by a crowd of strangers should be any good. The answer is that a market is not a simple poll. It is a weighted vote in which the people most confident in their view, and most willing to back it with money, move the price the most. Someone who believes the market is underpricing an outcome buys, lifting the price; someone who thinks it is overpriced sells, lowering it. The number settles where those pressures balance. That mechanism pulls in scattered private information, a poll here, a model there, a piece of local knowledge somewhere else, and compresses it into one figure that updates the moment anything changes.
This is the wisdom of crowds, with a sharpening twist: the votes that count most are the ones with money behind them, which tends to weight informed and committed participants more heavily than idle opinion. It is why reviews of these markets find prices track real frequencies reasonably well and tend to edge out bookmaker odds, and why a market with many active traders is more trustworthy than a thin one with few. But the same logic sets the limit. When a market is thin, there are fewer informed votes holding the number in place, so the price is softer and easier to push around. The mechanism is powerful where it is deep and weak where it is shallow, and reading a price well means knowing which you are looking at. We go further in the wisdom of crowds and markets and in reading prices as implied probability.
An informative price is not a promise. A contract at 60 cents still loses about four times in ten when the price is right, and prices are often wrong. The structure does not protect you from the outcome.
Fees and the spread between buy and sell prices reduce returns on every round trip, and frequent trading multiplies that drag. The cheaper the contract, the larger the cost looks in proportion.
If few people are trading, you may not be able to enter or exit at the price you see, and a single order can move a thin book. That gap between the quote and the fill is a hidden cost.
A contested resolution, an oracle dispute, or a question about where and how your funds are held are real risks, and they differ sharply between a regulated exchange and an offshore or onchain venue.
Availability and legal status differ by region and move quickly. What is allowed for one reader may not be for another, and what is allowed today may change. Verify your own eligibility before acting.
Event contracts carry a real risk of loss, and understanding the mechanics does not change that. Trade only with money you can afford to lose, decide your limits before you start, and stop if it stops being a considered decision. Participation is for those who are 18 or the legal age in their region. If trading is affecting your wellbeing or your finances, free and confidential support is available in the United States through the 1-800-GAMBLER helpline run by the National Council on Problem Gambling.
The mechanics are the same idea everywhere, but the venue, the fees, and the legal status are not. Availability depends on where you live, so check legality first. These reference pages are information, not an introduction to any single platform.
A prediction market is a venue where people trade contracts tied to a defined future event. The contract pays a fixed amount, usually 1 dollar, if the event happens and nothing if it does not, so its price sits between those values and reads as the market estimate of how likely the event is.
A yes contract that settles at 1 dollar trades between 1 cent and 99 cents. A price of 60 cents implies the market thinks there is roughly a 60 percent chance the event happens, per the way Kalshi and similar venues describe their contracts as of June 2026. It is an estimate set by traders, not a guarantee, and it moves as people trade.
No. On an exchange style prediction market there is no house setting odds and taking the other side of your position. You trade with other participants through an order book and the venue earns from fees. That difference is central to how the category is regulated in the United States.
Each market has defined resolution criteria and a stated source of truth set before trading. When the event is decided, winning contracts redeem for full value and losing contracts expire at zero. On decentralised venues an oracle reports the outcome and a dispute process can challenge a contested result.
Federally regulated event contract exchanges are overseen by the Commodity Futures Trading Commission as designated contract markets under the Commodity Exchange Act. The CFTC reported volume across these markets exceeded 25 billion dollars in 2025 and issued proposed rulemaking in 2026, so the framework is active and evolving.
No. They carry a real risk of losing money, and an informative price is not a promise. Fees, thin liquidity, settlement risk, and contested legality all matter. This page is general information, not financial, legal, or tax advice.
Reviewed by Fredrik Filipsson, Editor, on 26 June 2026. Mechanics, venue families, and the regulatory position checked against platform terms and CFTC material current at that date.