A prediction market is a place where you trade contracts on the outcome of a real-world event — an election, next month’s inflation print, a football game, or where Bitcoin closes on Friday. Each contract pays $1 if the event happens and $0 if it doesn’t, so its price between 1 and 99 cents reads directly as the market’s estimate of the odds: a contract at 63 cents means the crowd prices the event at roughly a 63% chance.

That single mechanic — a price that is a probability — is what makes prediction markets different from both gambling and traditional investing. This guide explains how the contracts work, how prices are set, how markets resolve and pay out, whether the odds are actually accurate, and what to watch before you trade, as of July 2026.

Key takeaways

  • A prediction market trades event contracts — yes/no positions on whether a defined future event happens, each settling at $1 if you’re right and $0 if you’re wrong.
  • The price is the probability. A contract at 30 cents implies a 30% chance; because “Yes” and “No” always sum to $1, the price is the crowd’s live odds.
  • Your loss is capped at what you paid. Buy a contract for 40 cents and 40 cents is the most you can lose — there’s no margin call and no unlimited downside.
  • The odds are usually well-calibrated. Across 28,407 resolved Polymarket markets (January 2024–May 2026), events priced at 40 cents resolved yes 41% of the time, per research summarized on SSRN — a mean calibration error of about 2 percentage points.
  • The two giants are Kalshi and Polymarket, which together traded roughly $44.8 billion in June 2026, a category that barely existed at that scale a year earlier.

What is a prediction market?

A prediction market is a financial exchange where the “asset” you trade is the outcome of a specific future event rather than a company or a commodity. Instead of buying a share of Apple, you buy a share of “Yes, the Federal Reserve cuts rates in September 2026.” If that turns out to be true, your share is worth $1; if not, it’s worth nothing.

The two largest venues in 2026 are Kalshi, a US exchange regulated by the Commodity Futures Trading Commission (CFTC), and Polymarket, a crypto-native platform that settles trades in stablecoins. They list thousands of markets across politics, economics, sports, crypto prices, and world events. We break down each venue in our guide to what Kalshi is and our head-to-head on Kalshi vs Polymarket.

The label “prediction market” captures the whole point: the market’s job isn’t just to let people bet — it’s to produce a prediction. The price is the output.

How event contracts work

Every market is built around a question with a clear, objective resolution rule — for example, “Will US CPI come in above 3.0% for July 2026?” You take a side by buying “Yes” or “No” contracts. Here is the mechanic that makes the whole thing work:

  • Each contract settles at $1 or $0. If your side is correct, every contract you hold pays out exactly $1. If it’s wrong, it expires worthless.
  • Contracts trade between 1 and 99 cents. A “Yes” at 30 cents costs you 30 cents and pays $1 if you’re right — a bit more than 3x your money. A “Yes” at 90 cents costs 90 cents to make 10, because the market thinks it’s very likely.
  • Yes and No always add up to $1. If “Yes” is 30 cents, “No” is 70 cents. A buyer of “Yes” is effectively matched against a buyer of “No,” and the exchange settles the pair when the outcome is known.
  • Your downside is fixed. Unlike margin trading or leveraged futures, you can never lose more than you paid. Buy at 40 cents and 40 cents is your maximum loss, known in advance.

This binary, loss-capped structure is why prediction-market contracts are often called “event contracts.” It’s also why they behave differently from a sportsbook bet, where the house sets the odds and profits from your losses.

Why the price equals the probability

The elegant part is that the price is the forecast. Because a winning contract is worth exactly $1, a rational buyer will pay up to their honest estimate of the odds and no more. If you think there’s a 70% chance of an event, you’ll buy “Yes” below 70 cents and sell above it. Multiply that behavior across thousands of traders staking real money, and the price settles at the crowd’s collective probability.

Prices then move in real time as new information arrives. When a jobs report drops or a candidate stumbles, buyers and sellers reprice within seconds, and the contract’s cents-value tracks the shifting odds live — the way a stock reprices on earnings news. By the time the event resolves, the market has produced a continuously updated, crowd-sourced probability estimate. This “markets find truth” idea is the thesis we explore in Kalshi and the new Wall Street.

How trades are matched: order books and market makers

Under the hood, a prediction market matches buyers and sellers in one of two ways.

  • Central limit order book (Kalshi). Like a stock exchange, Kalshi keeps a live book of bids and offers. You can place a limit order to buy “Yes” at 32 cents and wait for a seller, or take the best available price immediately. The two sides of every market — Yes buyers and No buyers — provide each other’s liquidity.
  • Automated market maker and order book (Polymarket). Polymarket uses a blockchain-based system where trades settle in USDC, a dollar-pegged stablecoin, on the Polygon network. Liquidity providers and an order-matching layer keep quotes available so you can enter and exit. If you’re new to how dollar-pegged tokens function, our explainer on stablecoins covers the settlement layer underneath.

Either way, the platform earns money from trading fees, not from taking the other side of your position. That is a structural difference from a bookmaker: an exchange profits from volume regardless of who wins, so its incentives point toward liquidity rather than toward your loss.

How a market resolves and pays out

When the event’s outcome becomes known, the market “resolves,” and every contract is redeemed at $1 or $0. Resolution depends on a pre-defined source of truth spelled out in the market’s rules:

  • Objective data sources. A market on CPI resolves to the official Bureau of Labor Statistics release; a market on a game resolves to the final score. Regulated exchanges like Kalshi name the settlement source in each contract’s terms.
  • Oracles (on crypto-native platforms). Polymarket relies on a decentralized resolution system — an “oracle” — where the reported outcome can be proposed and, if disputed, adjudicated before funds are released. Most resolutions are routine; contested edge cases are the ones that make headlines.

Because you don’t have to wait for resolution, you can also sell before the event. If your “Yes” contract climbs from 30 to 55 cents as the news turns your way, you can close the position and lock in the gain — you’re trading the probability, not just holding a bet to the end.

Are prediction market odds accurate?

Mostly, yes — and there’s now hard data. Research summarized on SSRN examined 28,407 Polymarket markets that resolved between January 2024 and May 2026 and found the prices were well-calibrated: contracts priced at 40 cents resolved “Yes” about 41% of the time, those at 70 cents resolved “Yes” about 72% of the time, with a mean absolute calibration error of roughly 2 percentage points. In plain terms, the price is a reliable read of the true odds.

But why they’re accurate is more nuanced than the popular “wisdom of crowds” story. A separate 2026 study covering 292 million trades across 327,000 binary contracts on Kalshi and Polymarket, highlighted by Yale Insights, found that accuracy comes not from the broad crowd but from a small, persistently skilled minority — roughly 3% of traders — who react fastest to news and correct mispricing. Those same traders also capture most of the profits. Two practical caveats follow: thinly traded markets (low volume, low liquidity) are far less reliable, and political markets in particular tend to be underconfident, with prices compressed toward 50%.

Prediction markets vs betting and stocks

It helps to place prediction markets between the two things people confuse them with:

  • Versus a sportsbook. A bookmaker sets the odds and profits when you lose; a prediction market is a peer-to-peer exchange where other traders set the price and the platform earns fees. Odds on an exchange move with supply and demand, not with a house margin. The distinction is real, though the risk of short-term trading feels similar — a collision we cover in are prediction markets replacing sportsbooks?.
  • Versus stocks. Like a stock, a contract has a live price you can trade in and out of. Unlike a stock, it has a hard expiry and only two final values — $1 or $0 — so there’s no dividend, no compounding, and no “hold forever.”

What to know before you trade

Prediction markets are real markets with real risk. As of July 2026, keep a few things in mind:

  • You can lose your whole stake. A losing contract goes to zero. The trade-off is that your downside is fixed and known upfront.
  • Liquidity matters. Thinly traded markets can swing sharply and be hard to exit at a fair price. Well-funded markets are better calibrated.
  • Regulation is still moving. Kalshi operates under CFTC oversight and Polymarket now runs a regulated US arm, but sports contracts face state-level challenges. We map the picture in are prediction markets legal in the US?.
  • It’s a tool for views, not free money. Because the sharpest traders drive the accuracy, beating a well-calibrated market consistently is hard. Treat a trade as a way to price a probability, not as an edge.

The bottom line

Prediction markets work by turning a question into a tradeable contract that pays $1 if the answer is yes and $0 if it’s no. The price you pay — somewhere between 1 and 99 cents — is the crowd’s live estimate of the odds, updated in real time as information arrives and settled against an objective source when the event resolves. That design makes them a genuinely useful forecasting tool and a fast-growing corner of finance, but the same data that shows the odds are accurate also shows the profits flow to a skilled few. Trade them to express a view, understand the resolution rules, and never stake more than you can afford to lose.

Frequently asked questions

How do prediction markets work in simple terms?

You buy a yes/no contract on whether a future event will happen. Each contract costs between 1 and 99 cents and pays out $1 if you’re right or $0 if you’re wrong. The price is the market’s implied probability, so a contract at 65 cents means the crowd thinks there’s about a 65% chance of the event, and you can sell your position before the event resolves.

How is the price of a prediction market contract set?

The price is set by supply and demand between traders on an exchange, not by a bookmaker. Because “Yes” and “No” contracts always add up to $1, the price of “Yes” is the crowd’s collective probability estimate. When new information arrives, buyers and sellers reprice the contract in real time, so the odds update continuously until the event resolves.

Are prediction market odds accurate?

Generally yes. Research summarized on SSRN found that across 28,407 resolved Polymarket markets from 2024 to 2026, prices were well-calibrated to within about 2 percentage points — events priced at 40 cents happened roughly 41% of the time. Accuracy is highest in liquid, high-volume markets and weaker in thinly traded ones, and studies suggest a small minority of skilled traders drives most of the accuracy.

Can you lose money on a prediction market?

Yes. If your contract resolves against you it settles at $0 and you lose everything you paid for it. The upside is that your loss is capped at your stake — you can never lose more than the price you paid — so there is no margin call or unlimited downside as there can be with leveraged trading.

What’s the difference between a prediction market and sports betting?

A sportsbook sets the odds and profits when you lose, taking the other side of your bet. A prediction market is a peer-to-peer exchange where other traders set the price through buying and selling, and the platform earns trading fees regardless of who wins. That makes the odds market-driven rather than set by a house margin.

Sources