Prediction Markets Guide 2026-2027: How They Work, Platforms, Categories, Legal Status and How to Start

Prediction markets went from a niche corner of crypto and political forecasting to an industry trading tens of billions of dollars a month in under three years. Kalshi traded $13.07 billion in March 2026 and about $37 billion in August on The Block’s count, while Polymarket’s two venues together did about $8.2 billion that month. Sports drive most of Kalshi’s volume, DraftKings, FanDuel and Robinhood all run their own products, and the Supreme Court has been asked to decide whether states can treat sports contracts as gambling. This guide was last fact-checked in late October 2026.

This guide covers everything someone approaching prediction markets in 2026 actually needs to understand: how they work, which platforms matter, what categories have the most liquidity, the legal status (it is complicated), the real risks, and how to start trading if you decide to. It is written for both new entrants and experienced traders looking for a grounded 2026 view without the hype.

For deeper reading on specific platforms, see our Polymarket review, Kalshi review, or our full comparison of the best prediction markets platforms.

Quick Verdict

Prediction markets in 2026 are a legitimate, fast-growing asset class that sits between traditional derivatives trading and sports betting. They offer genuine edge for traders who understand probability and market microstructure, produce forecast accuracy that beats most pollsters, and create clear entertainment value for casual users. They also carry real risks: legal uncertainty in multiple US states, opaque oracle resolution in crypto-based markets, and wash trading or information asymmetries that punish unprepared participants. Approach them as a new category of financial market, not as “better sports betting.”

What Are Prediction Markets

A prediction market is a marketplace where users trade contracts tied to the outcome of real-world events. Each contract pays $1 if the event resolves “yes” and $0 if it resolves “no.” The price between $0.01 and $0.99 represents the market’s implied probability of the event happening.

For example, a contract “Will the Fed cut rates in June 2026?” priced at $0.42 implies the market assigns a 42% probability to that outcome. If the Fed cuts rates, “Yes” holders receive $1 per contract. If not, the contract expires worthless.

This structure does three things simultaneously. It lets users speculate on outcomes they believe are mispriced. It aggregates information across thousands of participants into a single live probability. And it creates a settlement mechanism that pays winners and collects losses automatically.

For a formal definition, see our what are prediction markets page. This guide focuses on the practical landscape.

A Short History

Prediction markets are not a 2024 invention. They trace back to informal political betting on Wall Street dating to 1884, with some forms of event betting going back to 1503. The modern digital version started with the Iowa Electronic Markets in 1988, a University of Iowa research project that still operates with small-stakes political and economic contracts.

The key milestones that brought prediction markets to where they are in 2026:

• 1988: Iowa Electronic Markets launches as the first modern electronic prediction market
• 2003: Intrade goes live and becomes the largest real-money prediction market globally before shutting down in 2013
• 2014: PredictIt launches with CFTC no-action relief for political event contracts
• 2020: Polymarket founded, later running on the Polygon blockchain with USDC for trading. Kalshi is designated by the CFTC as a contract market built for event contracts
• 2021: Kalshi opens for trading
• 2024: Polymarket processes over $3.5 billion during the US presidential election cycle. Federal court rules in Kalshi’s favor against the CFTC over political event contracts
• December 2025: Polymarket US relaunches via the QCX acquisition as a CFTC-regulated exchange for US users. DraftKings and FanDuel launch their own prediction products
• March 2026: Combined Kalshi and Polymarket monthly volume reaches about $23.6 billion
• April to September 2026: Federal appeals courts split on whether states can apply gambling law to sports contracts, and New Jersey asks the Supreme Court to decide
• August 2026: Kalshi trades about $37 billion in a single month

This is the fastest growth trajectory of any new financial market category since crypto derivatives. The combination of regulatory clarity in some areas, political interest driving awareness, and sports contracts creating volume has accelerated everything.

How Prediction Markets Work: Mechanics

Binary contracts and pricing

Almost every major prediction market contract is binary. Each contract has a “Yes” side and a “No” side. Prices range from $0.01 to $0.99 and sum to $1.00 between the two sides (minus small platform fees or spreads). If you buy 100 “Yes” shares at $0.65, you pay $65 for the position. If “Yes” resolves true, you receive $100 ($35 profit). If it resolves false, you lose your $65.

The price reflects the market’s aggregate probability estimate. Prices move as new information arrives, traders place new orders, and sentiment shifts. Active markets on Kalshi and Polymarket routinely see spreads of 1-2 cents, comparable to institutional-grade financial products.

Order books and liquidity

Prediction markets use open order books, like stock exchanges. You can place limit orders (buy or sell at a specific price) or market orders (execute immediately at the best available price). Maker orders that provide liquidity to the book typically pay no fees. Taker orders that remove liquidity pay fees that depend on the price, peaking at $1.75 per 100 contracts at 50 cents on Kalshi and on Polymarket’s crypto markets.

Liquidity is highly concentrated. The top 50 markets on Polymarket and Kalshi account for most of the daily volume, while long-tail markets can have spreads of 10+ cents and minimal depth. The practical implication is that trading strategy differs dramatically between high-volume markets and niche ones.

Resolution and settlement

Every contract has a resolution date and clearly defined settlement rules set at creation. When the event occurs, an oracle (a system that bridges real-world information with the platform) checks trusted data sources, verifies the outcome, and reports the result. This triggers automatic payout to winners.

Resolution mechanics differ by platform. Kalshi uses internal settlement tied to official data sources (government reports, sports league data, verified news). Polymarket uses UMA’s Optimistic Oracle, a decentralized protocol where an approved proposer posts a bond, typically $750 USDC, behind an outcome. If no one disputes within a 2-hour challenge period, the proposal becomes final. Disputed outcomes go to UMA token-holder voting.

The oracle mechanism matters because it determines trust. Kalshi’s settlement is operationally simpler and less contentious. Polymarket’s UMA system has resolved the vast majority of markets without controversy but has produced occasional high-profile disputes on ambiguous outcomes.

Making money

Users profit three main ways. First, by holding winning contracts to resolution and receiving $1 per share. Second, by trading in and out of positions as probabilities shift, locking in gains before resolution. Third, by market making (providing liquidity on both sides of a market to capture the spread).

Losses happen when you buy contracts at prices higher than the final outcome probability, when you panic sell at unfavorable prices, or when oracle disputes resolve against your position. Experienced traders treat prediction markets as a probability game where the edge comes from better information, better pricing discipline, and better risk management than the average participant.

The Major Prediction Market Platforms

The prediction markets landscape in 2026 centers on two dominant platforms plus a growing set of secondary players.

Kalshi

Kalshi is the largest prediction market in the world and the biggest CFTC-regulated one. It traded $13.07 billion in March 2026 and about $37 billion in August, mostly sports. It raised at $11 billion in December 2025 and about $22 billion in March 2026, and Bloomberg reported in September that it was finalising a round at $40 billion, which Kalshi had not confirmed. The taker fee is 0.07 × contracts × price × (1 − price), which works out to anywhere from under 1% to nearly 7% of the stake depending on the price, and US users earn a variable 3.50% APY on cash balances.

Kalshi’s strength is regulatory clarity in the US. Its weakness is geographic restriction (mainly US users) and the ongoing state-level legal battles over sports contracts. For a full breakdown, read our Kalshi review.

Polymarket

Polymarket is the largest prediction market outside the US, with March 2026 volume of $10.57 billion, a record at the time. Politics ($2.97 billion) and crypto ($2.72 billion) made up most of it. The platform operates as two products: Polymarket International (global, crypto-native, unregulated) and Polymarket US (CFTC-regulated via QCX, open without a waitlist since May 2026).

Polymarket’s fee structure was originally zero. Taker fees now peak at between $1.00 and $1.75 per 100 shares depending on category, and geopolitics markets are free. Settlement uses the UMA Optimistic Oracle. Funding is USDC on Polygon for international users, dollar deposits for US users. For the full review, see our Polymarket review.

DraftKings Predictions

Launched on December 19, 2025, first on CME and since June 2026 on DraftKings’ own exchange, DKeX, DraftKings Predictions brings CFTC-regulated event contracts to DraftKings’ existing customer base. It is available in 48 states, with sports contracts in 18 of them. Volume is growing but remains well below Kalshi’s level. FanDuel Predicts, Flutter’s US product built with CME, launched three days later and follows a similar model, with sports contracts in 18 states since January 2026.

Robinhood Predictions

Robinhood offered US presidential election contracts in October 2024 and opened a prediction markets hub powered by Kalshi in March 2025, covering politics, economic data and sports. In June 2026 it launched its own exchange, Rothera, a joint venture with Susquehanna. Users traded 13.6 billion contracts in the second quarter of 2026. Robinhood charges a commission of up to a cent per contract, and exchange fees can apply on top. For Robinhood’s existing user base, this is the lowest-friction entry point to prediction markets in the US.

OG by Crypto.com

Crypto.com launched OG as a standalone prediction markets platform on February 3, 2026, running on its CFTC-registered exchange, Crypto.com | Derivatives North America (CDNA). Focus is US sports. Crypto.com has announced margin trading for OG, which would be a first for a regulated prediction market, but it was not live as of October 2026.

Manifold

Manifold is a reputation-based prediction market using play-money “mana”. It ended its sweepstakes cash option in March 2025, so nothing on the platform converts to dollars any more. The platform focuses on long-tail markets, user-created questions, and community-driven forecasting. Unlike Kalshi or Polymarket, Manifold targets forecasting culture rather than trading volume.

Secondary and emerging platforms

• Truth Predict: Trump Media’s prediction market, announced with Crypto.com’s CDNA infrastructure and still in beta rather than fully launched as of October 2026
• Betfair Predicts: Flutter’s UK product, in invite-only beta since April 2026
• Matchbook: UKGC-licensed exchange running a B2B prediction markets platform since December 2025; easyBet Predictions went live on it in January 2026
• ADI Predictstreet: Licensed in Gibraltar in March 2026 as a betting intermediary, with a FIFA World Cup 2026 partnership
• PredictIt: Political markets under CFTC no-action relief; since July 2025 the cap is $3,500 per contract and the 5,000-trader limit is gone
• Augur: Decentralized Ethereum-based platform, dormant for years, with a reboot announced in September 2025

Prediction Markets vs Sports Betting vs Traditional Gambling

This distinction matters for regulation, user experience, and trading strategy. The boundaries are genuinely fuzzy, and the answer depends on who you ask.

Structural differences

Sports betting operates through bookmakers who set odds and take the opposite side of user bets. The house has a built-in margin (the vig or overround) that guarantees profit over volume. When you lose a bet, the bookmaker wins directly.

Prediction markets operate through peer-to-peer exchanges. The platform matches buyers and sellers and collects fees, but does not take positions. When you lose a contract, another user wins directly. Pricing is driven by supply and demand rather than set by an operator.

Traditional gambling (roulette, slots, most casino products) has a fixed house edge built into the mathematics of the game. The outcome depends on randomness, and the long-term expected return is negative.

Practical differences

• Pricing: Sportsbooks use odds (-110, +150). Prediction markets use dollar prices between $0.01 and $0.99.
• Liquidity management: Sportsbooks always accept your bet (up to limits). Prediction markets require matching counterparties.
• Exit before resolution: Prediction markets allow easy position exit at current market price. Sports bets typically require “cash out” features at disadvantageous prices.
• Fee structure: Sportsbooks bake margin into odds. Prediction markets charge transparent percentage fees.

Regulatory framing

The debate over whether prediction markets are gambling or financial instruments is not academic. It determines which regulator has jurisdiction, what licensing is required, and who can legally participate. In the US, the federal appeals courts are split on this question, which is why Kalshi’s sports contracts are live in most states but geofenced in a few.

The regulatory category question matters because different license frameworks apply depending on how a platform is classified. For the full picture of license categories operating globally, see our types of gambling licenses classification guide.

Market Categories

Understanding where liquidity concentrates tells you where the serious money is and where strategies work.

Sports

Sports dominate prediction markets volume in 2026. Between March 2 and 29, DeFi Rate counted $9.9 billion of Kalshi’s $11.38 billion in volume coming from sports contracts, and Polymarket lists thousands of sports markets at any given time. Categories include game winners, championship outcomes, player performance milestones, season records, and in-game events updated in real time.

European football, US major leagues (NFL, NBA, MLB, NHL), and international events (World Cup, Champions League, Olympics) drive the deepest liquidity. Sports markets are where prediction markets most directly compete with traditional sportsbooks.

Politics and elections

Politics is the category that built Polymarket’s reputation, and it still lists a deep catalogue of political markets. Categories include US elections at federal and state level, UK political developments such as Labour leadership contracts, European elections, cabinet appointments, legislative votes, approval ratings, and geopolitical outcomes.

Politics is where prediction markets most visibly compete with polling for forecast accuracy. During the 2024 US election, prediction markets called the outcome in real time while major pollsters were still showing it as a toss-up.

Crypto and finance

Crypto-adjacent contracts are one of Polymarket’s two biggest categories, generating $2.72 billion over most of March 2026. Polymarket runs thousands of crypto markets covering Bitcoin price milestones, Ethereum targets, altcoin outcomes, and token launch predictions. Financial markets cover Fed rate decisions, inflation targets, equity milestones, commodities, and corporate earnings outcomes.

Geopolitics

Geopolitical contracts cover international conflicts, diplomatic events, sanctions outcomes, and treaty decisions. The 2026 Strait of Hormuz crisis made oil and shipping one of Polymarket’s fastest-growing subcategories.

Technology and science

Tech markets cover AI breakthroughs, IPO outcomes, product launches, corporate milestones, and scientific discoveries. SpaceX launch success, AI model releases, and IPO pricing are consistent volume drivers. Scientific markets (climate, medical, research) tend to have thinner liquidity but attract dedicated forecasting communities.

Culture and entertainment

Culture markets cover awards show winners, box office performance, music chart outcomes, celebrity events, and viral moments. These markets tend to see attention-driven volume spikes rather than sustained activity.

The unusual

Some markets get creative. Kalshi has hosted contracts on whether specific words appear in earnings calls, celebrity outfit choices, legal case outcomes, and weather milestones. The long tail of prediction markets is where forecasting communities experiment with the edge of what can be priced.

Legal Status by Jurisdiction

This is where things get genuinely complicated. Prediction markets occupy different regulatory categories in different places, and the situation is evolving fast.

United States

Federally, prediction markets are regulated by the CFTC as event contracts or swaps. Kalshi holds a Designated Contract Market license. Polymarket US operates as a CFTC-regulated exchange via the QCX acquisition. This gives both platforms federal legitimacy.

The state-level picture is messier. Multiple states argue that sports event contracts are gambling under state law, regardless of CFTC status. Kalshi has geofenced sports contracts in Nevada, Washington and Michigan under state court orders. A Massachusetts injunction from January 2026 was stayed on appeal and the state’s top court has not ruled. Arizona filed criminal charges against Kalshi in March 2026, and the CFTC itself sued Arizona, Connecticut and Illinois in April and later Minnesota, Wisconsin and New York. The federal appeals courts have split: the Third Circuit sided with Kalshi, while the Ninth and Sixth sided with the states. New Jersey has asked the Supreme Court to decide. In October 2026 the CFTC also issued an interim final rule and a proposed rule that would treat sports, political, cultural and weather event contracts as swaps.

United Kingdom

The UK Gambling Commission said in February 2026 that prediction market products would appear to fall within the definition of a “betting intermediary” under UK gambling law. In practice that means operators serving UK customers need a Gambling Commission licence, not FCA authorisation. So far the licensed route runs through exchanges: Matchbook launched a B2B prediction markets platform in December 2025, easyBet Predictions went live on it in January 2026, and Betfair Predicts has been in invite-only beta since April 2026. Polymarket is close-only for UK users and Kalshi restricts the UK.

European Union

Most EU member states that have taken a position treat prediction markets as unlicensed gambling. France, Germany, the Netherlands, Belgium, Portugal, Italy, Spain, Denmark, Poland, Romania, Hungary and the Czech Republic have all warned against, blacklisted or blocked Polymarket, and Italy, Spain and the Czech Republic have moved against Kalshi too. In June 2026 nine national regulators pledged coordinated action against unlicensed prediction markets, and in July ESMA said some of these contracts may be binary options under EU financial law.

Geoblocked jurisdictions

Polymarket publishes its own restriction list. As of October 2026 it fully blocks Iran, Syria, Cuba and North Korea and sets 34 countries to close-only, meaning existing positions can be exited but no new ones opened. That group includes the US (for the international platform), the UK, France, Germany, Italy, Belgium, Poland, Singapore, Taiwan, Thailand, Australia, New Zealand, Brazil and Argentina, plus Ontario, Quebec, British Columbia and Alberta. National regulators have separately ordered blocks in places such as Spain, Denmark, Portugal, South Korea, India, the Philippines and Indonesia.

Canada

Canada splits by province. Ontario’s securities regulator settled with Polymarket’s operators in April 2025, and Polymarket is now close-only in Ontario, Quebec, British Columbia and Alberta. Kalshi restricts Canada directly, but Wealthsimple Predict, which offers Kalshi contracts, was authorised by the investment industry regulator CIRO in March 2026, giving Canadians a regulated route through the securities side.

Rest of world

The rest of the world is closing fast. Brazil blocked 27 platforms including Polymarket and Kalshi in April 2026 and banned derivatives on sports, election and entertainment events. Argentina, Colombia, South Korea, India, the Philippines, Indonesia, Singapore and Taiwan have all blocked Polymarket, and Australia and New Zealand treat it as illegal gambling. Grey areas remain in places such as Chile, Mexico, Peru and Japan, and in much of Africa, where the practical question is often whether local payment rails support crypto-based platforms.

For a deeper dive on this topic, see our prediction markets regulation guide.

How to Start Trading

The practical path depends on where you are and what you want to trade.

US users

• For sports and politics with regulatory clarity: open a Kalshi account, verify identity, fund via bank transfer or debit card
• For the Robinhood interface: trade event contracts in the Robinhood app if you already have an account
• For Polymarket’s markets in a regulated wrapper: open a Polymarket US account (no waitlist since May 2026; not available in Nevada, Connecticut or Tennessee)
• For sports inside the Crypto.com ecosystem: sign up to OG by Crypto.com
• If you already bet with DraftKings or FanDuel: their prediction apps are available in most states, but sports contracts only in 18

International users

• Outside the US, UK, EU restricted states, and geoblocked jurisdictions: Polymarket International is the primary option
• Setup requires a crypto wallet (MetaMask works well), USDC on Polygon for funding, and familiarity with crypto mechanics
• Alternatively, MetaMask Prediction Markets offers a wrapped version of Polymarket with simpler onboarding

UK users

• easyBet Predictions, running on Matchbook’s UKGC-licensed exchange, has been live since January 2026
• Betfair Predicts has been in invite-only beta since April 2026 under Flutter’s licence
• Polymarket is close-only for UK users and Kalshi restricts the UK, and neither holds a UK licence

Starting strategy

For new users, the practical recommendation is to start small (under $100), focus on markets you genuinely understand, and track your decisions over at least 50-100 trades before concluding anything about your skill. Prediction markets reward discipline and pattern recognition over time. Most new users lose money in their first 3-6 months because they overweight narrative conviction and underweight liquidity dynamics.

Risks and Common Mistakes

Regulatory risk

Laws change and platforms can lose access to markets overnight. US state-by-state enforcement against sports contracts is an active risk. A Supreme Court ruling against federal CFTC authority would force major platform restructuring. Non-US users face similar risks as jurisdictions clarify their positions.

Platform risk

Centralized platforms (Kalshi, DraftKings, Robinhood, OG) face operational risk if the company fails. Decentralized platforms (Polymarket) face smart contract risk and oracle dispute risk. Both models have failure modes, and users should size positions accordingly.

Liquidity risk

Markets outside the top tier can have wide spreads and thin order books. Exiting a large position in an illiquid market can be expensive or impossible. New users consistently underestimate this and end up trapped in positions they cannot easily exit.

Information asymmetry

Prediction markets attract sophisticated traders with better information than casual users. Pricing often reflects this information advantage. Users who trade based on gut feel or mainstream media narratives typically lose to traders with deeper research or real-time data access.

Wash trading and manipulation

Polymarket’s anonymous structure allows wash trading (traders buying and selling with themselves to create apparent volume). A Columbia University study published in November 2025 estimated that about a quarter of Polymarket’s historical volume showed signs of it. Kalshi’s KYC requirements make wash trading harder but not impossible.

Oracle disputes

Polymarket’s UMA oracle has resolved the vast majority of markets without controversy but has produced high-profile disputes on ambiguous outcomes. A dispute can delay payout or flip the resolution in ways that are not always intuitive. Understand the resolution mechanism before trading significant positions.

Common mistakes

Overtrading illiquid markets. New users often trade niche markets where their information edge does not compensate for the wide spreads and execution risk. Stick to top-volume markets until you understand the liquidity dynamics.

Confusing conviction with edge. Being confident about an outcome does not mean the market is mispriced. Your job is to find gaps between market probability and true probability, not to prove you are right.

Ignoring fees on short-duration trades. Dynamic fee structures mean frequent trading on low-margin markets can eliminate your edge entirely. Model the break-even before high-frequency strategies.

Not understanding resolution criteria. Read the full resolution rules before trading. Edge cases (Was the outcome announced by the deadline? Does the data source count?) decide close markets.

Using prediction markets as sports betting without adjusting mental model. Prediction markets reward probability-first thinking. Users who approach them like sportsbooks miss the different liquidity dynamics and fee structures.

Why Prediction Markets Are Growing

The 1,100% year-on-year growth on Kalshi is not an accident. Several structural factors are driving adoption simultaneously.

Forecast accuracy. Prediction markets consistently outperform expert predictions and polling averages on verifiable outcomes. The 2024 election was the most public demonstration of this, but the pattern holds across politics, economics, and sports. For institutions and serious analysts, this creates real value beyond speculation.

Regulatory clarity in the US. The CFTC designation of Kalshi as a DCM and Polymarket’s QCX acquisition gave both platforms federal legitimacy that earlier prediction markets lacked. This opened the door to institutional participation, mainstream media coverage, and partnership deals (Kalshi with CNN and CNBC, ICE’s $2 billion investment in Polymarket, FIFA’s World Cup deal with ADI Predictstreet).

Sports contracts unlock volume. The legal permissibility of sports event contracts under federal CFTC authority (contested by states but currently operating) turned prediction markets into a competitor for the $100B+ US sports betting market. This is what drove 87% of Kalshi’s March 2026 volume.

Political engagement. Polymarket became a cultural phenomenon during the 2024 election and the Trump administration’s second term. Political markets create passionate user communities that drive engagement beyond pure financial speculation.

Crypto infrastructure. Polymarket’s Polygon-based architecture allowed it to scale globally in a way that traditional financial infrastructure could not match. USDC settlement, smart contract enforcement, and permissionless participation created a product that works in 160+ countries simultaneously.

The 2026-2027 Outlook

Three trends will shape prediction markets over the next 18 months.

The Supreme Court question. The split is now real. The Third Circuit sided with Kalshi in April 2026, the Ninth Circuit sided with Nevada in August and the Sixth Circuit with Ohio and Tennessee in September. New Jersey (No. 26-299) and Robinhood (No. 26-338) have petitioned the Supreme Court, and Kalshi’s response in the New Jersey case is due on November 9. If the Court takes it, a ruling would most likely land in 2027 and decide whether sports event contracts are federal-only or subject to state law.

Consolidation and institutional entry. Bloomberg has reported Kalshi raising at $40 billion and Polymarket at $21 billion, neither confirmed by the companies at the time of writing. ICE has put $2 billion into Polymarket, CME runs FanDuel’s prediction product, Robinhood built its own exchange with Susquehanna, and DraftKings moved onto its own exchange in June 2026.

Regulatory response internationally. The UK Gambling Commission treats prediction markets as betting. Nine European regulators pledged coordinated action in June 2026, and Brazil, South Korea, India and others blocked the big platforms this year. Canada has opened a securities route through CIRO, and more countries are likely to pick a side during 2026-2027.

Who Should Trade Prediction Markets

Prediction markets work for:

• Traders with probabilistic thinking who understand market microstructure
• Users with genuine information advantages in specific categories
• Speculators looking for a more transparent alternative to traditional sports betting
• Researchers and analysts using market prices as forecast signals
• Crypto-native users comfortable with wallet mechanics (for Polymarket International)

Prediction markets do not work well for:

• Users seeking guaranteed returns or low-volatility investment
• Participants who cannot distinguish between conviction and edge
• Users in jurisdictions with unclear or hostile regulatory treatment
• Casual bettors who prefer fixed-odds simplicity over exchange mechanics
• Anyone who cannot afford to lose the capital they deploy

Related Guides

• What Are Prediction Markets (Definition)
• Best Prediction Markets Platforms
• Polymarket Review
• Kalshi Review
• Prediction Markets Regulation
• Are Prediction Markets Legal
• Prediction Markets vs Betting

Final Verdict

Prediction markets in 2026 are a legitimate asset class with tens of billions of dollars in monthly volume, more than $30 billion in last confirmed valuations for the two leaders alone, and federal regulation in the US under the CFTC, even as states fight over sports contracts. They offer real advantages over traditional betting (transparent pricing, exchange-based mechanics, genuine information aggregation) and create forecast accuracy that outperforms most experts on verifiable outcomes.

They also carry real risks. The legal picture is unsettled in the US, restrictive in most of the EU, and evolving in the UK. Liquidity is highly concentrated in top markets, punishing users who spread thin across niche contracts. Information asymmetries favor sophisticated traders. Oracle disputes can flip positions unexpectedly.

Approach prediction markets as a new type of financial market rather than a better sportsbook, size positions conservatively until you understand the mechanics, and focus on categories where you have genuine edge. The category will be materially bigger and more regulated in 18 months than it is today. The question is whether you will have learned enough about how it works to participate effectively when it matures.

FAQ

Are prediction markets gambling?

It depends on the jurisdiction. In the US, the CFTC treats them as event contracts (derivatives), while multiple states argue they constitute gambling. In the UK, the Gambling Commission classifies them as betting intermediaries (gambling). In the EU, treatment varies by member state. The legal classification determines which regulator has authority and what licensing applies.

How do prediction markets make money?

Platforms earn revenue mainly through trading fees, which on both Kalshi and Polymarket are taker fees that depend on price and peak around 50 cents, plus bid-ask spreads on certain market structures and interest on customer cash (Kalshi passes a variable 3.50% APY to US users). Some platforms also monetize data and API access for institutional users.

How accurate are prediction markets?

Research consistently shows prediction markets outperform expert forecasts and polling averages on verifiable outcomes. The 2024 US election is the most prominent recent example. Accuracy is strongest in high-volume markets with diverse participants and weakest in thin markets dominated by small groups. Markets with clear resolution criteria and deep liquidity typically produce probability estimates within 2-3% of the ultimate frequency of the events.

Can you make money on prediction markets?

Yes, but consistent profits require real skill. The average casual user loses money due to fees, spreads, and information disadvantages. Profitable traders typically have domain expertise (sports analytics, political data, economic modeling), disciplined risk management, and patience. Market making can generate consistent returns for users with capital and infrastructure but requires managing inventory risk.

What is the biggest prediction market?

Kalshi, by a wide margin. It traded about $37 billion in August 2026 on The Block’s count, against roughly $8.2 billion across both Polymarket venues. Kalshi’s last confirmed valuation was about $22 billion in March 2026, with a $40 billion round reported in September. Polymarket was valued at about $9 billion when ICE invested in 2025, and a raise at $21 billion was reported in August 2026.

Are prediction markets legal in the US?

Kalshi and Polymarket US are CFTC-regulated and operate legally at the federal level. Sports event contracts are another matter. Kalshi has geofenced them in Nevada, Washington and Michigan, around a dozen other states are in court, the CFTC is suing several states to block their enforcement, and the federal appeals courts have split. New Jersey has asked the Supreme Court to settle it. See our are prediction markets legal guide for detail.

Are prediction markets legal in the UK?

Only operators licensed by the UK Gambling Commission can serve UK users, and the UKGC treats prediction markets as betting intermediaries, like betting exchanges. In practice that means easyBet Predictions on Matchbook’s exchange, live since January 2026, and Betfair Predicts, in invite-only beta since April 2026. Polymarket is close-only for UK users and Kalshi restricts the UK.

What is the difference between Kalshi and Polymarket?

Kalshi is CFTC-regulated, US-based, primarily fiat-funded, and most of its volume is sports. Polymarket International is crypto-native (USDC on Polygon), available in most countries outside the 38 it blocks or sets to close-only, and skews toward politics and crypto markets. Polymarket US relaunched in December 2025 as a separate CFTC-regulated product for US users. For a detailed comparison, see our Polymarket review and Kalshi review.

How do prediction markets resolve outcomes?

Each platform uses an oracle mechanism. Kalshi uses internal settlement tied to official data sources and verified news. Polymarket uses UMA’s Optimistic Oracle, where an approved proposer posts a bond, typically $750 USDC, behind an outcome. If undisputed within 2 hours, the proposal becomes final. Disputed outcomes go to UMA token-holder voting, which usually takes four to six days. The vast majority of markets resolve without dispute.

Can I short a prediction market?

Yes. Buying “No” shares is functionally equivalent to shorting “Yes.” You can also sell “Yes” positions you previously bought to exit before resolution. Crypto.com has announced margin trading for OG, which would allow leveraged positions, but it was not live as of October 2026.

What fees do prediction markets charge?

Fees vary by platform. Kalshi charges takers 0.07 × contracts × price × (1 − price), which peaks at $1.75 per 100 contracts at 50 cents and works out to anywhere from under 1% to nearly 7% of the stake depending on price. Polymarket International peaks at $1.75 per 100 shares on crypto, $1.25 on sports and $1.00 on politics, with geopolitics free, and Polymarket US has its own schedule. Robinhood charges a commission of up to a cent per contract plus exchange fees. Resting limit orders are free on most markets.

Can prediction markets predict the future?

Prediction markets aggregate information from participants into probability estimates, which on average are more accurate than individual expert forecasts or polls on verifiable outcomes. They do not predict the future with certainty. They produce a probability distribution reflecting collective knowledge. Users who treat prediction market prices as probabilities (not certainties) extract the most value from the information.