Prediction markets processed $28.4 billion in combined monthly volume in May 2026, up from under $5 billion in September 2025. Lifetime trading volume across Kalshi and Polymarket crossed $150 billion in April. A Bernstein analyst projects 2026 full-year volume at $240 billion — a 370% year-over-...
"Nothing is more valuable than the truth." — Shayne Coplan, CEO, Polymarket
Prediction markets processed $28.4 billion in combined monthly volume in May 2026, up from under $5 billion in September 2025. Lifetime trading volume across Kalshi and Polymarket crossed $150 billion in April. A Bernstein analyst projects 2026 full-year volume at $240 billion — a 370% year-over-year increase.
The capital structure behind this growth is substantial. Kalshi closed a $1 billion Series F in May at a $22 billion valuation. Intercontinental Exchange, the NYSE parent, has committed approximately $2 billion to Polymarket at a $9 billion valuation. DraftKings filed its first CFTC event contract templates on May 22. Jupiter launched Forecast, a Solana-native prediction market, on June 4. On June 2, Polymarket executed its first institutional block trade — a six-figure GPU compute hedge settled on Polygon.
The growth trajectory raises questions about where genuine economic value resides. Prediction markets charge 0% to 0.75% in fees. Kalshi's annualized revenue exceeds $1.5 billion on $178 billion in annualized volume. But a Pew Research Center analysis found that sports account for 80% of Kalshi's volume and 39% of Polymarket's — categories that overlap directly with regulated sportsbooks. A congressional probe opened in May over insider trading on policy contracts. Ten countries have banned or restricted the platforms in 2026 alone. This report examines the economic structure, competitive dynamics, and sustainability of a sector that has outgrown its infrastructure.
Combined monthly global trading volume on Kalshi and Polymarket rose from under $5 billion in September 2025 to $24 billion in April 2026 and $28.4 billion in May, according to Pew Research Center data published May 27. Kalshi accounted for $17.3 billion of the May total (approximately 61%), with Polymarket handling $8.4 billion.
The growth curve is steep. According to TRM Labs, monthly volume grew from $1.2 billion in early 2025 to over $20 billion by January 2026, with unique wallets tripling to 840,000 in the six months through February 2026. Geopolitics, macroeconomics, and politics — not cryptocurrency — now drive the majority of activity.
Pew Research Center's platform-level breakdown is instructive. Sports account for 80% of Kalshi's total historical volume since July 2024 and 39% of Polymarket's. Combined with politics and cryptocurrency, these three categories represent 91% of Kalshi's volume and 90% of Polymarket's. The concentration matters: it means the sector's headline growth figures are substantially driven by sports betting under a different regulatory label.
Lifetime volume across both platforms crossed $150 billion in April 2026. Kalshi posted $14.81 billion in monthly notional volume that month, with Polymarket clearing $9.01 billion. This combined monthly figure now exceeds what legal U.S. sportsbooks process in a typical month, according to Pew.
Kalshi's annualized revenue exceeds $1.5 billion, per the company's May disclosure alongside its $1 billion fundraise. Annualized trading volume has more than tripled in six months, from $52 billion to $178 billion. Kalshi charges taker fees that peak at approximately 1.75 cents per contract at the 50-cent price point (50% implied probability).
Polymarket's fee structure differs. In March 2026, the platform introduced a probability-based taker fee for sports markets using the formula: Taker Fee = 0.0625 × Price × (1 − Price). The peak effective fee is 0.75% at the 50/50 price point. Makers pay 0% and receive daily USDC rebates from collected taker fees. Geopolitical and world events markets remain entirely fee-free.
The subsidy dynamics are notable. Polymarket's zero-fee geopolitical markets — arguably the platform's most high-profile category — generate no direct revenue. The platform's data licensing arrangement with ICE, which launched the Polymarket Signals and Sentiment tool in February 2026, represents an alternative monetization channel: selling normalized prediction data as structured market signals for institutional traders.
Kalshi's revenue model is more conventional — it charges both sides of trades and earns from volume. But the company's $22 billion valuation implies a price-to-revenue multiple above 14x, suggesting investors are pricing future growth, not current cash flow.
DraftKings provides a useful comparison point. CEO Jason Robins disclosed on Q1 2026 earnings that annualized predictions consumer volume exceeded $1 billion in April, with total annualized trading volume above $2.3 billion. DraftKings' DKeX subsidiary filed its first CFTC event contracts on May 22, covering "GAMEPROPERTY" and "GAMEWIN" contract classes. The company's established sportsbook user base — approximately 24 million — represents a distribution advantage that pure prediction market platforms lack.
The prediction market sector has consolidated around two dominant platforms while attracting entrants from adjacent industries.
Kalshi ($22B valuation, $1B raised in May from Coatue, Sequoia, a16z, Paradigm, Morgan Stanley, ARK Invest) has pursued institutional adoption. Institutional trading volume on the platform grew 800% over six months. Clear Street and Interactive Brokers both announced integration of Kalshi contracts onto their platforms in May. The company waived fees on block trades above 100,000 contracts through September 1. In February, Kalshi partnered with Solidus Labs for surveillance and enforcement infrastructure. In March, it formalized a partnership with ARK Invest.
Polymarket ($9B valuation, $2B from ICE) has leveraged its on-chain architecture for differentiation. On June 2, the platform closed its first institutional block trade — a six-figure transaction between FalconX and AneraLabs settling against the Ornn Compute Price Index tracking Nvidia H100 GPU rental pricing, recorded on Polygon. FalconX will serve as a dedicated market maker for future block trades. The platform also announced plans for a POLY token, with a launch expected in Q2 2026.
Jupiter Forecast, launched June 4 on Solana, represents the first major on-chain-native competitor. Unlike Polymarket (which runs on Polygon), Forecast uses competing market makers — proprietary automated market makers (Prop AMMs) post independent quotes and the system routes users to the most competitive price. The platform starts with 15-minute crypto price prediction markets. Total volume through the Jupiter Prediction tab reached approximately $17 million by April 2026. Jupiter emphasized that Forecast "complements rather than competes" with Polymarket and will maintain integration of Polymarket markets.
The rivalry between Kalshi and Polymarket has turned hostile. On June 4, Polymarket accused Kalshi of corporate espionage and copying platform features. Despite this, both CEOs co-invested in a $35 million prediction markets venture fund in March 2026, signaling shared interest in category growth.
The institutional thesis for prediction markets rests on two premises: hedging and data.
Polymarket's June 2 block trade illustrates the hedging application. By settling a GPU compute price contract on-chain, the trade demonstrated that prediction markets could function as commodity futures for AI infrastructure costs — a market with no existing derivatives product. The counterparties were FalconX (a crypto prime brokerage) and AneraLabs (an AI risk clearinghouse), not traditional commodity traders.
ICE's data play is more significant by scale. The February 2026 launch of Polymarket Signals and Sentiment delivers crowd-sourced probability assessments as structured data feeds for institutional and professional traders. FinTech Weekly described the $2 billion commitment as being about "financial data infrastructure, not prediction markets." This positions Polymarket as a data vendor rather than an exchange — a model that, if validated, would be a substantial pivot from its retail origins.
On the Kalshi side, institutional adoption is occurring through traditional distribution channels. Interactive Brokers and Clear Street — firms serving institutional and high-net-worth clients — integrated Kalshi contracts in May. Wintermute, one of the largest crypto market makers, began quoting prediction markets as event-contract volume topped $60 billion in 2026, per The Defiant.
The institutional penetration remains early. ARK Invest sized the medium-term opportunity at $1 trillion to $5 trillion over a 3-5 year horizon. Current annualized volumes ($240 billion projected for 2026, per Bernstein) represent less than 5% of the low end of that estimate.
The regulatory landscape is multi-layered and adversarial.
CFTC: On March 12, the CFTC's Division of Market Oversight issued Advisory Letter 26-08 detailing compliance requirements for event contracts, alongside an Advance Notice of Proposed Rulemaking. DCMs certified approximately 1,600 event contracts for listing in 2025 alone. The comment period closed April 30. As of late May, the White House Office of Management and Budget was reviewing the CFTC's proposed framework. The current CFTC chair withdrew a prior rule proposal that would have banned political and sports event contracts.
SEC: In May, the SEC delayed 24 prediction market ETF launches from Bitwise, Roundhill, and GraniteShares, requesting additional time to review pricing models, risk disclosures, and event contract mechanics. SEC Chairman Paul Atkins testified in February that prediction markets represent "overlapping jurisdiction" between the CFTC and SEC — a turf issue that adds regulatory uncertainty.
Congress: House Oversight Committee Chairman James Comer opened a probe on May 22 into insider trading on Kalshi and Polymarket, seeking records on KYC controls, suspicious activity, and contracts tied to military and geopolitical events. Separately, the Senate advanced restrictions barring members from trading on prediction markets. On June 4, U.S. lawmakers urged the FTC to investigate both platforms' business practices.
International bans: More than 10 countries have banned or restricted Polymarket or Kalshi, with seven acting in 2026 alone, according to CCN.
The platforms have responded. Both rolled out enhanced integrity measures in March 2026, restricting politicians from trading on their own campaigns, athletes on their own leagues, and employees on contracts tied to their employers. Kalshi partnered with Solidus Labs for market surveillance. But the structural tension remains: products that overlap with both regulated futures and regulated gambling face dual or triple regulatory exposure.
A peer-reviewed study — the most comprehensive analysis of live prediction market data to date — found that accuracy on these platforms comes from approximately 3% of traders, not from broad crowd wisdom. This small group's trades predict outcomes, move prices toward resolution values, and react to information first, according to CoinDesk's April 26 report on the findings.
A Vanderbilt University study found significant divergence in accuracy by platform. PredictIt led with 93% contract accuracy, while Polymarket registered 67%. The constrained, lower-volume environment of PredictIt appeared to produce more reliable forecasts.
Market integrity remains a concern. TRM Labs analysts observed "behaviors that resemble forms of market manipulation, including coordinated wallets entering positions ahead of major news and thin markets where a single participant dominates pricing." Kalshi publicly claimed 70% wash trading in Polymarket's top markets, per MSN. A New York Times investigation found more than 80 Polymarket users made bets with suspicious characteristics, including trades placed hours before U.S. and Israeli strikes in Iran. A U.S. soldier was arrested for using inside information to place bets on Polymarket about the ouster of Venezuela's Nicolás Maduro, reportedly netting roughly $400,000.
Calibration data tells a more nuanced story. Across platforms, the average calibration error runs 2-3%, with events priced at 70% occurring approximately 68-72% of the time. Brier scores often fall below 0.10, and in the final days before contract expiration, approach 0.00-0.01. The mechanism works — but its reliability depends on which market, which timeframe, and which participants.
Volume growth is real but concentrated. May 2026 saw $28.4 billion in combined monthly volume, but sports betting accounts for 80% of Kalshi's trading and 39% of Polymarket's. The sector's growth story is substantially a sports betting reclassification story.
Valuations imply aggressive future growth. Kalshi's $22 billion valuation on $1.5 billion in annualized revenue (14.7x) and Polymarket's $9 billion valuation on undisclosed but likely lower revenue require sustained volume expansion and fee monetization to justify current pricing.
Institutional adoption has begun but remains minimal. Polymarket's first block trade and ICE's data feed product are proof-of-concept, not scale. Interactive Brokers and Clear Street integrations on Kalshi represent distribution rather than institutional demand.
Regulatory risk is multi-vector. SEC ETF delays, CFTC rulemaking, congressional insider trading probes, and international bans create compounding jurisdiction conflicts. The absence of a finalized federal framework leaves the sector exposed.
Accuracy is driven by a 3% trader minority, not crowd wisdom. The information discovery function — the sector's core value proposition — depends on a thin layer of informed participants, not democratic aggregation.
New entrants are fragmenting the market. Jupiter Forecast on Solana, DraftKings via DKeX, and potential ETF products from Bitwise and others expand access but also dilute network effects for incumbents.
Prediction markets are the fastest-growing segment in financial services by volume growth rate, expanding from $1.2 billion in monthly volume in early 2025 to $28.4 billion by May 2026. The capital pouring in — $1 billion to Kalshi, $2 billion to Polymarket from ICE, $35 million to a shared VC fund — reflects institutional conviction that this category will persist.
The economic sustainability question mirrors the broader pattern identified in blockchain ecosystem analysis: how much of this activity represents genuine value creation versus subsidized growth? Polymarket's zero-fee geopolitical markets, Kalshi's waived block trade fees, and the sector's heavy reliance on sports categories (which compete with established sportsbooks offering similar products) suggest the revenue model is still being found, not proven.
The sector's most defensible value proposition may not be the trading venue itself but the data it generates. ICE's $2 billion investment in Polymarket is explicitly positioned as a data infrastructure play — structured probability signals for institutional decision-making. If prediction market data becomes standard input for financial analysis (the way Bloomberg Terminal data is today), that justifies the current valuations. If the sector remains primarily a retail sports and political betting venue with regulatory gaps, it does not.
At $240 billion in projected 2026 volume, prediction markets have achieved scale. Whether they have achieved a sustainable business model is a separate — and still open — question.