One-line version: prediction markets crossed the threshold from "election curiosity" to "the second financial-markets layer" in the past six months. Bernstein's $1T-by-2030 projection just became plausible because the structural pattern is now identical to how AI consolidated: top-layer duopoly + fragmented distribution + dedicated infrastructure. Middle layer dies. Sports is the entry point. Macro is the endgame. The CFTC is the most powerful financial-market regulator no one is watching. The next Supreme Court case in this space is already inevitable.

This essay is a material reading of one specific question: with everything in the public record as of late April 2026, where are prediction markets actually, who are the main players, what are the moves that just happened or are about to happen, and where does the alpha sit?

The numbers below — every figure, valuation, contract value, court ruling, share, and headcount — are sourced. See footnotes.

The shape: from $51B all of 2025 to $66B in four months of 2026

Set these facts side by side:

Combined Kalshi + Polymarket year-to-date trading volume through April 20, 2026: roughly $66.7B (Kalshi $37.49B + Polymarket $29.23B). That number is already greater than the entire 2025 industry total ($51B in Polymarket alone, ~$60B industry-wide).[1]
The week of April 6–11, 2026 set the all-time record: Kalshi $3.54B + Polymarket $2.48B = $6.02B combined. Driven by The Masters, U.S. election-cycle markets, and Iran geopolitical contracts. The following week dropped 13–17% on a lighter sports calendar but still cleared $5.10B combined.[1]
Super Bowl Sunday 2026 (February): Kalshi did $871M in single-day volume, with combined NFL Super Bowl–related markets clearing over $1B in 24 hours on Kalshi alone, and $1.38B across Kalshi + Polymarket combined.[2]
Bernstein analyst Gautam Chhugani projects $240B in industry trading volume for 2026 (a 370% YoY increase) and ~$1T by 2030 at a roughly 80% CAGR. Industry revenue projected at $400M (2025) → $2.5B (2026) → $10.8B (2030) at current take rates.[3]
Kalshi raised $1B in March 2026 at a $22B valuation (Coatue-led; up from $11B in December 2025; up from $5B in October 2025 — three valuation steps in six months).[4]
Polymarket has taken $2.6B from Intercontinental Exchange (NYSE parent): $2B at a $9B valuation in October 2025, plus an additional $600M in March 2026, plus a reported $400M raise in negotiation at a $15B valuation.[5]
CFTC approved Polymarket's full U.S. relaunch on November 25, 2025, ending a near-three-year hiatus that began with a 2022 settlement.[6]
Robinhood's prediction-markets line is now its fastest-growing revenue segment, generating roughly $350M in annual recurring revenue and accounting for ~30% of Kalshi's total volume. Over 9 billion contracts have been traded by 1M+ Robinhood customers in twelve months.[7]
Robinhood + Susquehanna acquired 90% of MIAXdx (a CFTC-licensed DCM and DCO) on January 20, 2026, rebranding it Rothera. Q2 2026 launch.[8]
DraftKings entered prediction markets on December 19, 2025 via the acquisition of Railbird (CFTC-licensed). Live in 38 states. Fanatics launched Fanatics Markets in early December 2025 via the acquisition of Paragon Global Markets, with Crypto.com supplying backend infrastructure. FanDuel entered via its own CFTC acquisition in late 2025.[9][10]
Coinbase rolled out prediction markets to all U.S. customers on January 27, 2026. Coinbase had previously closed the $2.9B Deribit acquisition in August 2025, building the largest dedicated derivatives moat in U.S. crypto.[11]
Both Kalshi and Polymarket announced crypto perpetual futures launches in April 2026. Polymarket launched its perpetuals product on April 21; Kalshi's "Timeless" product launched April 27. Both go directly at Coinbase's derivatives business.[12]
Jump Trading took equity stakes in both Kalshi and Polymarket in February 2026 in exchange for committed liquidity provision. Susquehanna disclosed market-making at Kalshi in 2024; Jane Street is publicly active in the space. The "Big Three" prop trading firms (SIG, Jane Street, Jump) are all building dedicated prediction-market trading desks.[13]
5c(c) Capital announced a $35M infrastructure-only fund on March 23, 2026. Backers: Kalshi CEO Tarek Mansour, Polymarket CEO Shayne Coplan, Marc Andreessen (a16z via Moneta Luna), Micky Malka (Ribbit), and Kyle Samani (formerly Multicoin). Investment thesis: 20 early-stage startups over two years, infrastructure only — explicitly not exchanges.[14]
AI agents now represent roughly 30% of wallet activity on Polymarket, with reported profitability around 37% — versus 7–13% for human traders.[15]
CFTC Chair Mike Selig was confirmed by the Senate on December 18, 2025 (53-43). His predecessor, Acting Chair Caroline Pham, served from January 20, 2025 through December 22, 2025 and laid the regulatory foundation for the current expansion. Selig has explicitly directed staff to draft new event-contract rules, withdrew the prior staff advisory restricting sports event contracts, and has filed CFTC suits against the states of Arizona, Illinois, and Connecticut asserting "exclusive regulatory authority" over commodity derivatives.[16]
Wisconsin became the latest state to sue major prediction-market operators on April 23, 2026 — joining 19+ existing federal cases. Defendants in Wisconsin's complaint: Kalshi, Coinbase, Polymarket, Robinhood, Crypto.com. Mixed court results so far: Kalshi has won preliminary injunctions in Nevada (April 2025) and at the federal appeals level against New Jersey, but lost a preliminary motion against Maryland in 2025 — the first time a state regulator successfully convinced a federal judge that federal preemption likely doesn't apply.[17]
DraftKings, FanDuel, Fanatics, and Bet365 jointly funded a $48M Super PAC ("Win for America") in early April 2026 targeting Texas, Georgia, and 15 other states. The political driver is the prediction-market threat — sportsbooks are spending hard money to keep prediction-market platforms out of state-by-state sports betting.[18]
Kalshi suspended and fined three congressional candidates on April 22, 2026 for "political insider trading" — Mark Moran (VA Senate, Independent after Democratic primary), Matt Klein (MN-02 Democrat), Ezekiel Enriquez (TX-21 Republican). Fines ranged from $539 to $6,229; five-year suspensions.[19]

Each fact, taken alone, is a discrete event. Stacked together, they describe a single structural transition: prediction markets are crossing the threshold from "asset class A people occasionally talk about during election years" to "asset class B that has its own bond-grade infrastructure, regulated venues across distribution channels, dedicated institutional desks, and a Supreme Court case loading."

The single observation underneath all of it: the industry is consolidating into the same shape AI is consolidating into.

Top layer = winner-take-most duopoly. Kalshi at $22B + Polymarket at $15B together hold over 95% of U.S. retail prediction-market volume. Everything else is plumbing.
Distribution layer = fragmented, becoming the actual battleground. Robinhood, DraftKings, FanDuel, Fanatics, Coinbase, Crypto.com, Webull, Kraken, ForecastEx (via Interactive Brokers), and the upcoming Rothera. Each one is a different distribution surface for the same underlying contracts. Margins compress. Volumes compound.
Infrastructure layer = small specialists, capital-thin. 5c(c) Capital, Paradigm (institutional terminal), Adjacent News (data API), UMA (oracle for Polymarket), the in-house liquidity desks at SIG/Jane Street/Jump. Three orders of magnitude smaller than the venue or distribution layers. Spray-and-pray VC economics.
No middle-layer winner. "General prediction-market tooling without a venue or a hyperscaler partnership" is structurally a losing position. The five-year capital allocated to that middle layer will mostly evaporate, exactly as the corresponding middle layer in AI is evaporating.

The rest of this essay walks through each layer.

A short history: from papal elections to Polymarket

Prediction markets are old. The new thing is the regulated, scaled, financialized version of them.

Sixteenth-century Vatican bookmakers ran odds on papal elections. Eighteenth-century London coffeehouses ran prediction markets on military campaigns and life-table actuarial questions. Nineteenth-century New York had organized political-betting venues at every major election, with public posting of odds. Through the early twentieth century, U.S. presidential elections had documented betting markets larger by volume than the Stock Exchange's daily turnover during election weeks. Federal anti-gambling legislation in the 1930s suppressed visible betting; offshore and informal markets persisted.

The modern wave begins with two specific predecessors that everyone in the current ecosystem has internalized as cautionary tales:

Intrade operated from 2001 to 2013 — a Dublin-based platform with a US-active user base. Famous for correctly predicting the 2008 Obama victory and 49 of 50 states in the 2012 Obama re-election. The CFTC sued Intrade in November 2012 for operating an improper off-site options exchange. Intrade banned U.S. users in response. By March 11, 2013, the platform halted all trading citing potential "financial irregularities" — its founder had died in late 2011 and post-mortem audits revealed accounting issues. Total wagered notional over the platform's lifetime: roughly $200M+. Total active accounts: ~50M+. Intrade's failure was foundational. It taught the industry two lessons: (1) you cannot operate an unregulated event-contract venue with U.S. customers — the CFTC will eventually move; (2) the founder-key-person risk in this category is severe because trust in resolution depends on operational continuity.[20]

PredictIt launched in 2014 as an academic research project run by Victoria University of Wellington, operating in the U.S. under a CFTC no-action letter that explicitly limited it to academic / research use. PredictIt's design constraints — $850 maximum per single contract, 5,000-person cap per market — kept it small but durable. In 2022, the Biden-era CFTC revoked the no-action letter. PredictIt litigated. In July 2025, the U.S. District Court for the Western District of Texas entered final judgment in favor of PredictIt, declaring the CFTC's prior efforts invalid and barring future shutdown attempts. In September 2025, Aristotle, Inc. received CFTC approval for full DCM and DCO designation under the umbrella of the Prediction Market Research Consortium — moving PredictIt from academic-research-with-caps to a real exchange. The contract cap was raised from $850 to $3,500. The 5,000-person cap was eliminated. Underdog Fantasy subsequently acquired Aristotle Exchange to bring prediction markets in-house.[21]

Augur launched in 2018 as the first decentralized prediction market on Ethereum. v1 was effectively unusable due to gas costs and unresolvable edge cases. v2 ("Turbo") improved UX but never gained meaningful traction. Augur is now rebooting in 2025-2026 with new oracle technology — a legitimate technical direction but starting from an effectively zero user base. Augur's value to the current industry is mostly retrospective: it proved that the decentralized-resolution problem was harder than the design papers had anticipated, which is why every modern decentralized prediction market either uses an oracle network (UMA for Polymarket) or accepts a centralized resolution mechanism.

The current generation begins with Kalshi (founded 2018, CFTC-approved 2020, launched 2021) and Polymarket (founded 2020, originally on Polygon). Both founders explicitly cited Intrade as the model and the cautionary tale. Both designed their platforms specifically to avoid Intrade's regulatory and key-person risks: Kalshi by going through the full CFTC DCM approval process; Polymarket by operating offshore initially and using a decentralized oracle for resolution.

The 2024 inflection point was the U.S. presidential election. Polymarket processed $3.6B+ on the election outcome alone — the largest single resolved-event prediction-market position in industry history. Kalshi won a federal court ruling in late 2024 that election-based event contracts do not constitute illegal gambling under the Commodity Exchange Act — the foundational legal precedent that opened the entire U.S. event-contract market.[22]

The 2026 inflection point is what we are inside now. The volume curve is steepening on a roughly weekly basis. The regulatory architecture is consolidating around CFTC exclusive jurisdiction. The capital flows are institutional-scale. The asset class is being renamed in the financial press as "Information Finance." All of this happened in roughly twelve months.

The Top Layer (1): Kalshi

Tarek Mansour and Luana Lopes Lara founded Kalshi in 2018 out of MIT. Mansour graduated in 2018 with bachelor's degrees in computer science and mathematics plus a master's in engineering. He spent his early career as a quantitative trader at Goldman Sachs and a global macro trader at Citadel before concluding that "many trades were driven by expectations about future events even though there was no streamlined market for such trading." Kalshi secured CFTC approval in 2020, launched in 2021, and won a federal court ruling in late 2024 that election-based event contracts do not constitute illegal gambling under the Commodity Exchange Act — the foundational legal precedent that opened the U.S. event-contract market.[23]

Mansour's stated long-term vision: "to financialize everything and create a tradeable asset out of any difference in opinion."[23] That is the literal extension of the financial-instrument creation pattern that produced equities (claims on future cash flows), bonds (claims on future repayment), futures (claims on future delivery), options (claims on future contingent payoffs), credit default swaps (claims on future default events). Event contracts are the natural next category — claims on the realized truth-value of any specifiable future proposition. A category that, if Mansour is right, has no inherent ceiling beyond the population of specifiable propositions about the future.

The Kalshi numbers as of April 2026:

Valuation: $22B (March 2026, Coatue-led $1B raise; up 4.4× from $5B in October 2025)
Annualized revenue: roughly $1.5B
U.S. market share: ~90% of regulated U.S. prediction-market volume
Weekly volume: $3B+ consistently, with a record of $3.54B in the week of April 6–11
YTD volume through April 20: $37.49B
Sports as % of volume: >70%
Single-day record: $871M on Super Bowl Sunday (February 2026)
Single NFL game volume: $10M+ in notional volume routinely
Robinhood integration share: ~30% of Kalshi total volume
Other distribution partners: Webull, Crypto.com, various retail brokerages
Strategic advisor: Donald Trump Jr. (since January 2025), via 1789 Capital connections
Recent product additions: margin trading approval (allows ~10× exposure on $1 of capital), "Timeless" crypto perpetual futures (launched April 27, 2026, directly competitive with Coinbase, Kraken, Robinhood derivatives)
International expansion: announced expansion into Brazil
Notable backers: Sequoia Capital, Coatue Management, Andreessen Horowitz, Henry Kravis (KKR co-founder), Charles Schwab, plus various sovereign-adjacent capital flows[4][24]

Kalshi's fee structure is straightforward and aggressive:

Taker fee formula: 0.07 × contracts × price × (1 - price)
Maker fee: 0% (with daily USDC rebates from collected taker fees)
Implication: 50/50 markets cost the most per contract; extreme-odds markets cost almost nothing
A $1,000 position at even odds runs about $35 in taker fees at the 0.07 coefficient[25]

The structural advantages Kalshi has built:

First-mover regulatory precedent. The 2024 federal court ruling was won by Kalshi. Every subsequent venue that wants to offer election or political event contracts cites Kalshi v. CFTC.
Full CFTC DCM + DCO licensing, as opposed to no-action letters or partial approvals. Kalshi can list contracts unilaterally subject to CFTC review rather than per-contract approval.
The Robinhood distribution monopoly. The KalshiEX integration in Robinhood drove that partnership to $350M in ARR in twelve months. While Robinhood + Susquehanna will operate Rothera as a separate exchange in 2026, the existing KalshiEX integration is durable.
Sports market depth. No competing venue offers the breadth of NFL/NBA/MLB/NHL/college/soccer event contracts. >70% of Kalshi's volume is sports — a dominant share that competing venues would need years to replicate.
Susquehanna market-making relationship. Disclosed in 2024. Means execution-quality order books that pure retail platforms cannot match.
Self-policing on insider trading. Kalshi's April 22, 2026 enforcement against three congressional candidates is structurally significant — it positions Kalshi as a self-regulating exchange, exactly the posture the Senate Banking Committee wants to see before approving expansion. The fines were small ($539-$6,229) but the suspensions are 5 years and the institutional signal is clear.[19]

The structural challenges:

State-level regulatory exposure is the most acute risk. Twenty-plus federal lawsuits, plus the Wisconsin case, plus Arizona criminal charges, plus the inevitable Supreme Court federal-preemption case (more on this below).
Sports betting industry hostility. The DraftKings/FanDuel/Fanatics/Bet365 $48M Super PAC is the visible tip of a much larger industry response, and that hostility translates into state-level political pressure on prediction markets.
The Robinhood-Susquehanna divorce. Rothera is structurally a competing venue. As Rothera ramps in Q2 2026, some portion of Kalshi's 30% Robinhood-channel volume will migrate to Rothera's independent venue.
The Brazilian expansion path. Brazil is a regulatory test case for international growth — the first non-U.S. market for Kalshi. The execution risk is non-trivial.

The Top Layer (2): Polymarket

Shayne Coplan founded Polymarket in 2020 after dropping out of his computer science studies at NYU. In October 2025, Coplan became the world's youngest self-made billionaire on the Bloomberg Billionaires Index, with an estimated net worth of approximately $1.0B according to Forbes.[26]

The Polymarket arc through 2024–2026 is the most dramatic narrative in the prediction-markets industry:

Spring 2024: Polymarket secured ~$70M across two funding rounds, backed by Peter Thiel's Founders Fund, Vitalik Buterin (Ethereum cofounder), and General Catalyst.[27]
November 2024 election cycle: Polymarket processed more than $3.6B in bets on the U.S. presidential election outcome — the largest single resolved-event prediction-market position in industry history.[22]
November 13–14, 2024: The FBI raided Coplan's New York City apartment. Phones and electronics were seized. The investigation was reportedly connected to a DOJ inquiry into whether Polymarket had violated its 2022 CFTC settlement by continuing to serve U.S. customers via VPNs and other workarounds. No criminal charges were filed.[28]
July 2025: The DOJ and CFTC formally closed both investigations (one civil, one criminal) with declination notices — no further action.[29]
October 2025: ICE (NYSE parent) committed up to $2B in strategic investment at a $9B valuation, alongside agreements to distribute Polymarket data and collaborate on tokenization.[5]
November 25, 2025: CFTC granted Polymarket an Amended Order of Designation, enabling regulated U.S. operations through registered brokerages and FCMs.[6]
End of 2025: Polymarket relaunched for U.S. residents after a near-three-year hiatus.
March 2026: ICE invested an additional $600M.
April 2026: Polymarket reportedly raising another $400M at a $15B valuation; launched perpetual futures product on April 21.
POLY token + airdrop confirmed publicly by Polymarket CMO Matthew Modabber on Degenz Live (October 24, 2025): token launch sequenced after U.S. relaunch stabilizes. ICE collaboration includes future tokenization initiatives.[30]

The Polymarket numbers as of April 2026:

Valuation: $9B (October 2025) → $15B (April 2026 reported negotiation)
2025 contract volume: $51B+ (the figure Bernstein cites for Polymarket alone all of 2025)
YTD volume through April 20, 2026: $29.23B
Weekly volume: $2B+ consistently; record $2.48B in the week of April 6–11
Monthly volume October 2025: $2.76B
Iran-related markets: $377.8M+ total trading volume across 20 active Iran markets since the war began (February 28, 2026); the single most-traded Iran market — "US x Iran ceasefire by...?" — has done $280.1M alone[31]
World Cup 2026 markets: $769.9M+ on the FIFA World Cup Winner market alone; 197 active markets total; $770.9M+ across all World Cup markets — and the tournament itself starts June 11, 2026[32]
Active markets by category (as of April 7, 2026): 5,400+ crypto, 3,600+ football, 197+ World Cup
Category mix: roughly 42% politics, 31% crypto, sports + macro the remainder
Notable backers: Founders Fund (Peter Thiel), Vitalik Buterin, General Catalyst, ICE (NYSE parent), 1789 Capital (Donald Trump Jr.'s firm — eight-figure investment), Jump Trading
Treasury Secretary Howard Lutnick: longstanding crypto advocate, Cantor Fitzgerald's prior role as Tether reserve custodian, generally supportive policy environment
Donald Trump Jr.: simultaneously Strategic Advisor at Kalshi (since January 2025) and Polymarket Advisory Board member with 1789 Capital investment — the obvious conflict-of-interest is acknowledged across reporting[33]

Polymarket's fee structure is more nuanced than Kalshi's:

Pre-March 30, 2026: zero fees on most markets (the Polymarket positioning was "free for users; revenue from peripheral integrations and data licensing")
Post-March 30, 2026: taker fee = 0.0625 × Price × (1 - Price), applied to a subset of markets — 15-min and 5-min crypto direction markets, NCAAB markets created after February 18, 2026, Serie A markets created after the same date. Politics, economics, and most other sports remain zero-fee on the offshore product.
For the CFTC-regulated Polymarket US venue: flat 0.10% taker fee, 0% maker fee, $0.001 minimum per trade.
Daily revenue post-fee-update: ~$1M, implying ~$365M annualized revenue trajectory[34]

The structural advantages Polymarket has built:

Crypto-native composability. Polymarket runs on Polygon (formerly) and is moving toward its own chain. Smart-contract-resolved markets enable agent trading at a level no centralized exchange can match — which is why AI agents are 30% of wallet activity.
Outright global volume leader before regulatory clearance. $51B in 2025 happened entirely without U.S. retail access. The U.S. relaunch is the volume-amplifier on top of an already-dominant offshore base.
ICE strategic alignment. ICE owns the NYSE, NYSE Arca, and ICE Futures U.S. Polymarket data flowing into ICE's institutional distribution channels means Bloomberg-Terminal-class distribution into traditional finance.
POLY token as long-tail capital tool. The pending airdrop and token launch convert long-time users into tokenholders, creating a financial incentive structure orthogonal to anything centralized exchanges can replicate.
The geopolitical/macro market depth. Iran ceasefire markets at $377M, Trump approval markets at meaningful volume, election cycle markets with full coverage — Polymarket has been the de facto source of geopolitical-event price discovery for two cycles.
The agent-native trader base. AI agents already 30% of wallet activity, with profit rates 3-5× human traders. Whatever the next wave of LLM-agent finance looks like, Polymarket has a head start.

The structural challenges:

Oracle vulnerability. The March 24–25, 2026 Ukraine mineral deal market was a $7M dispute where a single UMA whale with 5M tokens (across three accounts) cast 25% of the resolution votes, forcing a "Yes" resolution despite Ukraine not having agreed. Polymarket called it "an unprecedented governance attack" but did not refund affected positions. UMA subsequently transitioned the resolution mechanism from Optimistic Oracle V2 to Managed Optimistic Oracle V2, with a 37-address whitelist of authorized proposers (Risk Labs and Polymarket employees). The vulnerability is partly closed, but the fact that a single-actor governance attack succeeded once is a permanent reputational issue for institutional adoption.[35]
Regulatory tail risk. The DOJ/CFTC declination is favorable but state-level lawsuits remain. New York Attorney General Letitia James separately sued Coinbase and Gemini for "illegally" running gambling operations through their prediction-market integrations.[36]
Token-launch governance complexity. The POLY airdrop will create thousands of new governance-token holders with their own opinions about market resolution. The same dynamic that produced the Ukraine dispute will be pricier to manage at scale.
The blockchain dependency. Migrating from Polygon to a Polymarket-native chain is a non-trivial engineering effort with execution risk.

The case studies: how specific markets actually behave

The structural facts about prediction markets are most legible through specific live markets. Five examples that map the asset class.

The 2024 U.S. presidential election

Polymarket processed $3.6B+ on the resolved-question "Will Trump win the 2024 election?" — the largest single resolved-event prediction-market in history. Kalshi ran parallel markets at lower volume. Both platforms had Trump priced consistently above polling-average implied probability throughout October 2024. In late October, Trump's odds on Kalshi were 58%; on Polymarket, similar levels. Both platforms diverged from polls' "toss-up" framing.

The post-election analysis was contested. Some research (UCSD working paper) found that prediction markets "overestimated Trump's win probability with significant variance" while options-implied probabilities provided more accurate estimates. Other reads emphasized that the markets correctly identified a non-toss-up state when polls did not. Either way, the election cemented Polymarket's brand as the venue people watched in real time during a major political event — a positioning that Kalshi has worked aggressively to match in 2026.[22]

Iran ceasefire markets

The most active geopolitical market category in 2026. Twenty active Iran-related markets on Polymarket aggregating to $377.8M+ in trading volume since the U.S.–Iran war began February 28, 2026.[31]

Specific market-level breakdown:

"US x Iran ceasefire by...?" — $280.1M traded since launch
"Iran x Israel/US conflict ends by...?" — $100.6M traded since launch (February 28, 2026)
"US x Iran ceasefire extended by...?" — $52.3M traded as of April 26, 2026
"US x Iran permanent peace deal by...?" — $56.5M traded as of April 27, 2026

President Trump announced an indefinite extension of the U.S.-Iran ceasefire on April 21, 2026, just before the initial two-week truce (agreed April 8, mediated by Pakistan) was set to expire on April 22. Hours later, Iran attacked commercial ships in the Strait of Hormuz, prompting the U.S. to sustain its naval blockade while halting strikes. The market reactions to those events happened within minutes — a real-time price signal that produced more accurate, more timely information than any traditional news source.

This is the empirical case that prediction markets are better than news for tracking fast-evolving geopolitical events. CNN reports the ceasefire announcement; Polymarket reprices on the announcement; CNN reports the Iran ship-seizure response; Polymarket reprices again. The aggregate sequence of price moves is a higher-fidelity record of real-time event probabilities than the corresponding sequence of news headlines.

Trump approval rating

Polymarket runs continuous markets on Trump's approval rating crossing specific thresholds (above 35%, above 40%, above 45%) by specific dates. Kalshi has parallel markets. Volume on these markets has been moderate ($10–50M aggregate) but the prices are watched closely by political-strategy professionals as a counter-check to traditional polling averages.

The 2026 reading: Polymarket-implied probabilities of Trump approval crossing 40% by mid-2026 dropped to single digits as approval fell to AP-NORC's 33% reading in April. The market was in some sense ahead of mainstream political commentary on the depth of the approval decline.

2026 midterm elections

The single largest political-event market category for 2026. Markets opened in mid-2025 and have accumulated meaningful volume. As of late April 2026:

Polymarket Democratic House control probability: 85.5% — driven by widening generic-ballot leads (Democrats +5.8 average as of April 26) and Trump's sub-40% approval
Polymarket Democratic Senate control probability: separate market — lower than House given GOP map advantage, but rising
Polymarket "Democratic sweep both chambers" probability: 50.5%
Polymarket "Republican Senate / Democratic House" probability: 36.5% (the structural status quo given map asymmetry)
Polymarket "GOP sweep both chambers" probability: 12.5% — implies the GOP holds the House — not the base case at current pricing

Kalshi runs similar markets with slightly different pricing. The dispersion between Kalshi and Polymarket on identical political markets is typically 1-3 percentage points, suggesting cross-venue arbitrage is mostly working — i.e., the two venues are price-discovery equivalents.[37]

Super Bowl 2026

Single-event sports-market peak. Kalshi processed $871M on game day; combined Kalshi + Polymarket exceeded $1.38B in Super Bowl-related volume across the day. By comparison, Las Vegas sportsbooks processed roughly $1.5–2B on the Super Bowl. For the first time, regulated U.S. prediction markets approached parity with the entire Las Vegas sports-betting handle on a single event. This is the moment the sportsbook industry began responding seriously to the prediction-market threat — the $48M Win for America Super PAC was funded immediately afterward.[2][18]

World Cup 2026 (June-July)

The next major sports-event volume catalyst. The 2026 FIFA World Cup begins June 11, 2026 (co-hosted by U.S., Mexico, Canada). Polymarket alone has $769.9M+ in volume on the World Cup Winner market since launch in July 2025; 197 active markets total; $770.9M+ across all World Cup markets. Industry expectation is that the tournament drives $2-3B+ in combined Kalshi + Polymarket volume during June-July, plus additional volume routed through DraftKings, FanDuel, Fanatics, Robinhood, and the other distribution surfaces.[32]

The structural significance: this is the first World Cup with U.S.-regulated prediction-market platforms operating at scale. The volume comparison to U.S. sportsbooks during the tournament is the next major data point in the prediction-market-vs-sportsbook structural battle.

The Distribution Layer: where the actual fight is

The Kalshi/Polymarket duopoly captures the venue layer. The distribution layer — where retail users actually reach the contracts — is where 2026's competitive dynamics are being decided.

Robinhood (HOOD)

The reference distribution surface. Q1 2026 earnings (reported April 28, 2026, on the day of "Timeless" Kalshi crypto-perp launch): revenue grew ~22–26% YoY to over $1.14B, with prediction markets noted as the fastest-growing revenue line.[7]

The structural facts:

Prediction markets ARR: ~$350M
Kalshi-routed volume share: 30% of Kalshi's total volume
Twelve-month metrics: 9B+ contracts, 1M+ customers
MIAXdx acquisition completed January 20, 2026: 90% Robinhood + Susquehanna joint venture / 10% MIAX retained equity. Rebranded as Rothera. Q2 2026 product launch.
Rothera structure: independent JV operating as a Designated Contract Market (DCM) and Derivatives Clearing Organization (DCO) with full CFTC approval to list and clear fully collateralized futures, options on futures, and swaps.
Susquehanna role: market-making partner since 2024 disclosure; Rothera makes that relationship explicit at the venue level.[8]

Why Rothera matters: Robinhood has been a distribution surface for Kalshi for the past twelve months. Rothera is Robinhood + Susquehanna becoming their own venue — a third top-layer player on top of Kalshi and Polymarket. The economic logic is straightforward: 30% of Kalshi's volume is too valuable to leave entirely on Kalshi's economics. Operating their own DCM means Robinhood + Susquehanna keep a larger share of the take rate.

Whether Rothera fragments the venue layer or simply specializes in distinct contract categories will determine the next twelve months of competitive dynamics. The most likely shape is that Rothera focuses on macro, crypto, and Robinhood-specific user-acquisition flows, while Kalshi remains dominant in sports and Polymarket in geopolitical/political.

Coinbase (COIN)

Coinbase opened prediction markets to all U.S. customers on January 27, 2026.[38] That move was preceded by the August 2025 closing of the $2.9B Deribit acquisition, which gave Coinbase the largest dedicated derivatives moat in U.S. crypto. Together, Coinbase is positioning itself as the venue where crypto traders also bet on macro and event outcomes.

The strategic question for Coinbase: with Kalshi's "Timeless" and Polymarket's perpetuals both launching in April 2026, both venues are now directly attacking Coinbase's $2.9B Deribit moat. Coinbase had previously offered "perpetual-style" futures contracts with long-dated expirations — not true Binance-format perpetuals, which had been blocked from U.S. retail by regulatory framework. CFTC Chair Mike Selig has now stated the CFTC plans to formally authorize perps in the U.S. The question is not whether perps come to U.S. retail; it's whether Coinbase, Kalshi, or Polymarket captures the dominant share when they do.

DraftKings (DKNG)

Entered prediction markets December 19, 2025 via the acquisition of Railbird (CFTC-licensed). Live in 38 states. The DraftKings Predictions strategy is to leverage the existing DKNG sportsbook user base (~5M+ active monthly users in U.S. legalized sports betting markets) and re-route them to event contracts in jurisdictions where direct sportsbook operations are not legal.[9]

The trick is regulatory arbitrage: in California, Texas, and Georgia — three of the largest U.S. sports markets — direct sports betting is illegal at the state level, but CFTC-regulated event contracts are operational under federal preemption. So DraftKings can offer "essentially sports betting" in those states via the prediction-market wrapper, while a state-licensed DKNG sportsbook cannot exist there.

The structural tension: this is exactly why DraftKings, FanDuel, Fanatics, and Bet365 are jointly funding the $48M Win for America Super PAC targeting Texas, Georgia, and 15 other states. They are effectively betting both sides — operating prediction-market alternatives via federal preemption while simultaneously lobbying for state-by-state sports-betting legalization that would let them operate the more profitable sportsbook product.[18]

FanDuel (Flutter Entertainment)

Followed DraftKings's pattern in late 2025: acquired its own CFTC-registered exchange to launch state-bypassing event contracts. Less public detail than DraftKings's Railbird purchase, but similar structural positioning. FanDuel's user-acquisition advantage runs through Flutter's parent-company resources — Flutter is the world's largest online gambling operator.

Fanatics Markets

Launched in early December 2025 via the acquisition of Paragon Global Markets in July 2025. Crypto.com supplies the backend trading infrastructure under a strategic partnership. Positioning: "the first prediction market at the intersection of sports, finance, and culture."[10][39]

Phase 1 (live as of December 2025): event contracts on sports, finance, economics, and politics in 24 states.

Phase 2 (launching early 2026): expansion to crypto, stocks/IPOs, climate, pop culture, tech/AI, movies, and music event contracts.

The strategic difference from DraftKings: Fanatics is leveraging its existing 100M+ sports-merchandise customer database. The user-acquisition-cost economics differ materially from DraftKings's sportsbook-bridge approach.

Fanatics also launched a "Combos" parlay feature in 2026 — bundling multiple event contracts into a single multi-leg position. This is a structural step toward sportsbook-feature parity, again stretching the line between event contracts and sports betting.

Crypto.com

Operates Crypto.com Derivatives North America, a CFTC-registered DCM. Provides the white-label backend for Fanatics Markets. Caroline Pham's CFTC requested compliance documentation from Crypto.com early in 2025 but did not pursue enforcement. Mike Selig's CFTC has been notably permissive.[16]

Webull

Listed Kalshi prediction markets in a manner roughly parallel to Robinhood's Kalshi integration, though at smaller scale. Webull's value is a U.S.-Asia-Pacific-bridge user base — meaningful international distribution potential as Kalshi expands abroad.

ForecastEx (Interactive Brokers / ICE)

Operates inside Interactive Brokers' platform with ICE backing. Focus: macroeconomic and policy-driven event contracts (CPI prints, Fed rate decisions, GDP releases). Positioning as the "serious trader's" prediction market — institutional macro funds rather than retail. Specific volume figures not publicly disclosed.[40]

Kraken

Added prediction markets to its U.S. crypto offering in 2025. Smaller volume than Coinbase but follows the same crypto-exchange-becomes-event-contract-distribution pattern.

The pattern

The distribution layer has gone from "Robinhood + Webull plus the venues themselves" twelve months ago to at least nine distinct retail surfaces today: Robinhood, Coinbase, DraftKings Predictions, FanDuel Predictions, Fanatics Markets, Crypto.com, Kraken, Webull, ForecastEx, plus the venues' own apps (Kalshi, Polymarket, PredictIt). And Rothera as the upcoming tenth. Each surface acquires customers, routes them to underlying venues (or to its own venue), and shares some portion of the take rate.

This is the fragmented-distribution-on-top-of-concentrated-venue pattern that financial markets have repeatedly converged on — equities (NYSE, Nasdaq, ATSes) distributed via dozens of brokers; futures (CME, ICE) distributed via dozens of FCMs; options (NYSE Arca, NASDAQ PHLX, BOX, Cboe) distributed via dozens of broker-dealers. The venue layer concentrates because of network-effect liquidity dynamics. The distribution layer fragments because of customer-acquisition arbitrage.

The result: the actual margin in 2026 sits at the venue layer (Kalshi, Polymarket, soon Rothera), and the customer-acquisition battle sits at the distribution layer. Consumers will end up with five-to-ten different prediction-market apps on their phones, all routing to the same two-to-three underlying venues, all charging slightly different effective fees, all with different sport / category coverage.

The Liquidity Layer: Wall Street quants take their seat

In late 2024 and through 2025, the institutional liquidity providers who had been quietly building positions in prediction markets surfaced publicly.

Susquehanna International Group (SIG)

The earliest of the "Big Three" prop firms to take a public position. Disclosed in 2024 that it would act as a market maker on Kalshi. Subsequent moves:

2024: Acquired majority stake (with Robinhood) in LedgerX, a U.S. derivatives exchange.
January 2026: Joint venture with Robinhood acquired 90% of MIAXdx → Rothera.
Day-one liquidity provider role at Rothera.[41]

SIG's competitive advantage is twenty-plus years of derivatives market-making infrastructure — risk systems, position management, hedging algorithms — that translates with minimal modification to event contracts. SIG's stake in the venue layer (via Rothera) plus its market-making in Kalshi makes it the most structurally diversified prop firm in the prediction-markets space.

Jump Trading

February 2026: Acquired equity stakes in both Kalshi and Polymarket in exchange for committed liquidity provision. Structure: fixed equity stake at Kalshi; variable equity at Polymarket scaled to Jump's actual market-making activity.[13] The deal structure is significant — it explicitly trades capital commitment for permanent ownership stake, which is the same template equity-for-liquidity arrangements have historically taken in commodity exchanges (Cboe Global Markets equity originally distributed to liquidity providers, etc.).

Jane Street

Less publicly disclosed than SIG and Jump but generally believed to be the largest single market maker by realized notional volume on Kalshi. Jane Street's institutional positioning is mostly sub rosa — the firm releases minimal commentary by design. The firm is broadly understood to be running event-contract trading desks alongside its U.S. equities, options, ETFs, and U.S. Treasury market-making operations.

DRW

Chicago-based prop firm with strong derivatives heritage. Public positioning in prediction markets is more recent than SIG/Jump/Jane Street but DRW has been building dedicated desks. The firm's existing infrastructure in CME and ICE futures translates closely.

The "Big Three" + emerging tier

The financial-press standard formulation is SIG, Jane Street, and Jump as the "Big Three" of prediction-market liquidity provision, with DRW, Belvedere Trading, IMC, and similar firms in the second tier. The fact that this exact tiered structure is emerging — identical to the tiered structure of equity-options market making — is itself the signal that prediction markets have become institutional.[42]

Paradigm

Paradigm built a dedicated prediction-market trading terminal, marketed at hedge funds and prop firms. Real-time data, fast execution, complex position management across many markets simultaneously. The existence of this terminal is a leading indicator: institutional infrastructure precedes institutional flow. Bloomberg Terminal preceded mainstream institutional fixed-income trading by a decade in the 1980s.[43]

Hedge fund adoption

Goldman Sachs's global co-head of equities described predictions related to macro events and CPI prints as the categories where Wall Street attention is most focused. Tradeweb's co-head of global markets has publicly described a future where bulge-bracket banks have dedicated prediction-market trading desks with financial contracts as the anchor product. Margin trading on Kalshi enables a hedge fund to secure $50M of macro exposure for $5M of capital posted — making prediction markets useful as a macro hedge rather than merely as a retail entertainment product.[44]

The current state: institutional interest is high and infrastructure is being built, but most hedge funds are still in pilot/test mode rather than running material capital. The transition from "exploring" to "deploying" is the next twelve-month story.

The specific institutional use cases emerging:

Macro hedge construction — buying "Recession in 2026 YES" contracts as a tail-risk hedge against equity exposures.
CPI / Fed-decision binary trades — Kalshi and ForecastEx host markets resolving on specific CPI prints and Fed policy decisions; these are increasingly used as alternative to FOMC futures.
Geopolitical risk hedging — Iran ceasefire markets functioning as hedges for energy-exposed portfolios.
Election-cycle exposures — 2026 midterm markets used to hedge state-level regulatory regimes.

In each case, the prediction market provides cleaner, more directly-event-tied exposure than the corresponding traditional-finance instrument (FOMC futures, energy options, sector ETFs).

The Capital Layer: who's funding what

Track the venture capital flows and the picture clarifies further.

Kalshi's cap table

Coatue Management — lead investor in March 2026 $1B round at $22B valuation
Sequoia Capital — multi-round backer
Andreessen Horowitz (a16z) — multi-round backer
Henry Kravis (KKR co-founder) — personal investment
Charles Schwab — strategic backer
Susquehanna International Group — equity-for-liquidity
Robinhood — distribution-and-equity arrangements
Goldman Sachs alumni network — Mansour's prior employer; some informal capital flows
Notable advisor: Donald Trump Jr. (since January 2025)

Polymarket's cap table

Founders Fund (Peter Thiel) — early backer (2024)
General Catalyst — early backer
Vitalik Buterin (Ethereum cofounder) — early backer
ICE / NYSE parent — $2.6B cumulative ($2B October 2025 + $600M March 2026); also data distribution + tokenization partner
1789 Capital (Donald Trump Jr.) — eight-figure investment
Jump Trading — variable equity (February 2026)
Reportedly raising additional $400M at $15B valuation as of April 2026[45]

1789 Capital (Donald Trump Jr.)

The most interesting cross-venue capital story. 1789 Capital grew from $200M to $2B in 2025 and was then closed to new investors. Portfolio includes Elon Musk's SpaceX, Palmer Luckey's Anduril, AI company Groq, and Polymarket. Plus a separate 1789 Real Estate fund attracting $1B in months.[33]

The Kalshi-vs-Polymarket positioning: Trump Jr. is Strategic Advisor to Kalshi (since January 2025) AND Polymarket Advisory Board member with 1789 Capital eight-figure investment. The conflict-of-interest is acknowledged in mainstream reporting. The functional answer is that Trump Jr. and 1789 Capital are positioned to capture upside regardless of which venue dominates — a strategic hedge.

5c(c) Capital

The deepest signal of the structural maturation. $35M raise announced March 23, 2026. Twenty early-stage investments planned over two years. Infrastructure-only thesis: data tools, liquidity provision, compliance systems — explicitly not exchanges.

LP base reads as "every relevant prediction-markets and crypto-infra capital allocator":

Kalshi CEO Tarek Mansour (personal)
Polymarket CEO Shayne Coplan (personal)
Marc Andreessen (a16z, via Moneta Luna)
Micky Malka (Ribbit Capital founder)
Kyle Samani (formerly Multicoin)[14]

The implicit thesis is identical to the AI specialist VC thesis: top layer is winner-take-all, infrastructure layer is fragmented, capital allocated to the middle layer mostly evaporates. The fact that the CEOs of Kalshi and Polymarket are both LPs is a tell — they are betting on infrastructure that runs across their respective platforms rather than on either platform exclusively.

The implication: 5c(c)'s portfolio will tell us, more directly than any analyst report, where the prediction-market industry's fragmented-infrastructure layer is converging. Watch the specific 20 investments closely; they will collectively map the actual structural seams in the industry.

Strategic-investor ecosystem

The strategic-investor side of the cap table tells a cleaner story than the venture side:

ICE / NYSE (Polymarket): bringing 200+ years of futures-exchange operating experience and a global institutional-data distribution network into prediction markets.
Robinhood + Susquehanna (Rothera): bringing 25M+ retail accounts and tier-one derivatives-marketmaking infrastructure.
Jump Trading (Kalshi + Polymarket): bringing top-tier algorithmic trading capacity.
Founders Fund (Polymarket): bringing Thiel's network for political and macro market access.
Coatue (Kalshi): bringing public-market expertise — Coatue is one of the most active growth-stage investors in late-stage IPO candidates.

The pattern is the same as in AI: strategic capital follows operational capability, not just financial returns. ICE invests in Polymarket because Polymarket helps ICE sell macro-event-data subscriptions to ICE's existing institutional client base. Robinhood acquires Rothera because Rothera helps Robinhood capture margin from its own users that currently leaks to Kalshi. Jump invests in equity-for-liquidity because the equity stake compensates Jump for the marginal liquidity it provides at scale.

The Regulatory Layer: where the real shape is decided

This is the single highest-leverage layer for the entire industry's medium-term shape, and the place where the most decisive moves are happening.

CFTC — the central authority

Caroline Pham served as Acting Chair from January 20, 2025 through December 22, 2025. Her tenure produced:

A more permissive stance on prediction-market regulation (February 2025 statement: "current Commission interpretations regarding event contracts are a sinkhole of legal uncertainty")
A largely hands-off approach to sports event contracts
The Crypto CEO Forum
The framework for perpetual contracts and 24/7 trading
The Polymarket settlement that enabled CFTC re-approval (November 25, 2025)
The Aristotle / PredictIt full DCM + DCO approval (September 5, 2025)[16]

Mike Selig was confirmed as CFTC Chair on December 18, 2025 (53-43 Senate vote, sworn in shortly after). His background is unusually direct evidence of the administration's intentions:

Began his career in 2014 as a law clerk for then-CFTC Commissioner J. Christopher Giancarlo (now Chamber of Digital Commerce chair)
Partner at Willkie Farr & Gallagher LLP (2022–2025), in the Asset Management Department and Willkie Digital Works practice — specializing in crypto and event-contract regulatory work
Served as chief counsel of the SEC's Crypto Task Force and senior advisor to SEC Chairman Paul S. Atkins immediately prior to CFTC confirmation
Helped develop the post-Loper Bright regulatory framework for digital-asset securities markets
Participated in the President's Working Group on Digital Asset Markets and contributed to its report on "Strengthening American Leadership in Digital Financial Technology"
Education: Florida State University undergraduate; George Washington University Law School[46]

Selig's policy moves so far:

Directed staff to draft new event-contract rules with explicit goal of replacing the prior framework
Withdrew the 2025 staff advisory restricting sports event contracts
Has begun rulemaking on prediction markets
Filed CFTC suits against Arizona, Illinois, and Connecticut, asserting that "the CFTC has exclusive regulatory authority when it comes to commodity derivatives markets"
Stated publicly that the CFTC will not defer to courts on prediction-market jurisdictional disputes — the agency intends to protect federal preemption directly through litigation[16][47]

The Selig stance is the most consequential regulatory development in the prediction-markets industry's history. It positions the CFTC as the exclusive federal regulator and signals to states that any state-level enforcement action will face a coordinated CFTC defense. The result: the actual enforceability of Wisconsin's, Arizona's, Illinois's, Connecticut's, Maryland's, Nevada's, and New Jersey's various state-level lawsuits depends on whether federal courts ultimately accept CFTC's preemption argument or accept the states' framing that prediction markets are gambling rather than commodity derivatives.

The contrast with traditional CFTC posture is stark. Historically, CFTC chairs have preferred clarifying-rule approaches with light enforcement footprints. Selig's posture — actively litigating against state regulators while issuing aggressive new rules — is the most assertive CFTC stance in the agency's recent history. The administration explicitly chose this posture by appointing Selig from the SEC Crypto Task Force rather than promoting from within CFTC career staff.

State-level resistance

The states fighting prediction-market platforms (as of late April 2026):

Wisconsin (April 23, 2026) — sued Kalshi, Coinbase, Polymarket, Robinhood, Crypto.com. Cited Kalshi's own Instagram ads ("The First Nationwide Legal Sports Betting Platform") and Polymarket's marketing ("a platform where people can bet on the outcome of future events") as evidence that the platforms are gambling operations rather than commodity derivatives.
Arizona — filed criminal charges against Kalshi.
Illinois — sued by Kalshi (federal preemption suit).
Connecticut — sued by Kalshi (federal preemption suit).
New Jersey — Kalshi won a federal appeals court ruling on preemption.
Nevada — Judge Andrew Gordon (April 2025) granted Kalshi a preliminary injunction on preemption grounds.
Maryland — Judge Adam Abelson denied Kalshi's preliminary injunction motion. First state-level win on preemption against Kalshi.
New York — AG Letitia James sued Coinbase Financial Markets and Gemini for "illegally" running gambling operations.
Various others — cease-and-desist letters from at least 15 additional states.[17][36]

The cumulative federal-court docket is roughly 19–20+ active cases as of late April 2026.

The Supreme Court trajectory

Given the conflicting circuit-court rulings (NJ + NV for federal preemption; MD against), the question of whether federal preemption applies to state gambling regulation of prediction markets is on a near-certain Supreme Court collision course. Industry analysts expect the case to be granted certiorari within 12–18 months and decided within 24 months.

The Supreme Court ruling is the single most consequential single event in the prediction-markets industry's medium-term future. A federal-preemption win for Kalshi/Polymarket means:

All 50 states are open to regulated prediction-market operation (including Texas, Georgia, California, Florida, all of which currently lack state-licensed sportsbook frameworks)
The CFTC's exclusive jurisdiction is constitutionally settled
DraftKings/FanDuel/Fanatics's $48M Win for America Super PAC strategy collapses — there's nothing left to lobby states about
Bernstein's $1T-by-2030 projection becomes substantially more likely to clear

A federal-preemption loss means:

State-by-state legality, with major markets (Texas, California, Florida) effectively closed
Kalshi and Polymarket's value compresses 30–60%
DraftKings/FanDuel get to capture the closed states via state-licensed sportsbooks
The 2030 industry projection compresses to perhaps $300–500B rather than $1T

Given the structural tension and the public-policy stakes, the case will be argued. The Supreme Court's current median ideological position on federalism questions tilts toward states' rights but also toward predictable federal preemption when statutes are clear. The Commodity Exchange Act's preemption language has been previously interpreted broadly. Industry consensus probability of preemption being upheld: roughly 60–70%. That's a meaningful but not overwhelming margin.

The PredictIt resolution

PredictIt's case was the canary on Supreme Court direction. The CFTC revoked PredictIt's no-action letter in 2022 under the prior administration. PredictIt litigated in the U.S. District Court for the Western District of Texas, with rulings ultimately favorable. On July 22, 2025, the court entered final judgment in favor of PredictIt, declaring invalid the CFTC's prior efforts and barring the agency from future shutdown attempts based on PredictIt's current structure.[21]

Then on September 5, 2025, Aristotle, Inc. received CFTC approval for both DCM and DCO designation. PredictIt operates as a project of the Prediction Market Research Consortium, a nonprofit consortium of academic institutions, with operational support from Aristotle. The CFTC approval raised PredictIt's previous $850 single-contract cap to $3,500, and removed the 5,000-person-per-market cap entirely.[21]

Subsequently, Underdog Fantasy acquired Aristotle Exchange to bring prediction markets in-house, creating a new operational tier between PredictIt and the Kalshi/Polymarket duopoly.[21]

Self-policing as regulatory strategy

Kalshi's April 22, 2026 enforcement action — suspending and fining three congressional candidates for trading on their own races — is structurally significant. The fines were small ($539, $784, $6,229) but the suspensions are 5 years and the institutional signal is clear: Kalshi is positioning as a self-regulating exchange in the same template as the major U.S. financial exchanges. The CFTC structurally relies on exchange self-regulation as a precondition for granting and maintaining DCM status; Kalshi's enforcement actions support that posture.[19]

The deeper political-economy reading is that this enforcement pre-empts congressional pressure. Lawmakers from both parties had been raising concerns about insider-trading dynamics in prediction markets — particularly around congressional candidates trading their own races and around members of Congress trading markets resolving on legislation they vote on. By visibly enforcing against three small-dollar candidate cases, Kalshi creates the precedent for future enforcement against members of Congress without requiring CFTC or SEC action. The message: "We police this ourselves; you don't need to write new rules."

The Smaller Venues

Beyond the Kalshi/Polymarket/Rothera triangle, the venue landscape includes:

PredictIt / Aristotle Exchange

Now operates with full DCM + DCO designation, $3,500 per-contract cap, and academic-consortium backing (the Prediction Market Research Consortium, originally from Victoria University). Underdog Fantasy's acquisition of Aristotle Exchange brings new operational capital.

PredictIt's historical positioning as the academic/educational prediction market gives it a unique brand among researchers and political analysts. Recent product updates (new app, expanded market types, deeper liquidity) signal ambition beyond historical scale.

ForecastEx (Interactive Brokers / ICE-adjacent)

CFTC-licensed venue inside Interactive Brokers, focused on macroeconomic event contracts. Effectively the "institutional prediction market" — used primarily by Interactive Brokers' professional and institutional clients rather than retail.

Sporttrade

Sports-event-focused CFTC-licensed venue. Smaller volume than Kalshi's sports business but pure-play in approach.

Manifold Markets

~200,000+ active markets as of early 2026 — by far the largest market count, though most use Mana (virtual currency) rather than real money. Introduced "sweepstakes mode" using Sweepstakes Cash redeemable for prizes in 2025, creating a middle ground between play-money and real-money platforms.[48] Manifold's value is primarily as a research and market-design experiment platform rather than a primary financial venue.

Augur

The original decentralized prediction market on Ethereum. Augur v1 was effectively unusable due to gas costs and unresolvable edge cases. Augur v2 (Turbo) improved UX but never gained traction. Augur is rebooting as of late 2025/2026 with new oracle technology — a legitimate technical direction but starting from an effectively zero user base.

Hedgehog Markets

Solana-based decentralized prediction market with non-custodial smart-contract architecture. Politics, crypto prices, sports, tournament prizes. Smaller than Polymarket by orders of magnitude.

International venues

Smarkets (UK / Ireland / Malta / select European markets) — licensed by the UK Gambling Commission. Liquidity is limited. Eyeing U.S. expansion.

Betfair Predicts (UK) — Betfair began testing a new prediction-markets product in April 2026 as a UK launch experiment, currently in invite-only beta.

Brazil — Kalshi has announced expansion plans, putting Brazil on the global prediction-markets map.

Hong Kong, Singapore, UAE — these jurisdictions have not yet developed regulated prediction-market frameworks but are likely future regulatory targets given their general financial-markets sophistication.

The international layer is small relative to the U.S. layer and is structurally bounded by each jurisdiction's gambling-vs-financial-instrument regulatory framework. The U.S. is structurally the largest prediction-markets jurisdiction by virtue of CFTC's relatively permissive event-contract framework — which itself depends on the Selig CFTC's posture and the eventual Supreme Court ruling.

The Information Layer

A small but strategically important layer: companies aggregating prediction-market data and serving it to applications, agents, and institutional clients.

Adjacent News

API-first prediction-markets aggregator. Access to 40,000+ markets across multiple platforms (Kalshi, Polymarket, others) with rich metadata, AI-powered natural language queries, semantic search, news aggregation across 5,000+ sources, plus aggregate stats and platform health metrics. JWT-authenticated REST API at /api/v1/.[49]

The Adjacent News positioning is roughly analogous to Bloomberg's role in fixed-income markets in the 1980s — a single platform that consolidates data across fragmented venues and serves it to institutional and developer users.

Paradigm

The institutional trading terminal for prediction markets — the analog of Bloomberg Terminal at the high end of the market. Real-time data, complex position management, multi-market hedging tools.[43]

Polymarket Analytics

In-house analytics platform. The intersection of platform-native data and external research is where most institutional buyer-side intelligence currently sits.

ICE distribution

ICE's investment in Polymarket includes a data-distribution agreement. Polymarket event-driven data flows into ICE's institutional client distribution network, alongside the equities, futures, and options data ICE already distributes globally. This is the most institutionally significant data partnership in the prediction-markets industry — it pipes Polymarket signal into Bloomberg-Terminal-compatible institutional consumption.[5]

Bloomberg, FactSet, CapIQ

Have not yet directly licensed prediction-market data feeds at scale. The structural question is whether Bloomberg Terminal adds prediction-market price feeds as a default data category in 2026–2027. Industry expectation is yes — likely beginning with macro contracts (CPI, Fed rate decisions, GDP) where institutional demand is highest.

Open-source aggregators

Sites like the GitHub-hosted "Awesome Prediction Markets Tools" lists provide retail-grade aggregation. Smaller scale, useful for retail, less useful for institutional.

The AI Agent Layer: 30% of Polymarket already, and the missing oracle

This is the most underappreciated structural fact about prediction markets in 2026.

AI agents now represent roughly 30% of wallet activity on Polymarket. Approximately 37% of those agents report positive profit and loss, compared to 7-13% of human traders who consistently turn a profit.[15]

Read carefully: AI agents are both a larger share of wallet activity and materially more profitable than human traders on Polymarket. That asymmetry is unprecedented in retail financial markets — the typical pattern is that retail underperforms institutional, but agents specifically beating both retail and (apparently) most professionals is a genuinely new shape.

Polystrat

An autonomous prediction-market trading agent launched in early 2026. Uses natural language processing to let users specify high-level goals in plain text ("hedge my long crypto exposure with election outcome contracts," "buy whatever has +20% expected value from my model"). Selects markets across sports, politics, and economics using live data and on-chain liquidity. The user interface is essentially "prompt your way to a portfolio."

Baozi MCP Server

Open-source tooling enabling AI agents to discover, bet on, and manage prediction markets via the Model Context Protocol. Features include "Grandma Mei," an AI oracle for automated market resolution with on-chain proofs and tiered verification. The fact that this is built on MCP is structurally significant — MCP is the protocol Anthropic and other frontier labs are converging on for tool-using agent architectures.

Frontier-lab integration

OpenAI, Anthropic, and Google have not yet released first-party prediction-market clients. The current state is third-party integration: developers build agents using Claude/GPT/Gemini APIs, fetch market data from Adjacent News or directly from Kalshi/Polymarket APIs, and execute trades through brokerage accounts or on-chain wallets. The frontier labs themselves are not yet vertically integrated into the prediction-market stack.

This is one of the few specific gaps in the AI-x-financial-markets convergence that has not yet been closed by an existing major player. The structural reasoning is clear:

Frontier-lab agents need probabilistic inputs about future world states for forecasting, risk management, scenario analysis, and decision support.
Kalshi/Polymarket APIs serve trader-grade pricing data, not agent-consumable probability oracles.
The required interface is effectively a "probability oracle" — a clean, low-latency, agent-native signal of "what's the market's current implied probability of X?" with reliability guarantees, settlement clarity, and dispute-handling semantics.
That interface does not currently exist as a first-party product from either Kalshi/Polymarket or the major frontier labs.

The 5c(c) Capital infrastructure-only thesis is implicitly a bet that someone will build this. Whether it's Adjacent News extending into agent-native interfaces, Polymarket using its blockchain composability to ship native-agent SDKs, or a new specialist startup, the gap will be filled within 12–24 months.

The market-design implication is more interesting than the product-design implication. If 30% of prediction-market wallet activity today is AI agents, and AI capacity grows on the trajectory the AI economy is on, that share goes to 50% in 2027 and 70%+ by 2029. At that point, the prediction-market price is the AI agent consensus. Which means prediction markets have crossed from "human price discovery using markets" to "AI consensus operationalized as price."

That is a genuinely new species of market.

The further implication is that whoever provides the cleanest agent-native consumption layer for prediction-market signals captures a structurally important position. The opportunity is roughly analogous to the early-1990s opportunity Bloomberg captured by being the first to ship a unified terminal for fixed-income market data. Whoever does this for AI agents in 2026-2027 captures an analog position in a much larger eventual market.

The candidates for who builds this:

Adjacent News extending its API into native agent/MCP servers
Polymarket shipping a first-party agent SDK on top of its blockchain composability
Anthropic or OpenAI building it as a first-party product (unlikely near-term — frontier labs are vertically internalizing AI applications, not financial-data products)
A new specialist funded by 5c(c) Capital or similar

The structural odds favor candidate 4. Specialist startups built to serve agent consumption from day one will have the cleanest product, the cleanest economics, and the most aligned incentives. The big incumbents will partner with or acquire them.

The Volume Story: $51B (2025) → $66B in 4 months → $240B (2026e) → $1T (2030e)

The volume trajectory is the cleanest signal of structural transition.

2025 actuals

Polymarket alone: ~$51B in contract volume across all of 2025
Kalshi 2025: dramatic ramp from ~$100M weekly in early 2025 to $3B+ weekly by Q4 2025
Industry total 2025: roughly $60–80B (some industry estimates put the full number at $44B for Polymarket and proportionally lower for Kalshi during the U.S. relaunch period)[3][22]

2026 YTD (through April 20)

Kalshi: $37.49B
Polymarket: $29.23B
Combined: ~$66.7B
Already greater than the entire 2025 total in just four months

Bernstein's projection

2026 industry total: $240B (projected) — a 370% YoY increase
2030 industry total: ~$1T at ~80% CAGR
Industry revenue: $400M (2025) → $2.5B (2026) → $10.8B (2030) at current take rates[3]

What the projection implies

A $1T 2030 industry would put prediction markets at roughly:

1–2% of global equity-trading volume
10–20% of global commodities-futures volume
30–50% of global retail crypto-derivatives volume

These are non-trivial fractions of established asset classes. Crucially, prediction-markets revenue at $10.8B would put the industry at roughly the size of the entire U.S. cash equities market-making revenue pool today — i.e., on the order of CME Group's annual revenue. That's not a tail asset class. That's a major financial-markets segment.

The category distribution

Bernstein's estimate of the category mix:

Today: sports >60%, politics ~20%, crypto/macro/other ~20%
2030: sports cut in half; macro and finance grow most; politics steady[3]

The Polymarket-specific category breakdown:

~42% politics (the Polymarket DNA)
~31% crypto
~25% sports + economics + other

Kalshi's distribution:

~70%+ sports
Politics, crypto, macro, weather, entertainment splitting the remainder

The asymmetry is structural: Polymarket has the global geopolitical/macro/crypto market, while Kalshi has the U.S. sports market. Each platform's strengths reflect their respective regulatory and user-acquisition histories.

The 2026 calendar of catalysts

The remaining 2026 calendar has at least four major volume catalysts:

World Cup 2026 (June 11 – July 19) — Polymarket already has $769M+ on the Winner market alone. Industry expectation is $2-3B+ combined Kalshi + Polymarket volume during the tournament.
2026 midterm elections (November 3) — historically the highest-volume political-event window. Polymarket already prices Democratic House control at 85.5%; Senate at 50/50; full election volume will be 5-10× current pricing-window volume.
NBA Finals + NFL season — peak sports-event windows in October-November.
Iran ceasefire resolution — whichever direction it resolves will drive concentrated volume in the resolution window.

The combination of these catalysts plus the ongoing AI-agent participation growth makes a 2026 actual volume figure of $200-280B plausible, not $240B as Bernstein's central case. Bernstein's projection may turn out conservative.

The downward catalysts: a Supreme Court ruling against federal preemption, a major UMA-style oracle attack, another Polymarket-style regulatory event, a CoreWeave-style fragility event in adjacent infrastructure, or sustained Iran war escalation that disrupts global financial flows. None are base case but all are non-trivial tails.

The fee economics

Understanding the unit economics of the venues is essential for understanding which players survive a downturn.

Kalshi's fee math

The taker fee formula: 0.07 × contracts × price × (1 - price). Maker is 0%.

At a 50/50 contract with a $1 face value, a $1,000 position pays $35 in taker fees. The implied take rate at 50% probability is 3.5%. At 90/10 odds, the take rate is 0.63%. At 99/1 odds, 0.07%.

This is heavily back-loaded toward contested markets — exactly where most volume sits. The blended effective take rate across Kalshi's volume is in the 1.5–2.5% range. At $37B YTD volume, that implies revenue contribution roughly $550M–$925M YTD. Annualized implied revenue: $1.7B–$2.8B, which broadly matches the publicly reported $1.5B annualized figure (the gap is because actual mix-shifts and price-distribution effects compress effective take rate below the theoretical maximum).

Polymarket's fee math

The Polymarket fee formula: 0.0625 × Price × (1 - Price), applied to a subset of markets (crypto direction, NCAAB after Feb 18, Serie A after Feb 18). For Polymarket US (CFTC-regulated): flat 0.10% on contract premium, $0.001 minimum, 0% maker.

Polymarket's effective take rate is materially lower than Kalshi's because most markets remain zero-fee on the offshore product. Reported daily revenue post-March 30 fee update: ~$1M, implying ~$365M annualized. At $29B YTD volume, this is a 0.5% effective take rate — roughly one-third Kalshi's effective rate.

The implication: Polymarket subsidizes liquidity by accepting lower take rates, in exchange for capturing more of the AI-agent and crypto-native trader population. Kalshi optimizes for take rate, in exchange for accepting somewhat lower volume share. Both economic models work, but they converge to different equilibria over time. Polymarket's POLY token launch will eventually monetize the ecosystem differently — the airdrop creates a token holder base whose interests align with platform-volume growth.

What sustainable take rates look like

Comparison points:

Equity exchanges (NYSE, Nasdaq): take rates approximately 0.0001–0.0005% per trade after rebates
Futures exchanges (CME, ICE): take rates approximately 0.05–0.15% per trade
Options exchanges (Cboe): take rates approximately 0.10–0.30% per trade
Crypto spot exchanges (Coinbase, Binance): take rates approximately 0.10–0.50% per trade
Sportsbooks (DraftKings, FanDuel): take rates approximately 7–10% per wager (the "vig")

Prediction markets at 1–3% are between options/futures exchanges and sportsbooks — closer to sportsbook than to traditional exchange. Over time, with institutional adoption, take rates will compress toward the futures/options exchange range. The current take rates reflect early-stage venue economics where retail dominates and competition is limited; institutional flow will pull take rates down.

Bernstein's $10.8B 2030 revenue projection at $1T volume implies a 1.08% blended take rate — already significantly below current levels. The implicit assumption is that take rates compress 30–50% as the industry institutionalizes.

The competitive implication is that 2030's surviving venue economics depend on volume scale. A venue at $100B annual volume with a 1% take rate ($1B revenue) is viable. A venue at $20B annual volume with the same take rate ($200M revenue) is not, because the operational, regulatory, and liquidity-incentive costs of running a CFTC DCM are roughly $100–150M per year. The minimum viable venue size in 2030 is roughly $50B annual volume. Below that, the venue either consolidates into another or shuts down.

This is why the duopoly (Kalshi + Polymarket) is structurally durable: both are above the minimum viable scale already, and both have strong network-effect liquidity moats. The third venue (Rothera) has a credible path to viability via Robinhood's existing user base. Anyone else is fighting an uphill battle.

The Supreme Court question, in compressed form

The most consequential single near-term variable for the entire industry. Walking through the case structure:

The legal architecture

Federal: The Commodity Exchange Act (CEA) and CFTC regulations under the CEA establish federal authority over commodity-derivatives markets and event contracts. The CFTC has approved Kalshi, Polymarket (via amended order), Crypto.com, ForecastEx, Aristotle, MIAXdx/Rothera, and others as DCMs and/or DCOs. Federal-court decisions in late 2024 (election contracts) and through 2025 confirmed federal authority over event contracts.

State: Many state gambling laws define "wager" and "bet" broadly enough to capture event contracts. State attorneys general and gaming commissions argue that federal preemption does not apply because event contracts are functionally gambling, not commodity derivatives.

The conflicting circuit-court rulings

NJ federal appeals: sided with Kalshi on preemption (Kalshi win)
NV district: granted Kalshi preliminary injunction on preemption (Kalshi win)
MD district: denied Kalshi preliminary injunction; held federal preemption likely doesn't apply (state win — first major state-level legal victory)
Various other states: cease-and-desist letters, criminal charges (AZ), state-AG lawsuits (WI, NY against Coinbase/Gemini)

When circuit courts reach conflicting outcomes on a question of federal preemption with industry-wide stakes, the Supreme Court typically grants certiorari within 12–18 months and decides within 18–24 months.

The likely arguments

For preemption (Kalshi/Polymarket position):

The CEA's preemption clause is broad — Section 16(e)(2) of the CEA preempts state law that "treats" a futures contract or commodity option as gambling.
Event contracts are commodity options under longstanding CFTC interpretation.
The CFTC has explicitly approved these venues as DCMs and is exercising regulatory oversight.
A state-by-state regulatory regime would defeat the purpose of federal commodities-market regulation.

Against preemption (state position):

Sports event contracts are more analogous to traditional gambling than to commodity derivatives.
State sovereignty over gambling is a longstanding principle dating to Murphy v. NCAA (2018) and earlier.
The CFTC's interpretation expanded recently and exceeds the CEA's actual statutory grant.
Event contracts on outcomes that the trader may directly influence (sports involving local teams) are inherently gambling.

The industry consensus probability

Roughly 60-70% probability that federal preemption is upheld. The key swing factors:

Whether the Court treats sports event contracts differently from political event contracts
Whether the Court accepts CFTC's expansionary interpretation or limits it to the CEA's plain language
Whether the Court applies its post-Loper Bright deference framework narrowly or broadly

The decision tree

If preemption is upheld (60–70% probability):

All 50 states open to regulated prediction-market operation
Bernstein's $1T-by-2030 trajectory becomes substantially more likely
DraftKings/FanDuel/Fanatics's state-by-state lobbying strategy fails
Kalshi/Polymarket valuations expand (potential 50–100% upside on top of current marks)

If preemption is rejected (30–40% probability):

State-by-state legality
Texas, California, Florida, and similar large-population states close
Kalshi/Polymarket valuations compress 30-60%
DraftKings/FanDuel capture the closed states via state-licensed sportsbooks
2030 industry projection compresses to perhaps $300-500B
The 5c(c) Capital infrastructure thesis becomes more important rather than less — fragmented state-by-state regulation requires more compliance/data/integration tooling

The actual ruling will be the single largest valuation reset event in the industry's history. Markets currently price for the 60-70% preemption-upheld scenario; deviation in either direction will produce a sharp valuation move.

A subtle structural risk

Even if preemption is upheld, the CFTC could itself become less permissive under a future administration. The asset class's structural position depends on continued CFTC permissiveness. Selig is sympathetic, but a 2028 administration change with a different CFTC chair could re-tighten the framework. This is the single highest "what if politics changes" risk variable for the industry — and the reason 5c(c) Capital and the institutional capital flows are weighted toward infrastructure (which is regulator-agnostic) rather than venue equity (which is regulator-dependent).

The Trump administration's posture

The political-economy reading of the prediction-markets industry under the second Trump administration is unusually direct.

Regulatory: pro-prediction-markets

Mike Selig at CFTC: pro-event-contract, pro-perpetuals, pro-crypto, asserting exclusive federal jurisdiction
Caroline Pham (predecessor): laid the regulatory foundation with the Polymarket re-approval, the Aristotle DCM grant, and the hands-off sports-event-contract approach
Howard Lutnick at Treasury: longtime crypto advocate (Cantor Fitzgerald custodied Tether reserves), broadly aligned with prediction-market expansion
The administration has not pursued any restrictive action against prediction markets during 2025-2026

Capital: actively invested

Donald Trump Jr. is Strategic Advisor to Kalshi (since January 2025) and Polymarket Advisory Board member with 1789 Capital eight-figure investment in Polymarket
1789 Capital portfolio: SpaceX, Anduril, Groq, Polymarket, plus 1789 Real Estate
The capital flows make political-economy sense: the family is positioned to capture upside regardless of which prediction-market venue wins[33]

Adjacent crypto policy

Trump signed legislation supportive of stablecoin frameworks
The CFTC's perpetuals authorization is in flight under Selig
Coinbase's Deribit acquisition closed in August 2025 without regulatory pushback

The DraftKings/FanDuel countercase

The $48M Win for America Super PAC explicitly funded by DraftKings, FanDuel, Fanatics, and Bet365
Targeting Texas, Georgia, and 15 other states
The political driver is keeping prediction markets out of state-by-state sports-betting regimes
This is the only significant Trump-adjacent political pushback against prediction markets

The aggregate posture: the Trump administration is pro-prediction-markets at the federal regulatory and capital-allocation level, with the offsetting political pressure coming from sportsbooks (themselves significant Trump-aligned campaign donors). The federal-vs-state preemption fight is in some sense a fight between two Trump-aligned constituencies: prediction-market operators (with Trump Jr. as advisor + investor) vs. sportsbooks (with their substantial political donations). Which side wins at the Supreme Court will signal which constituency the administration ultimately favors.

The Information Finance reframing

A subtle but important shift in how the industry is positioning itself: the asset class is being publicly renamed from "prediction markets" to "Information Finance."

The financial press has begun adopting the new terminology. Bernstein's research uses "Information Finance" alongside "prediction markets." Wall Street institutional desks describe the asset as "Information Finance" rather than "prediction markets" or "event contracts." Tarek Mansour's "financialize everything" framing fits naturally under the Information Finance banner.

The reframing matters because:

It positions the asset class as a peer of equities, fixed income, FX, commodities, crypto, rather than as a peer of sports betting. This is the key distinction for institutional capital allocation.
It signals to regulators that this is finance, not gambling, reinforcing the federal-preemption argument.
It maps to the structural reality of the industry's evolution — sports may be the entry point but the structural trajectory is toward macro events (CPI, Fed decisions, GDP, geopolitical resolution) where the analogy to traditional finance is closest.
It makes the asset class brandable for institutional buyers in a way "prediction markets" never could be. A pension fund's investment committee can authorize allocation to "Information Finance"; it cannot authorize allocation to "prediction markets."

The reframing is doing real work. Watch the financial-press transition over 2026: by year-end, "Information Finance" will be the standard analyst-coverage terminology, and "prediction markets" will be the legacy-retail term.

A fragility map

Translating the structural picture into specific failure points:

Conduit	Failure mode	Cascade path
Supreme Court ruling	Federal preemption rejected	Kalshi/Polymarket valuations compress 30-60%; large states close; sportsbooks win state-by-state
UMA oracle	Second governance attack on a major resolved market	Polymarket reputation hit; institutional adoption pauses; retail trust deficit
CFTC posture change	2028 administration change with restrictive CFTC chair	Industry framework tightens; institutional flow stops; valuations compress
Polymarket regulatory reversal	DOJ/CFTC reopens (declination revoked)	Polymarket exits U.S.; valuations to zero; cap-table holders take 100% loss on U.S. exposure
State-level regulatory cascade	Multiple states win post-Wisconsin and copy-paste enforcement	Kalshi/Polymarket compelled to exit states piecemeal; volumes compress 20-40%; state-by-state legal cost balloons
Sportsbook lobbying cascade	Win for America Super PAC succeeds in TX, GA, additional states	State-licensed sportsbooks displace prediction-market platforms in major markets
Insider-trading scandal	Member of Congress trades own legislation market	Senate Banking Committee mandates additional regulation; framework tightens
Single-venue insolvency	One distribution surface (e.g., a smaller crypto exchange) goes insolvent with prediction-market positions	User confidence damaged industry-wide; regulatory pressure increases
Iran war escalation	Sustained Hormuz closure → financial flows shock → institutional risk-off	All risk assets compress; prediction-market volumes drop 20-40% during the shock window
Trump approval collapse	Approval below 30% sustained, midterm wipeout	Possible 2028 Democratic landslide → CFTC posture reset risk

Each conduit is, individually, a single point of failure for some substantial portion of the industry's value. The conduits are not independent — they correlate. A Supreme Court loss on preemption would coincide with a state-level enforcement cascade. An oracle attack would coincide with institutional pullback. A CFTC posture change would coincide with valuation compression across the entire venue layer.

The dependency structure is tight enough that the industry's medium-term shape is best understood as a single correlated mega-bet on the continued permissiveness of CFTC + the upholding of federal preemption + the absence of major operational failures. If any one of those three breaks, the entire valuation stack reprices.

Plain-English version

Prediction markets did $66B in trading volume in just the first four months of 2026 — already more than the entire industry total for 2025. Bernstein analysts project the industry hits $1 trillion in annual volume by 2030.

Two venues dominate: Kalshi at $22B valuation (Tarek Mansour, MIT 2018, Goldman/Citadel quant background, founded 2018, CFTC-approved 2020) and Polymarket at $15B in negotiation (Shayne Coplan, NYU dropout, founded 2020, world's youngest self-made billionaire as of October 2025, FBI raid in November 2024, CFTC re-approval in November 2025). Together they hold over 95% of U.S. retail prediction-market volume.

The distribution layer is fragmented and getting more fragmented every quarter. Robinhood's prediction-markets line is now its fastest-growing revenue segment ($350M ARR, 30% of Kalshi's volume). DraftKings, FanDuel, Fanatics, Coinbase, Crypto.com, Webull, Kraken, and ForecastEx have all entered. Robinhood + Susquehanna acquired MIAXdx in January 2026 and are launching their own venue (Rothera) in Q2 2026 — adding a third top-layer player.

The liquidity layer is institutional now. Susquehanna disclosed market-making at Kalshi in 2024. Jump Trading took equity stakes at both Kalshi and Polymarket in February 2026 in exchange for liquidity provision. Jane Street is broadly understood to be the largest single market maker by realized volume. Goldman Sachs's co-head of equities is publicly describing macro event contracts as the focus area for institutional attention. Tradeweb's co-head expects bulge-bracket banks to operate prediction-market trading desks.

Wall Street infrastructure is being built. Paradigm built a dedicated trading terminal. Adjacent News operates an aggregator API across 40,000+ markets. ICE distributes Polymarket data into traditional finance institutional channels.

The capital flowing in is enormous and coordinated. Coatue led Kalshi's $1B March 2026 round at $22B. ICE put $2.6B into Polymarket cumulatively. 5c(c) Capital raised a $35M infrastructure-only fund with Kalshi's CEO + Polymarket's CEO + a16z + Ribbit + Multicoin's former managing partner as LPs. Donald Trump Jr.'s 1789 Capital invested in Polymarket while Trump Jr. simultaneously serves as Strategic Advisor at Kalshi.

The CFTC is the most powerful financial-markets regulator no one is paying attention to. Caroline Pham (Acting Chair, January-December 2025) laid the foundation. Mike Selig (confirmed December 2025, formerly SEC Crypto Task Force chief counsel) directed staff to draft new event-contract rules, withdrew the staff advisory restricting sports contracts, and is suing Arizona, Illinois, and Connecticut to assert "exclusive regulatory authority."

State-level resistance is the biggest single risk. Wisconsin sued Kalshi, Polymarket, Coinbase, Robinhood, and Crypto.com on April 23, 2026 — joining 19+ existing federal lawsuits, plus Arizona criminal charges, plus the New York AG suit against Coinbase and Gemini. Mixed court results so far: Kalshi won preemption rulings in NJ federal appeals and NV district; lost the first state-side preemption ruling in Maryland. A Supreme Court federal-preemption case is essentially inevitable within 24 months.

AI agents are 30% of Polymarket wallet activity, with 37% of those agents profitable versus 7-13% of human traders. This is unprecedented in retail financial markets. Polystrat (autonomous prediction-market trading agent) and Baozi MCP Server (open-source AI agent infrastructure) are early indicators of where the layer goes. No major frontier lab has shipped a first-party agent-native consumption layer for prediction-market data — the structural opportunity that 5c(c) Capital and its peers are betting will be filled.

Sports is the entry point. Macro is the endgame. Kalshi is >70% sports volume today. Polymarket is 42% politics, 31% crypto, the rest sports+macro. Bernstein projects sports cut in half by 2030 as macro and finance contracts grow. The Iran ceasefire markets alone have done $377M+ in volume. The 2026 World Cup (June 11+) will likely add another $2-3B+. The 2026 midterms in November will be the political-cycle peak.

The pattern is identical to AI. Top layer winner-take-all (Kalshi+Polymarket, soon +Rothera). Distribution layer fragmented (Robinhood, DraftKings, etc.). Infrastructure layer specialist (5c(c), Paradigm, Adjacent News). No middle-layer winner. Capital allocated to the middle evaporates.

Sportsbook industry pushback is the visible counter-force. DraftKings, FanDuel, Fanatics, and Bet365 jointly funded the $48M Win for America Super PAC targeting Texas, Georgia, and 15 other states. The political fight is whether prediction markets stay federally preempted or whether states get to enforce gambling regulations against them.

Kalshi has begun self-policing on insider trading. April 22, 2026 enforcement action: three congressional candidates fined and suspended for trading their own races. The institutional signal is that Kalshi is becoming a self-regulating exchange, the same template as the major U.S. financial exchanges — preempting congressional pressure for additional regulation.

The asset class is being renamed from "prediction markets" to "Information Finance" in financial press and Wall Street institutional discourse. The reframing positions the asset as a peer of equities/fixed income/commodities/crypto rather than sports betting. By year-end 2026, "Information Finance" will likely be the standard analyst-coverage terminology.

What this all points to

The structural transition under way is from "prediction market" to "Information Finance." The category itself is being renamed, both in financial-press coverage and in how Wall Street institutional desks discuss the asset. The reframing matters because Information Finance positions the asset class as a peer of equities, fixed income, FX, commodities, and crypto — rather than as a peer of sports betting.

Mansour's vision — financialize every difference of opinion — is the literal extension of how every prior generation of financial instruments emerged. Equities financialized claims on cash flows. Bonds financialized claims on repayment. Futures financialized claims on delivery. Options financialized claims on contingent payoffs. Credit derivatives financialized claims on default events. Event contracts financialize claims on the resolution of any specifiable proposition. There's no theoretical ceiling beyond the enumeration of specifiable propositions, which is a much larger set than any prior asset class has supported.

The single most important structural fact is that the layer is converging on the same shape as AI. Top layer winner-take-all, distribution layer fragmented, infrastructure layer specialist. Middle layer dies. The implication is that the same investment logic applies: don't try to build "general prediction-market tooling for industry X" — build a specific venue, or a specific distribution channel, or a specific piece of infrastructure with a clean integration moat.

The single most important near-term variable is the Supreme Court federal-preemption case. Given the conflicting circuit-court rulings (NJ + NV for Kalshi vs. MD for the state), certiorari is near-certain within 18 months and a ruling within 24 months. The decision tree:

Upheld (60–70% probability): $1T-by-2030 trajectory clears; Kalshi/Polymarket valuations expand 50–100%; sportsbooks lose the state-by-state moat
Rejected (30–40% probability): state-by-state regulation; major-population states close; Kalshi/Polymarket valuations compress 30–60%

The single most underpriced structural fact is the AI-agent share. 30% of Polymarket wallet activity is AI agents today. The trajectory takes that to 50% by 2027 and 70%+ by 2029. The economic implication is that prediction-market prices increasingly reflect AI agent consensus rather than human price discovery. The market-design implication is that whoever builds the cleanest agent-native consumption interface for prediction-market data captures a structurally important position. No major frontier lab has shipped this product yet. Anthropic, OpenAI, and Google are not vertically integrated into the prediction-market stack. That gap will close — but who closes it determines whether prediction-market venues capture more of the AI value chain or whether frontier labs do.

The most uncomfortable observation about the layer is that it is structurally identical to the convergence happening in AI. The same set of capital allocators is investing in both — a16z is in both. Sequoia is in both. ICE is in Polymarket, partnering with NYSE-adjacent traditional finance for distribution. The same prop firms (SIG, Jane Street, Jump) are providing liquidity in both prediction-market venues and AI infrastructure. The regulatory dynamics rhyme — federal vs. state, exclusive jurisdiction claims, courts as the ultimate arbiter. The capital-flow dynamics rhyme — top layer winner-take-all, infrastructure layer specialist, middle layer dies.

The most actionable observation is that the alpha is not in betting on prediction markets succeeding or failing as an asset class. That outcome is already largely priced (Kalshi at $22B and Polymarket at $15B implicitly price for substantial future growth). The alpha is in:

The Supreme Court probability — direct optionality on the preemption ruling
The AI-agent infrastructure gap — who builds the agent-native consumption layer
Specific conduit vulnerabilities — UMA oracle attacks, single-state regulatory cascades, sportsbook industry political pressure
The international layer — Brazil, UK, eventually India/Japan as Kalshi/Polymarket expand abroad
The integration moats at distribution layer winners — Robinhood-Susquehanna's Rothera execution, DraftKings's Railbird utilization, Fanatics's 100M-customer-base monetization
The 5c(c) Capital portfolio — the explicit list of infrastructure bets that the most-informed industry insiders are making

That is what the current state actually points to. Prediction markets are no longer a curiosity layer. They are becoming the second financial-markets layer — Information Finance — a real-time, regulated, executable price for every specifiable proposition about the future. Kalshi and Polymarket are the venues that captured the inflection. Robinhood, DraftKings, FanDuel, Fanatics, Coinbase, and the upcoming Rothera are the distribution surfaces. SIG, Jane Street, and Jump are the institutional liquidity providers. The CFTC under Mike Selig is the regulator. The Supreme Court is the next pivot point. AI agents are already the largest single trader category by wallet share and are taking market share from humans every month.

Information Finance has arrived. The question is which actors — and which infrastructure layers — capture the most of its trillion-dollar trajectory.

The next eighteen months will decide it.

Sources

[1] Kalshi YTD volume $37.49B / Polymarket YTD volume $29.23B (through April 20, 2026); week of April 6–11 record $6.02B combined; week of April 13 $5.10B combined — DeFiRate, weekly volume aggregator.

[2] Kalshi Super Bowl Sunday 2026: $871M Kalshi single-day volume; over $1B combined Super Bowl markets on Kalshi alone; $1.38B across Kalshi + Polymarket combined — Fortune, CBS Sports, pm.wiki.

[3] Bernstein research: $240B 2026 industry volume projection (370% YoY); $1T 2030 projection (~80% CAGR); revenue $400M (2025) → $2.5B (2026) → $10.8B (2030); sports >60% today, cut in half by 2030 — CNBC, Coindesk, The Block.

[4] Kalshi $22B valuation March 2026 (Coatue-led $1B); $11B December 2025; $5B October 2025; ~$1.5B annualized revenue; 90% U.S. share — Bloomberg, Coindesk, Kalshi News.

[5] Polymarket: ICE $2B at $9B (October 2025) + $600M (March 2026) + reportedly $400M at $15B (April 2026); 2025 contract volume $51B+; ICE data distribution + tokenization partnership — Fortune, ICE press release, Coindesk.

[6] Polymarket CFTC Amended Order of Designation, November 25, 2025; U.S. relaunch — Coindesk, Bitcoin Magazine, Reason.

[7] Robinhood prediction markets: $350M ARR; 30% of Kalshi total volume; 9B+ contracts; 1M+ customers; Q1 2026 revenue ~$1.14B (~22-26% YoY) — Robinhood IR, Motley Fool, Yahoo Finance.

[8] MIAXdx → Rothera: 90% Robinhood + Susquehanna joint venture / 10% MIAX retained; closed January 20, 2026; Q2 2026 product launch — Robinhood newsroom, MIAX press release, Markets Media.

[9] DraftKings Predictions December 19, 2025 launch via Railbird acquisition; live in 38 states — Boston Globe, Sportico, Covers.

[10] Fanatics Markets December 2025 via Paragon Global Markets acquisition; Crypto.com partnership backend; 24 states; Phase 2 early 2026 expansion — Fanatics IR, Coindesk, Crypto.com.

[11] Coinbase prediction markets to all U.S. customers January 27, 2026; Deribit $2.9B acquisition closed August 2025 — Coindesk, Coinbase newsroom.

[12] Polymarket perpetual futures launch April 21, 2026; Kalshi "Timeless" perpetuals launch April 27, 2026; CFTC perpetuals authorization expected — CNBC, Coindesk, The Information, Yahoo Finance.

[13] Jump Trading equity-for-liquidity stakes in Kalshi (fixed) + Polymarket (variable), February 2026 — Bloomberg, FinanceMagnates.

[14] 5c(c) Capital $35M fund March 23, 2026; ~20 startups over 2 years; infrastructure-only thesis; LPs include Tarek Mansour, Shayne Coplan, a16z, Ribbit, Multicoin Samani — Crypto Briefing, Fortune, Coindesk.

[15] AI agents 30% of Polymarket wallet activity; 37% profitable agents vs 7-13% profitable humans; Polystrat autonomous trading agent — New York City Servers blog, Alphascope.

[16] Caroline Pham Acting CFTC Chair January 20, 2025 - December 22, 2025; Mike Selig confirmed December 18, 2025 (53-43 Senate vote); Selig directives on event contracts, withdrawal of sports advisory, Arizona/Illinois/Connecticut suits — CFTC press releases, Politico Pro, WilmerHale, Sidley Austin.

[17] Wisconsin v. prediction markets April 23, 2026 (Kalshi/Coinbase/Polymarket/Robinhood/Crypto.com); 19+ federal lawsuits prior; Arizona criminal charges; Maryland court denied Kalshi preemption preliminary injunction (first state win); NJ federal appeals + NV district preemption rulings (Kalshi wins) — Coindesk, KERA News, Brownstein Hyatt, Stinson LLP, Epstein Becker Green.

[18] DraftKings + FanDuel + Fanatics + Bet365 $48M Win for America Super PAC targeting Texas, Georgia, 15 other states — Bettors Insider, iGaming Business.

[19] Kalshi insider-trading enforcement April 22, 2026: Mark Moran (VA), Matt Klein (MN-02), Ezekiel Enriquez (TX-21); fines $539-$6,229; 5-year suspensions — CNBC, CNN, Al Jazeera, Kalshi News, NBC News.

[20] Intrade closed March 10-11, 2013; CFTC sued November 2012 for off-site options exchange; 50M+ accounts, $200M+ wagered lifetime; correctly predicted 2008 election + 49/50 states 2012 — Wikipedia, CNBC, BuzzFeed, MarketsWiki.

[21] PredictIt litigation final judgment July 22, 2025 (Western District of Texas); Aristotle DCM + DCO approval September 5, 2025; PredictIt $850 → $3,500 per-contract cap; 5,000-person-per-market cap removed; Underdog Fantasy acquired Aristotle Exchange — PredictIt press release, GlobeNewswire, Odds Shark, NextEventHorizon.

[22] Polymarket November 2024 election: $3.6B+ in election bets — Washington Post, CNBC.

[23] Tarek Mansour Kalshi background: MIT 2018 (CS + Math + Master's of Engineering); Goldman Sachs quant trader; Citadel global macro; co-founded Kalshi 2018 with Luana Lopes Lara; CFTC approval 2020, launch 2021; "financialize everything" vision — Kalshi News, MIT, Wikipedia, BillionaireReporter.

[24] Kalshi Brazil expansion announcement — Kalshi News.

[25] Kalshi fee structure: 0.07 × contracts × price × (1-price); 0% maker; daily USDC rebates — Polytrage, DeFiRate, Laika Labs, pm.wiki.

[26] Shayne Coplan: NYU computer science dropout; founded Polymarket 2020; world's youngest self-made billionaire on Bloomberg Billionaires Index October 2025; Forbes net worth ~$1.0B — Wikipedia, Fortune, Bloomberg.

[27] Polymarket Spring 2024 funding ~$70M from Founders Fund (Peter Thiel), Vitalik Buterin, General Catalyst — Britannica Money, Wikipedia.

[28] Polymarket FBI raid Coplan apartment November 13–14, 2024; phones/electronics seized; DOJ investigation re: 2022 CFTC settlement violation via VPN access — CNBC, Coindesk, Axios.

[29] DOJ + CFTC investigation closed July 2025 with declination notices — CNBC, The Market Periodical.

[30] POLY token + airdrop confirmed by CMO Matthew Modabber on Degenz Live (October 24, 2025); launch sequenced after U.S. relaunch; ICE collaboration includes tokenization — Yahoo Finance, Blockworks, BraveNewCoin.

[31] Polymarket Iran-related markets: $377.8M+ across 20 active markets since February 28, 2026 war start; "US x Iran ceasefire by...?" $280.1M alone; "Iran x Israel/US conflict ends by...?" $100.6M — Polymarket, StartupHub.

[32] Polymarket World Cup 2026: $769.9M+ on Winner market alone since launch July 2, 2025; 197 active markets; $770.9M+ across all World Cup markets; tournament June 11 - July 19, 2026 — Polymarket, Oddpool, VegasInsider.

[33] Donald Trump Jr.: Strategic Advisor Kalshi (January 2025); 1789 Capital eight-figure Polymarket investment + Advisory Board seat; 1789 Capital grew $200M → $2B in 2025; portfolio includes SpaceX, Anduril, Groq, Polymarket — Kalshi News, Front Office Sports, CCN, Fox Business.

[34] Polymarket fee structure: 0.0625 × Price × (1-Price) on subset of markets; 0.10% on Polymarket US; daily revenue ~$1M post-March 30 update — KuCoin, DeFiRate.

[35] Ukraine mineral deal $7M dispute March 24-25, 2025; UMA whale 5M tokens / 3 accounts cast 25% of votes; UMA transitioned to MOOV2 with 37-address whitelist — The Defiant, The Block, Coindesk, web3isgoinggreat.

[36] NY AG Letitia James sued Coinbase Financial Markets and Gemini for "illegally" running gambling operations through prediction markets — TipRanks, Cryptonews.

[37] Polymarket 2026 midterm odds: Democratic House 85.5%, Democratic sweep 50.5%, R Senate / D House 36.5%, GOP sweep 12.5% — Polymarket "Balance of Power 2026 Midterms" market.

[38] Coinbase prediction markets to all U.S. customers January 27, 2026 — Coindesk.

[39] Fanatics Markets Phase 1 + Phase 2 expansion; Combos parlay feature — Fanatics IR, Crypto.com, GlobeNewswire, US iGaming Hub.

[40] ForecastEx Interactive Brokers macro event contracts focus — ForecastEx, Legal Sports Report.

[41] Susquehanna 2024 Kalshi market making disclosure; LedgerX majority stake with Robinhood; "day-one liquidity provider" at Rothera — Bloomberg, Yahoo Finance.

[42] "Big Three" prop firms (SIG, Jane Street, Jump) prediction-market dedicated desks; DRW second tier — FinanceMagnates, Sandmark, Currency Analytics.

[43] Paradigm prediction-market trading terminal — DeFiRate.

[44] Goldman Sachs / Tradeweb institutional adoption; macro/CPI focus; margin trading 10× exposure — Sandmark, FinanceMagnates, BNP Paribas.

[45] Polymarket cap table: Founders Fund / Vitalik / General Catalyst / ICE / 1789 Capital / Jump Trading; reported $400M at $15B (April 2026) — Tech Funding News, Fortune.

[46] Mike Selig background: 16th CFTC Chair sworn in December 2025; SEC Crypto Task Force chief counsel + senior advisor to SEC Chairman Paul Atkins; clerked for CFTC Commissioner Christopher Giancarlo (2014); Willkie partner 2022-2025; FSU undergrad, GW Law — CFTC press release, Wikipedia, WilmerHale, Dentons Crypto, Reed Smith, FinanceMagnates.

[47] Selig "CFTC won't defer to courts on prediction markets" — Sportico.

[48] Manifold Markets ~200K active markets; Mana virtual currency; Sweepstakes Cash mode introduced 2025 — CB Insights, Manifold.

[49] Adjacent News API: 40K+ markets cross-platform; AI-powered NL queries; news from 5K+ sources; Auth0 JWT — Adjacent News docs, GitHub Awesome Prediction Market Tools.

— Patrick Liu, April 27, 2026