Prediction markets have become a money machine bolted to a political calendar: more than $1.4 billion traded on contracts tied to the 2026 midterms, and Robinhood turned that flow into $173 million of revenue in a single quarter. The regulatory file that decides how much of it is durable was rewritten on Oct. 9, and the comment window on it closes Nov. 19, sixteen days after the votes. Here is where the line item actually sits, who gets paid in either branch, and what the midterms do and do not decide.
The election is the wrong clock
The CFTC Already Wrote the Rulebook — the Midterms Only Decide Who Signs the Comment Record
The most consequential document in prediction markets was published on Oct. 9, and it was not a bill. The Commodity Futures Trading Commission issued an interim final rule excluding casino-style gambling products from the definition of a swap and simultaneously proposed defining sports, politics, cultural and weather event contracts as swaps. So the election no longer decides whether the rulebook exists; it decides who writes the record that finalises it and whether a new Congress reopens the file in 2027.
Midterms volume on prediction markets
$1.4B
Traded on2026 midterm contracts, ahead of Nov. 3 (CNN analysis, Oct. 23, 2026)
CFTC comment deadline
Nov. 19, 2026
30-day comment period on the swap-definition proposal, after the midterms
Event contracts as share of transaction revenue
20%
$173M of $867M transaction-based revenue in Q3 2026
Prediction-market bills in Congress
~30
Introduced since 2025, mostly by House Democrats; none has passed
AI PAC general-election spending
$8M vs $6.9M
Leading the Future vs Public First Action, through Oct. 20, 2026
The rulebook
The Oct. 9 Package Splits the Market in Two: Federal Supremacy for Binary Contracts, States for the Wager
| Instrument | What it does | Status | Date that matters |
|---|---|---|---|
| Interim final rule: 'swap' excludes casino-style gambling products | Takes wagers placed with sportsbooks and casinos out of the swap definition using a provider test (state or tribal licence) and a product test (not inherently tradable on an exchange) | In force immediately | Comment period closes Nov. 19, 2026 |
| Proposed rule: 'swap' includes event contracts | Would define event contracts based on sports, politics, cultural and weather events as swaps, pre-empting state gambling law for listed venues | Proposed, comment open | Final rule lands under a new Congress |
| Proposed Rule 40.11 (June 12, 2026) | Public-interest review of contracts tied to unlawful activity, terrorism, assassination, war and gaming; 10-day trigger, 90-day clock, 100-day listing deadline | Proposed | Every listing starts its own clock |
| Sixth and Ninth Circuit rulings | Sixth Circuit: sports outcomes too far afield to be a swap; Ninth Circuit: Rule 40.11 bars gaming contracts from designated contract markets | Decided | New Jersey petitioned the Supreme Court on Sept. 2, 2026 |
The two documents pull in opposite directions on purpose. The proposal would treat binary event contracts as federal swaps, which pre-empts state gambling law for the listed venues; the interim final rule pushes traditional parlay-style wagers back to state gaming regulators. That is a bargain, and itpulls in opposite directions on purpose: Robinhood and DraftKings get federal protection for binary contracts while the state books keep the parlay business they already run.
- Robinhood routes event contracts across Kalshi, Rothera, ForecastEx, Crypto.com and OG, so one venue's licence no longer decides its revenue.
- A sports carve-out would erase the federal claim to binary sports contracts — the exact product the Ninth Circuit already called a bet.
- State sportsbooks keep the wagers the interim final rule sends back to them, protecting parlay handle while capping the arbitrage.
- The comment record closes sixteen days after the midterms, so the composition of Congress on that date shapes part of the rule's final record.
The revenue
Robinhood Converted a Venue Licence Into a $173 Million Quarter
Event contracts produced $173 million of revenue in the quarter ended Sept. 30, 2026, up more than sixfold year over year, and now out-earns both crypto trading at $120 million and equities at $121 million. That makes prediction markets the third-largest transaction line at a crypto brokerage — and the one line whose value depends on how a rulemaking ends. Contracts traded hit a record 16.4 billion, up more than sevenfold, with the Rothera venue contributing more than $100 million of annualised revenue in its first full quarter.
Robinhood transaction-based revenue, Q3 2026 ($M)
Transaction-based revenue of $867 million in Q3 2026, up 19% year over year (10-Q for the quarter ended Sept. 30, 2026).
Unit: $M
Options
399
Event contracts
173
Equities
121
Cryptocurrencies
120
Other
54
| Platform | Scale marker | Position |
|---|---|---|
| Kalshi | More than 90% of U.S. prediction-market activity; annualised volume tripled from $52B to $178B | $1B Series F at a $22B valuation (May 7, 2026); a further round near $40B reported Sept. 30, 2026 |
| Polymarket | More than $340 million traded on midterm markets | U.S. platform plus an offshore site that sits outside CFTC registration |
| Robinhood Markets | $173 million of event-contract revenue; 16.4 billion contracts traded in Q3 2026 | Routes across Kalshi, Rothera, ForecastEx, Crypto.com and OG; 28.7 million funded customers |
| DraftKings | Third by notional trading volume in the week ended Oct. 23, 2026 | DKeX exchange live, with $200M–$300M of 2026 investment behind it |
The AI money
The AI PACs Are Not Split Over Safety — They Are Split Over Preemption
| Camp | Vehicle and backers | Raised | General-election spend through Oct. 20, 2026 | Policy ask |
|---|---|---|---|---|
| Pro-deployment | Leading the Future and its Think Big affiliates — Andreessen Horowitz, OpenAI president Greg Brockman, Perplexity, Ron Conway | $125M | About $8M | One broad national framework that preempts state AI law |
| Safeguards-first | Public First Action (Anthropic-aligned), plus Guardrails Alliance, which raised $11.7M | $80M | About $6.9M | Keep state authority; preemption only if Congress writes a comprehensive federal rule |
Both camps want guardrails; they disagree about who writes them. Leading the Future, funded by Andreessen Horowitz, OpenAI president Greg Brockman and others, wants a single national standard that preempts state AI law, and it backed 25 of the 28 primary candidates it supported. Public First Action, aligned with Anthropic, has raised $80 million and resists pre-emption, arguing states should keep authority unless Congress writes a comprehensive federal rule. The prediction-market fight is the mirror image of that argument: who gets to write the rules when federal agencies and state regulators disagree.
AI-aligned super PAC general-election congressional spending ($M)
Federal election filings through Oct. 20, 2026; both networks had raised well over $100 million for the cycle.
Unit: $M
Leading the Future
8
Public First Action
6.9
- Through the primaries the two networks put at least $44 million behind 40 House and Senate candidates, and the pro-deployment camp won 25 of the 28 primaries it entered.
- New York state's RAISE Act is the live test case, and the pre-emption question turns into an election question whichever network's candidates win.
- The stakes are large relative to the spending: $55.7 million of AI-aligned cycle advertising, $28 million of it for Democrats and $19 million for Republicans, so the camps are buying leverage rather than policy.
- Anthropic's $20 million went to Public First Action for public education rather than political activity, which keeps the company formally on the sidelines.
The AI leg
For the Chips, the Election Is Second-Order — For the Rules, It Is First-Order
The deployment decisions the AI PACs are fighting over show up in chip orders, and those numbers are already extreme. SK hynix reported revenue of79.3 trillion won for the June 2026 quarter, up 257% year over year, at an operating margin of 76%, and Advantest posted quarterly revenue of 367.5 billion yen with a 51.5% operating margin. None of that depends on who wins in November; all of it depends on AI deployment staying fast, which is exactly what the pre-emption fight decides.
- The trade-deal layer is where the election is loudest: Japan's $550 billion U.S. investment package and South Korea's $350 billion pledge are executive promises attached to 15% reciprocal tariffs, and a new Congress can reopen both.
- The test-equipment chokepoint is where the money actually lands, and no ballot removes the need to test advanced packaging and HBM.
- The honest read on the AI supply chain is that pre-emption changes the slope of deployment, not its direction, which is why the chip names trade on demand whilethe labs and their PAC proxies trade on rulemaking.
Supply chain
Who Gets Paid in Either Branch
The rulebook splits the industry but it does not split the cash. Settlement data, venue infrastructure and order routing are needed whichever side wins, which is why the picks-and-shovels layer is the most reliable way to own this trade — and why the sportsbook layer is the most exposed.
- Genius Sports supplies the official data used to settle contracts for Kalshi and Polymarket and launched Prediction.com on Sept. 17, 2026; June-quarter revenue of $195.5 million was up 65% year over year, though trailing operating losses still run $99 million.
- CME Group lists $1 event contracts across sports, economic indicators, crypto, commodities and financials, and retains a 51% stake in FanDuel Predicts.
- Interactive Brokers routes client orders to Kalshi, CME Group and ForecastEx event contracts, adding a fee-light toll road to a brokerage trading at 13.6x trailing EBITDA.
- Coinbase offers event contracts through Coinbase Financial Markets, giving a crypto venue a regulated product beyond trading volume.
- DraftKings runs a sportsbook and the DKeX exchange: the sportsbook keeps the parlay handle the interim final rule protects, while DKeX bets on federal supremacy holding.
- Flutter Entertainment moved every sports and novelty contract on FanDuel Predicts to Crypto.com's exchange in August 2026, which conceded that the sports line was the contested one.
- Flutter Entertainment's prediction pivot has run behind schedule and cost it its chief executive, with the shares at $78.26 against a 52-week high of $240.04 and $11.0 billion of net debt behind a $13.6 billion market capitalisation — the balance sheet limits patience.
Fundamentals
The Market Is Paying for Proof: 13.6x Sales for Robinhood, 1.8x for DraftKings
| Company | Market cap | EV / TTM sales | P / E (TTM) | Trailing result |
|---|---|---|---|---|
| Robinhood Markets | $74.8B | 13.6x | 43.0x | TTM net income $1.99B |
| DraftKings | $10.3B | 1.8x | n.m. (loss) | TTM operating loss $182M |
| Flutter Entertainment | $13.6B | 1.4x | n.m. (loss) | TTM net margin -4.4% |
| Interactive Brokers | $144.6B | 12.0x | 31.6x | TTM earnings yield 3.2% |
| Genius Sports | $1.7B | 3.0x | n.m. (loss) | TTM operating loss $99M |
| CME Group | $105.5B | 15.3x | 23.9x | TTM earnings yield 4.2% |
The gap is not about growth expectations; it is about who has already converted the rulebook into gross profit. Robinhood earned $1.99 billion of net income over the trailing four quarters on roughly $5.1 billion of revenue, while DraftKings lost $182 million at the operating line over the same window and cut revenue4.6% year over year in the June quarter. DraftKings is funding a third-ranked exchange out of a shrinking sportsbook, and its forecast of $200 million to $300 million of 2026 prediction-market investment lands long before the revenue does.
Horizons
Two Clocks: Days to Quarters, Then One to Three Years
- Nov. 3, 2026: the midterms. Kalshi's own midterms market puts Democrats at 91% for the House and 64% for the Senate — the market most likely to be regulated is pricing its own regulator into power.
- Nov. 19, 2026: the comment record closes on both rules, and the incoming Congress inherits a file written under a different one.
- Early November 2026: DraftKings' third-quarter print, the first test of a full quarter of DKeX and of the 2026 guidance it still has to hit.
- 2027: the final swap-definition rule, the Supreme Court's answer to New Jersey's Sept. 2 petition, and the first federal AI preemption bill — three decisions that together settle who writes American betting and AI rules.
- One to three years: if federal supremacy hardens into statute the listed venues compound, and if the carve-out goes the other way volume migrates offshore to venues no American regulator can reach.
The practical consequence is that the days-to-quarters trade is about the Nov. 19 record and two sets of Q3 numbers, while the one-to-three-year trade is about whether the federal venue model survives the Supreme Court and a new Congress. Pricing a November headline as a permanent change of regime is the mistake; the rulebook is already written, andthe comment record is the only part still being written.
The call
Own the Licence, Not the Argument
Facts: the CFTC has published its rulebook, the comment record closes Nov. 19, and Robinhood has $173 million of quarterly revenue riding on it. Inference: a framework that preempts state gambling law for binary contracts ismore likely to survive a Democratic Congress than to be undone, because the agency, the venues and the state books each want a different piece of it. Speculation, clearly labelled: a Democratic House would more plausibly use its roughly30 bills and the GAO workforce review to reopen the sports carve-out than the political and economic contracts, where the volume and the margin sit.
Where the election beta actually lands
- Event contracts earned $173 million in Q3 2026, up over sixfold, and now exceed crypto at $120 million and equities at $121 million.
- Event contracts now supply 20% of transaction revenue, and the Nov. 19, 2026 comment record decides whether the venues it routes to keep federal supremacy.
- Near term, 16.4 billion contracts traded in one quarter scale straight into revenue; over one to three years, preemption of state gambling law is the structural moat.
- Trailing operating losses of $182 million against $6.2 billion of revenue show the DKeX build is funded out of margin.
- DKeX ranked third by notional volume in the week ended Oct. 23, 2026, but $200M–$300M of 2026 investment lands before the revenue does.
- Mixed forces: DKeX optionality if federal supremacy holds, while the sportsbook base is the asset most exposed to a federal carve-out.
- Over one to three years DKeX must out-run the 1.8x enterprise-value-to-sales multiple now applied to a revenue line that fell 4.6% year over year.
- Routing across Kalshi, CME Group and ForecastEx event contracts makes prediction markets a fee-light addition to a 13.6x-EBITDA brokerage.
- Federal supremacy widens the set of venues it can route to, and a sports carve-out removes sports contracts without touching the economics.
- Near term, nothing in the Nov. 19 record moves earnings; over one to three years, event-contract routing adds commission without balance-sheet risk.
- Lists $1 event contracts across sports, economic indicators, crypto and commodities, and retains 51% of FanDuel Predicts.
- Any rule keeping binary contracts inside CFTC-registered venues concentrates volume in cleared exchanges rather than offshore books.
- At 15.3x trailing sales and a 4.2% earnings yield, the venue is priced like an exchange rather than like a growth bet.
- Settlement data for Kalshi and Polymarket plus the Sept. 17, 2026 launch of Prediction.com make it the data toll on every contract that trades.
- June-quarter revenue of $195.5 million rose about 65% year over year, but trailing operating losses of $99 million keep the equity story unproven.
- Mixed forces: volume growth is the tailwind, platform concentration and continued losses are the offset.
- Moving every sports and novelty contract on FanDuel Predicts to Crypto.com in August 2026 conceded the contested sports line.
- A sports carve-out protects the state-licensed parlay handle; the loss is the federal-contract optionality the prediction pivot was bought for.
- With $11.0 billion of net debt against a $13.6 billion market capitalisation and the shares at $78.26 versus a $240.04 52-week high, the balance sheet limits patience.
- June-quarter revenue of 79.3 trillion won at a 76% operating margin price the AI build-out the PACs are legislating around.
- Watch the catalyst: whether Congress writes federal AI preemption into statute in 2027, and whether the $350 billion Korea investment pledge survives the new Congress.
- Election sensitivity is second-order here; direct exposure is to deployment speed, not to prediction-market rules.
- June-quarter revenue of 367.5 billion yen, up about 39% year over year at a 51.5% operating margin, with test-equipment lead times tight on HBM and AI chip demand.
- Watch the catalyst: disbursement of the $550 billion Japan-U.S. AI and chip investment fund, and federal AI preemption in the 2027 Congress.
- The election changes the slope of AI deployment, not the demand for testing it, and no ballot removes the test bottleneck.
