👋 Hey, welcome to Predicted, the newsletter covering the business of prediction markets.
Forwarded this email and want it in your inbox? Hit the subscribe button below! 👇
Discussed in this piece:
A record for US venues
Why volume gets politics wrong
The biggest political markets
Where did the money go?
Were the markets right?
What we are watching
This piece was written and researched by Omar El Safy and edited by Pet Berisha.
———————————————————————————
On Monday, we published Kalshi Raise $1 Billion at a $40 Billion Valuation. Check it out!
———————————————————————————
On the night before the midterms, US prediction markets held $1.0B in open political contracts.
That is the most they have ever held on an election, about 30% more than the $778M at the peak of 2024.
You wouldn't know it from the coverage.
On election day, Bloomberg reported that political betting had slumped to about 1% of October 2026 volume on Kalshi and Polymarket, down from roughly two-thirds in 2024.
On volume, they are right, as politics was 1.3% of the total notional volume traded on Kalshi from 1 October 2026 to election night.
But volume is the wrong way to measure politics. Volume counts every time a contract changes hands, and open interest counts the money sitting in contracts that haven't settled.
A sports bet is often placed and settled in an evening, so it shows up as volume. A Senate contract is bought in the spring and can be held until polling day, so open interest is likely a better measure.
Measured that way, the midterms were the biggest political event US prediction markets have ever had.
Fig. 1. Political open interest on US prediction markets, daily, September 2024 to 2 November 2026. Source: The Prediction.
01 - A record for US venues
Political open interest across the US venues we track hit $1B on 2 November 2026. The previous high was $778M the day after the 2024 election.
Kalshi held most of the new record, with $838M on election eve, up from $279M before the 2024 vote. Kalshi were only allowed to list US election contracts from October 2024, after winning their court case against the CFTC (Commodity Futures Trading Commission), so their 2024 election markets ran for just five weeks.
Kalshi's political open interest averaged $22M a day in June 2025, $131M in March 2026 and $521M in October 2026.
Two smaller venues also had their first big election.
Polymarket US went from a $5.2M daily average in Q3 2026 to $108.6M on election day, in the first federal election where Polymarket was legal for Americans.
Rothera went from $6.1M to $113.7M, more than Polymarket US. That figure is at face value, and because Rothera's book leans towards long shots, it overstates the money at stake more than anywhere else.
Polymarket International, the offshore exchange, held far less this time. On the eve of the 2024 vote its presidential market held $248M, more than twice the $111M in Kalshi's. On the eve of these midterms, its House and Senate markets held $25M, and Kalshi's held $103M.
Count every venue and 2024 probably still held a little more political money in total, roughly $1.1B to $1.3B compared to about $1.1B now.
For the first time, most of it sat on US-regulated venues.
In 2024 the US political money sat on two venues, with $279M on Kalshi and $204M on ForecastEx, owned by Interactive Brokers ($IBKR ( ▼ 6.25% )), the night before the vote. This time Kalshi held $838M, Rothera $82M and Polymarket US $80M, while ForecastEx was down to $6M.
Fig. 2. Open interest in the headline election markets on election eve, Kalshi and Polymarket International, 2024 presidential winner and 2026 House and Senate control. Source: The Prediction.
Fig. 3. Political open interest on the eve of each election, by US venue, compared to Polymarket International's headline election markets (the 2024 presidential winner, and 2026 House and Senate control). Source: The Prediction, Kalshi, Polygon blockchain.
That money now sits on venues Robinhood and Coinbase customers can reach from their apps, and Kalshi's prices are the ones S&P Dow Jones Indices signed up to build indices from.
Fig. 4. Political open interest by venue, Q3 2026 daily average compared to the 1 to 3 November 2026 peak. Source: The Prediction.
02 - Why volume gets politics wrong
Politics is a bigger share of the money held than of the money traded on every US venue that lists it. On Kalshi it was 32.7% of open interest and 1.3% of volume.
Fig. 5. Politics' share of notional volume and of open interest, by venue, 1 October 2026 to 3 November 2026. Source: The Prediction.
The reason is turnover: how many dollars trade for each dollar held. In October 2026 Kalshi's crypto markets turned over $427 for every dollar held and sports $35, while politics turned over $1, so a dollar put on an election mostly sits there.
Fig. 6. Notional volume traded for every $1 of average open interest, by category, Kalshi, October 2026. Source: The Prediction.
That makes a political contract closer to a long-dated option than a bet, and Kalshi's 2028 presidential markets already hold $270M, two years before anyone knows the answer.
03 - The biggest political markets
The biggest political market ever by volume is Polymarket's 2024 presidential winner, at $3.69B traded. It never held more than $298M, so it traded about twelve times more than the held amount.
On Kalshi, the midterm markets held $504M on election eve, more than their 2024 presidential markets held on election day, at $387M.
Fig. 7. Kalshi's largest political markets by open interest, the 2024 presidential on election day, the 2026 midterms on election eve and the 2028 presidential on 2 November 2026. Source: The Prediction.
Polymarket's 2028 Democratic nominee market has traded $1.31B, five times Kalshi's, but holds $12M, while Kalshi's holds $177M.
Fig. 8. Open interest in the 2028 Democratic presidential nominee market, Kalshi and Polymarket International, daily, July 2025 to November 2026. Source: The Prediction.
04 - Where did the money go?
The Texas Senate race alone drew $187M of trading on Kalshi and held $126M, more than the Senate control market.
Fig. 9. Kalshi's ten largest 2026 midterm markets, lifetime volume and open interest at settlement. Mississippi Senate still uncalled. Source: Kalshi.
Across Kalshi's ten biggest midterm markets, about two-thirds of what traded was still held at settlement, which means people bought and held to the result.
The Senate control market held about $200,000 a year out and $21.8M on election eve, while its price for a Democratic Senate climbed from 28% to 76%.
Kalshi earn fees when contracts trade, and in October 2026 every dollar held in sports traded $35 of volume, compared to $1 for every dollar held in politics. So, per dollar held, politics earns them a small fraction of what sports does.
Fig. 10. Kalshi House and Senate control markets, price of a Democratic win (left) and open interest (right), November 2025 to 2 November 2026. Source: The Prediction.
05 - Were the markets right?
Democrats took the Senate and the House, and both venues had them as favourites. At 6pm ET on election day Rotowire recorded Democrats at about 94% for the House and 75% for the Senate on Kalshi and Polymarket alike.
Favourites won almost every race, 479 of 487 on Kalshi and 419 of 426 on Polymarket.
Covers' tracker had Republican Darline Graham at about 70% in South Carolina's Senate race and Greg Abbott at around 60% to 65% in Texas, and both lost to Democrats.
Does more money make a market more accurate? Kalshi's own calibration study of 2.24M markets says it does. Markets with over $200,000 traded scored about six times better on the Brier score (a measure of forecast error where 0 is perfect) than markets under $10,000.
We checked the midterms ourselves, scoring 1,339 Kalshi markets on their price the night before. If anything, they were too cautious, as contracts priced between 70¢ and 90¢ came in 88% of the time.
Fig. 11. Kalshi's 2026 midterm markets, price the night before compared to how often the outcome happened, 1,339 markets grouped by price. Source: The Prediction.
Among the close races, priced between 20¢ and 80¢, the markets holding over $1M were more accurate than those holding under $100,000, with Brier scores of 0.140 and 0.183. That is 28 markets in one election, so treat it as a pointer.
Fig. 12. Brier score of contested Kalshi midterm markets (priced 20¢ to 80¢ the night before), by open interest on election eve. Lower is more accurate. Source: The Prediction.
Across 964 polls between 1988 and 2004, the Iowa Electronic Markets were closer to the result than the poll 74% of the time, and their lead was biggest more than 100 days before the vote.
In Brazil, Polymarket priced Lula at about 20% before he won on 25 October 2026, in a small market priced mostly from outside the country.
Our read is that the more money is held in a market, the more accurate its price tends to be.
06 - What we are watching
Whether political money stays between elections. Politics is already about a third of the money held on Kalshi, and the 2028 presidential markets hold $270M two years out.
Whether the money makes the markets deeper. We are watching whether more open interest brings more buyers and sellers at every price, so a fund could place a big order without moving it.
Who it attracts. A contract held for a year rewards anyone who knows the result early, and the Pentagon has already banned its staff from betting on what they learn at work.
Political markets outside the US. Polymarket International traded $397M on non-US elections in Q3 2026. We are watching whether any of them start holding money the way the US markets now do.
———————————————————————————
Disclaimers
This newsletter is for informational purposes only and is not financial, business or legal advice. These are the author's thoughts & opinions and do not represent the opinions of any other person, business, entity or sponsor. Any companies, platforms, markets or projects mentioned are for illustrative purposes unless specified.
The contents of this newsletter should not be used in any public or private domain without the express permission of the author.
The contents of this newsletter should not be used for any commercial activity, for example - research report, consultancy activity, or paywalled article without the express permission of the author.
Please note, the services and products advertised by our sponsors (by use of terminology such as but not limited to; supported by, sponsored by or brought to you by) in this newsletter carry inherent risks and should not be regarded as completely safe or risk-free. Third-party entities provide these services and products, and we do not control, endorse, or guarantee the accuracy, efficacy, or safety of their offerings.
When you trade on prediction markets (including event contracts, opinion markets and other speculative instruments) your capital is at risk. Risks include price volatility, loss of capital, illiquidity, complexity, evolving regulation and lack of protection. Many prediction market operators do not currently operate in a fully regulated industry, and availability varies by jurisdiction. You may not be protected under financial compensation schemes typically afforded to investors dealing with regulated entities. Nothing in this newsletter constitutes a recommendation to place, hold, or close any position on any market.
