Kalshi’s market makers made big profits at takers’ expense during last week’s midterm elections, according to an InGame analysis, with two Texas statewide races among the biggest wins for market makers and loses for takers. Millions more went in market makers’ direction on markets for House and Senate control, as well as Senate elections in Michigan and Maine.
The successful performance for market makers, at the expense of takers, may be a datapoint in the debates over the performances of prediction markets and models during the midterms.
While prediction markets had a significant amount of success in forecasting races, it was market makers who had the most success in doing so, while the typical Kalshi users may not have been so predictive.
Takers lost $44.6 million after fees, while market makers made $34.9 million in profit, across all midterm markets.
Market makers win big on Texas Senate
On the Texas Senate market, where Democrat James Talarico defeated Ken Paxton, Kalshi’s takers lost $13.6 million before fees, or $15.3 million after.
Unlike sports markets, Kalshi does not charge market maker fees for politics. Therefore, taker losses before fees are exactly equal to market maker wins before or after fees.
Kalshi’s market for the Texas Senate race was the biggest win for market makers dating back to the start of 2026, when InGame’s records begin. As the 2024 presidential election was believed to be a win for takers, and few other non-sports events have attracted similar volume, it may have been the biggest ever non-sports win for market makers.
Among all markets, including sports, it was the fifth-biggest win for market makers since InGame began keeping records, only behind four FIFA World Cup markets (including the World Cup winner, which attracted $1.51 billion in volume) and the July UFC fight between Conor McGregor and Max Holloway.
Big block trades on Texas governor lose
The next biggest profit for market makers was the Texas governor’s race, where takers lost $10.8 million before fees or $11.7 million after fees as Democratic challenger Gina Hinojosa defeated incumbent Greg Abbott. However, this market may have been affected by a series of large block trades, all on Abbott.
Block trades are negotiated off the exchange and then reported to Kalshi. They are limited to financial institutions, regulated insurers, investment funds, and entities with more than $10 million in assets, and they allow parties to take large positions without moving the public price.
There were two block trades on the election in late October, with the taker putting up $1 million to win an additional $500,000. That was then followed by five more block trades made on the day of the election, after votes had started being counted and Hinojosa became the favorite, in which the taker put down $2 million to potentially win another $6.1 million on top of the original stake.
All seven block trades were on Abbott. It is not clear if they were all placed by the same individual or group, or placed by unrelated parties.
There were a few taker wins
The market for control of the House cost takers another $5.3 million before fees, or $6.1 million after. The Senate control market added a further $3.6 million in losses before fees, or $4.3 million after fees. Takers lost an additional $3.2 million before fees, or $3.7 million after, on Michigan’s Senate race, and another $1.8 million before fees, or $2.1 million after, on the Senate race in Maine.
Takers also lost $4.4 million before fees, or $5.7 million after, on the Los Angeles mayoral race, though many of these losses were from bets placed during the primaries.
The main exception to the trend of market makers winning was in South Carolina, where Annie Andrews unexpectedly defeated Darline Graham for the Senate seat previously held by Graham's brother before his death. Takers won $3.3 million before fees, or $3 million after fees.
In the Florida governor’s race, where Republican Byron Donalds defeated David Jolly in a closer-than-expected contest, takers also won big, netting $2.9 million before fees, or $2.4 million after fees.
Senate races in Alaska and Mississippi remain uncalled, and their Kalshi markets remain open, with just under and just over $40 million, respectively, traded. In both races, Democratic candidates are heavily favored by Kalshi to win, with odds above 95%. If Mary Peltola wins in Alaska, as expected, takers would lose a further $1.4 million.
The Mississippi race would be a modest win for takers, not enough to be one of the 500 biggest taker wins of the year. Most other open markets would be modest wins for market makers.
Parlay performance
Kalshi began offering political parlays in October, in the lead-up to the election.
Despite market makers having less historical data to price correlations than with sports markets, it appears that they were still able to profit.
Takers on parlays with at least one non-sports leg lost $1 million on net Tuesday, including $70,000 in fees. Taker losses were 19.6% of stakes, a larger percentage than Kalshi parlay takers typically lose.
These figures include crypto parlays, but these were only a small share of total parlay volume.
Record Kalshi volume
The combination of the election and an already busy sporting calendar helped ensure that Kalshi achieved yet another record week in volume and fees.
In the seven days to Nov. 9, Kalshi’s volume hit $29.7 billion. The vast majority of that, however, was from sports and sports parlays. Parlay volume was $22.1 billion and straight sports was another $4.3 billion. Politics volume was $965.5 million, which was also less than crypto straight bets.
The business made $134.5 million in trading fees. If that average could be maintained over a full year, Kalshi would make $7 billion in fee revenue. Of last week's total, $4.4 million came from politics.
Kalshi taker-side volume for the week exceeded $6.4 billion, or $4.4 billion if likely parlay cash-out trades are excluded.
