Kalshi offers up to $1 million in grants for research on midterm prediction markets
Kalshi will fund up to 50 research projects on how prediction markets behaved in the 2026 midterms, with applications due November 18.
Kalshi opened a research grant program on November 10, 2026, offering up to $1 million across as many as 50 projects that study prediction markets during the 2026 midterm elections. The program page says the exchange will provide any publicly available Kalshi data a project needs.
How it works
Proposals of up to two pages are due by 11:59 p.m. ET on November 18. Applicants can also enter by quoting Kalshi’s announcement post on X with a research idea by November 18, then completing the full application by November 20. Judging runs from November 21 to 26, and Kalshi expects to notify grantees on or about November 27.
Each proposal must set out the research question, why it matters, the data and methodology, and the public output it will produce, such as a paper, dataset, dashboard or code. No Kalshi account, deposit or trading is required to apply, and grants are subject to official rules and a grant agreement.
What Kalshi wants studied
The page lists questions Kalshi is thinking about, while leaving room for others:
- how quickly prices reacted to early returns and race calls on election night
- how market prices and polling averages moved against each other during the campaign
- what markets showed in House and Senate races with little or no polling
- how well market probabilities matched results across race types
- which debates and campaign events moved prices, and for how long
- how trading activity differed across races and states
Why it matters
Kalshi has already published a paper on the calibration and accuracy of its midterm markets. The grants open those questions to other researchers working with Kalshi’s data, and because every proposal must name a public output, funded projects would put more of that analysis on the record.
Our AI news desk researched and wrote this article, and software checked every factual sentence against the records it links to, following the WriterOps methodology. Report an error.