Kalshi’s Tradable Midterms Index Settles to Pyth Values

Kalshi, a U.S. exchange regulated by the Commodity Futures Trading Commission, has listed contracts on KPOW Midterms, a gauge of market expectations for the Nov. 3 U.S. elections, with Pyth Network calculating and publishing the values that determine payouts.

Kalshi Research said Tuesday that the product was “Now tradable,” calling it the exchange’s first tradable political index. Pyth said the index is sourced and published exclusively by Pyth and “every contract on it settles to Pyth’s values.”

The contracts let traders take positions on a basket of election expectations rather than the winner of a single race. Under Kalshi’s contract terms, a race call or certification does not itself determine the payout: the published settlement reading governs.

What Goes Into the Index

KPOW Midterms combines 13 election events. House control carries a 20% weight and Senate control 25%. Five governor races—Ohio, Nevada, Florida, Texas and Georgia—and six Senate races—Texas, Michigan, Maine, Ohio, Kansas and Iowa—each carry 5%.

The scale runs from −50 for certainty of Democratic victories across all weighted races to +50 for Republican victories. Each event’s score is 50 times the difference between its Republican and Democratic contract prices; the index adds those scores according to their weights. A +10 reading means roughly a 60% weighted average implied Republican win probability, not a 60% chance in every race.

Pyth calculates the index using Kalshi’s public trades and Kalshi’s code, publishing a new value every 30 seconds. It is narrower than the broader American Power Index, which also incorporates current officeholding, the presidency and government-shutdown risk.

What Determines a Payout

Contracts compare the settlement reading with their specified threshold or range. A winning contract pays $1 and the losing side receives nothing, according to Kalshi’s product filing.

The default reading is a time-weighted average over the 15 minutes ending at the contract’s specified reference time—not the last index update. Certain contracts listed on a 15-minute basis for Nov. 3 can instead use a 60-second window, but only if that window was specified when listed.

For example, an active contract in Kalshi’s public API resolves Yes if the 15-minute average ending at noon ET on Nov. 4 is at least +5.0.

The calculation also filters abrupt price changes. A trade more than five cents from a market’s last accepted price is held until the price comes back, the opposite-party market confirms the move within 60 seconds, or both markets keep trading at the new levels for five minutes. A held move does not enter the index before acceptance.

The reference time does not guarantee immediate payment. The rules allow expiration up to a week after the covered period, with settlement no later than the following day unless the outcome is under review.

If an underlying election market stops trading, its last accepted price remains in the index—even after the race has been decided.