What the price really pays

A Kalshi price is a probability, but not always a fair one. We checked every settled market: how often each price really won, and what it paid after fees. Cheap longshots lose most of the money put into them, and the traders who take liquidity pay the most.

What a 1-10 cent longshot returned

Market-order traders, all prices

Limit-order traders, all prices

What a 90-99 cent favourite returned

Calibration and returns by price

How often each price wins

Dots on the dashed line would be a fair price. Below it, the price was too high.

Return after fees, by price

Per contract bought at that price, held to settlement.

MakersTakers

Price	Contract prices	Won	Return before fees	After fees	Makers after fees	Takers after fees

Makers buying favourites, by category50-97c, return on capital, out of sample

Category	Return on capital	Strength (t)	Events	Fills	Capital traded

Limit orders buying favourites (50-97 cents), by category, after fees, weighted by the dollars at risk. A strength (t) above about 2 is unlikely to be chance for one category alone, but with this many categories a few will cross it by luck. Sports is where most of the money trades.

Measured on real settlements

A bias that held in the past can shrink as more traders lean against it.

How the prices are sampled

Each contract's last trade on each of its final days, both sides of every market, fees at Kalshi's current schedule - the construction of the published study (Burgi, Deng and Whelan, 2025), extended to markets it never saw.

Data

Kalshi's public trade data (the open kalshi-trades dataset, CC BY 4.0), cross-checked against Kalshi's own API. Combination markets are excluded. Not financial advice.