Getting filled, and what the fill costs.

A market shows a best bid and a best offer — say 47 bid, 49 ask. Those are real orders someone has already placed. You can trade against them immediately by paying 49 to buy, which is a taker order, or you can add your own order at 47 or 48 and wait for someone to trade against you, which makes you a maker.

A market order simply takes whatever the book offers, which is fine when the book is tight and expensive when it is not — it will walk up through resting offers to fill your size. A limit order names the worst price you will accept. Practically every execution decision on the exchange is some version of this trade-off: certainty of getting filled against the price you get filled at.

Takers pay the full fee, 0.07 × p × (1 − p) per contract, and pay the spread by crossing it. Makers pay a quarter of that fee and, if filled, collect the spread rather than paying it. On a 50¢ contract in a one-cent market that is roughly 1.75¢ plus half a cent for the taker against roughly 0.44¢ less half a cent captured for the maker — a swing of about 2.3¢ per contract on the same view.

That difference is not free money, and treating it as such is the standard way to lose it. A resting order is an option you have written for the rest of the market: it gets exercised when someone wants the other side, which is more often when they know something or when the price is about to move through you. The maker discount is compensation for that, not a discount on the same trade.

Every Kalshi market has a YES and a NO side, and their prices sum to $1. A market where YES trades at 60¢ has NO at 40¢, necessarily — if it did not, buying both sides for less than a dollar would lock in a profit, and someone would.

The consequence trips up nearly everyone at first: buying NO at 40¢ is the same trade as selling YES at 60¢. Same exposure, same payoff, same risk. There is one book, not two, and the NO quotes are the YES quotes read from the other end. When you cannot get filled on one side, look at the other — you may find the size you wanted sitting there, described differently.

This is also why “shorting” on a prediction market carries no special risk. Selling YES is buying NO, and the worst case on a contract bought at 40¢ is the 40¢. There is no margin call and no unbounded loss, which is a real structural difference from shorting an equity.

The quote is only good for the size resting behind it. A market showing 49 ask might have 20 contracts there and the next 200 at 52 — so an order for 220 contracts does not fill at 49, it fills at an average well above it. On thin markets this slippage is routinely larger than the fee and the spread combined, and it is invisible until the fill comes back.

The practical reading is that a displayed price is a price for a size, and the size matters as much as the number. It is also why a large order broken into pieces over time behaves differently from the same order sent at once — and why reading the tape is mostly the study of how size arrives.