How to read Kalshi Up/Down odds and fees

Kalshi Up/Down contracts look like simple coin flips priced in cents. This TickTilt guide shows how to read those cents as implied probabilities, how the published taker fee works, and why break-even is a little worse than the sticker price. Pair it with the live lean cards and the Scorecard—leans are probabilities, not certainties.

Information only, not betting or financial advice. Always confirm fees on Kalshi's current schedule.

Price as an implied probability

Each contract pays $1.00 if that side wins and $0 if it loses. Buying Up at 70¢ means you risk 70¢ to make 30¢ if Up wins. The market is roughly saying "about a 70% chance Up wins." Down near 30¢ says the opposite. Up ask + Down ask usually adds to a bit more than $1 because of the bid–ask spread.

"Implied" matters: the price is the crowd's current estimate, not a fact about the future and not advice.

The taker fee formula

Kalshi's general taker fee (orders that take liquidity) is often written as:

fee ≈ 0.07 · C · p · (1 − p)
then rounded up to the next cent per order.
C = number of contracts, p = price in dollars (so 95¢ → p = 0.95).

The fee peaks near 50¢ and shrinks toward 0¢ or $1. Maker orders that rest and fill later are often fee-free on many series—check the live schedule and any series multiplier.

Worked examples (1 contract, taker)

Price p	Raw fee 0.07·p·(1−p)	Rounded up	Total cost	Break-even win rate
50¢	$0.0175	2¢	52¢	~52%
80¢	$0.0112	2¢	82¢	~82%
90¢	$0.0063	1¢	91¢	~91%
95¢	$0.0033	1¢	96¢	~96%
97¢	$0.0020	1¢	98¢	~98%

Example A. You take 1 Down at 95¢. Fee rounds to 1¢. Cost = 96¢. Win pays $1 → profit 4¢. Lose → −96¢. One loss wipes out about 24 such wins.

Example B. You take 10 contracts at 90¢ in one order. Raw fee = 0.07 · 10 · 0.90 · 0.10 = $0.063 → rounds to 7¢ total (~0.7¢ each). Rounding hurts less on larger clips.

Example C. At 50¢ the fee is largest. The same formula that looks tiny at 97¢ costs real money in the middle of the book.

Connecting fees to TickTilt leans

A Strong lean on TickTilt (≥~95%) can line up with a high Kalshi ask. If the ask is already 96–98¢, even a lean that is often right may only match break-even after fees. Use the Scorecard to see how often Strong leans held in recent samples, and use this fee math to see what "often enough" would have to mean at the price on screen. No signal is ever 100% certain.

Maker vs taker, in one minute

A taker hits a resting order and usually pays the fee above. A maker posts a limit that sits on the book; on many Kalshi series, makers pay $0 when that order later fills. The trade-off is speed: makers may not fill at all before the round ends. Always read the series fee line on Kalshi—multipliers and exceptions exist.

For learning from TickTilt alone, you do not need maker/taker detail. It only matters if you compare a lean on the card to a real fill price you might see on the exchange.

Spread is a fee you do not see on the schedule

If Up's best ask is 96¢ and the best bid is 94¢, buying Up costs 96¢ even before Kalshi's taker fee. Selling would get 94¢. That 2¢ gap is the spread. Near the end of a quiet round, spreads can widen. A TickTilt Strong lean does not remove spread risk; it only describes price-versus-target odds in TickTilt's model.

Putting a live card next to the book

Open the live cards. Suppose BTC shows Strong lean Down at 97% with two minutes left, and Kalshi's Down ask is 96¢. After a 1¢ rounded fee, total cost is about 97¢—close to break-even if the lean's historical hit rate (see the Scorecard) is similar. If the ask is 99¢, the math is harsher even when the lean looks loud. That comparison is educational: it shows why "often right" and "profitable after costs" are different questions.

TickTilt does not place orders, hold funds, or recommend sizes. Event contracts can lose your full stake. Eligibility and rules are Kalshi's.

Quick checklist

Read price as roughly implied probability.
Add the rounded taker fee (and any series multiplier).
Remember one loss at 95¢+ wipes many small wins.
Check Scorecard sample sizes—early data can look too clean.
Confirm the live fee schedule before relying on any example here.