To break even on a Kalshi BTC 15-minute contract you have to win at least as often as the price you pay, in cents, plus the trading fee. Buy at 96¢ and you need to win 96% of the time before fees. Kalshi's taker fee adds about 1¢ to a small order at that price, which lifts the bar to 97%. At 99¢ a one-contract order pays a 1¢ fee, so its breakeven is 100%: it cannot come out ahead even when it wins.
We checked that arithmetic against 11,537 graded signals from our own Kalshi BTC 15-minute feed, March 24 to October 4, 2026. The feed won 91.21% at a 91.74¢ average entry: 0.53 points under breakeven before fees, and further under after them. Sorted into ten entry-price buckets, no bucket won more often than its price by more than chance explains. One bucket, 90 to 92¢, won less often than its price by more than chance explains.
One formula, one rounding rule, and a term that shrinks toward zero at exactly the prices where most BTC 15-minute trading happens.
As checked on October 4, 2026, Kalshi's published fee schedule (effective February 5, 2026) sets the general taker fee as round up(0.07 × C × P × (1 − P)), where C is the number of contracts and P is the price in dollars, and "round up" means to the next cent. The schedule also says there is no settlement fee. Kalshi's public series record for KXBTC15M lists fee type "quadratic" with a multiplier of 1, which is the standard schedule.
Kalshi's API documentation adds a detail that matters at these prices. The fee is first computed to a millionth of a dollar, and a separate rounding adjustment then moves a retail account's balance to the next whole cent. For a single fill at a whole-cent price, that works out to the fee rounded up to the next cent. Larger orders spread the rounding across more contracts, and Kalshi carries any overpayment across the fills of one order.
The P × (1 − P) term is what makes this worth a page. It peaks at 50¢ and shrinks toward zero at the extremes, so the raw fee on a 99¢ contract is tiny: 0.07 × 0.99 × 0.01 = $0.000693. Rounded up, a one-contract order still pays a full cent, and a full cent is the entire profit a 99¢ contract can make.
| Price paid | Fee, 1-contract order | Breakeven, 1 contract | Fee per contract, 100-contract order | Breakeven, 100 contracts |
|---|---|---|---|---|
| 50¢ | 2¢ | 52.0% | 1.75¢ | 51.75% |
| 70¢ | 2¢ | 72.0% | 1.47¢ | 71.47% |
| 80¢ | 2¢ | 82.0% | 1.12¢ | 81.12% |
| 90¢ | 1¢ | 91.0% | 0.63¢ | 90.63% |
| 95¢ | 1¢ | 96.0% | 0.34¢ | 95.34% |
| 96¢ | 1¢ | 97.0% | 0.27¢ | 96.27% |
| 97¢ | 1¢ | 98.0% | 0.21¢ | 97.21% |
| 98¢ | 1¢ | 99.0% | 0.14¢ | 98.14% |
| 99¢ | 1¢ | 100.0% | 0.07¢ | 99.07% |
The arithmetic is simple. The question is whether any part of a real record clears it.
Every row below is a signal our feed actually sent, priced at the live Kalshi ask when it fired and graded at settlement. Each win rate sits beside the average price paid for it, because a win rate means nothing without its price.
| Entry price | Signals | Won | 95% range | Avg price = breakeven before fees | Breakeven after 1-contract fee | Read (1-contract fee) |
|---|---|---|---|---|---|---|
| 79¢ or less | 1,369 | 66.1% | 63.6 to 68.6 | 68.9% | 70.9% | Below, within chance |
| 80 to 85¢ | 772 | 83.5% | 80.8 to 86.0 | 82.6% | 84.0% | Above before fees, below after |
| 86 to 89¢ | 743 | 85.7% | 83.0 to 88.1 | 87.6% | 88.6% | Below, within chance |
| 90 to 92¢ | 1,039 | 88.5% | 86.4 to 90.3 | 91.0% | 92.0% | Below, beyond chance |
| 93 to 94¢ | 916 | 94.9% | 93.2 to 96.1 | 93.6% | 94.6% | Above before fees, within chance |
| 95¢ | 525 | 93.3% | 90.9 to 95.2 | 95.0% | 96.0% | Below, within chance |
| 96¢ | 809 | 96.7% | 95.2 to 97.7 | 96.0% | 97.0% | Above before fees, below after |
| 97¢ | 898 | 97.6% | 96.3 to 98.4 | 97.0% | 98.0% | Above before fees, below after |
| 98¢ | 1,649 | 97.6% | 96.8 to 98.3 | 98.0% | 99.0% | Below, within chance |
| 99¢ | 2,817 | 99.0% | 98.6 to 99.3 | 99.0% | 100.0% | At breakeven; cannot clear a 1-contract fee |
Testing ten buckets means one of them landing well above or below its price is what chance alone produces now and then, so the bar for "beyond chance" has to be raised. With a correction for ten tests the bar is 2.81 standard errors. The 93 to 94¢ bucket sits 1.59 standard errors above its price, which does not clear it, and its margin shrinks to about a third of a cent per contract after a one-contract fee. The 90 to 92¢ bucket sits 2.86 standard errors below its price, which does clear it: in this sample that bucket lost money before fees by more than noise accounts for.
The 99¢ row deserves its own sentence. 2,817 signals, a quarter of the sample, were priced at 99¢ and won 99.0% of the time. Before fees that is exactly breakeven. After a one-contract fee, the breakeven is 100%.
Add the rows back together and price every signal as one contract bought at the ask.
| Across all 11,537 signals | Before fees | After fee, 1-contract orders | After fee, 100-contract orders |
|---|---|---|---|
| Result per contract | −0.53¢ | −1.68¢ | −1.00¢ |
| Return on money staked | −0.58% | −1.84% | −1.09% |
That is the honest headline. A 91.21% win rate at a 91.74¢ average entry describes a market pricing near-certain outcomes about right, and the fee turns a small loss into a larger one. Order size changes how much: the rounding costs small orders the most, and it costs them most at the extreme prices where this market spends much of its time. 58.1% of the signals in this sample were priced at 95¢ or higher.
At high prices a win pays a few cents and a loss costs nearly a dollar.
| Price paid | A win pays | A loss costs | Wins needed to cover one loss, before fees |
|---|---|---|---|
| 90¢ | 10¢ | 90¢ | 9 |
| 95¢ | 5¢ | 95¢ | 19 |
| 96¢ | 4¢ | 96¢ | 24 |
| 97¢ | 3¢ | 97¢ | 32.3 |
| 98¢ | 2¢ | 98¢ | 49 |
| 99¢ | 1¢ | 99¢ | 99 |
Our feed lost 1,014 of the 11,537 signals in this sample, about one in eleven. Losses do not arrive on a schedule. A streak of two or three losing windows at 96¢ wipes out 48 to 72 four-cent wins, and every single window can lose 100% of what was paid for it.
This is a measurement, not a recommendation.
For the earlier study of the same feed, including why near-flat windows behave like coin flips, read What 9,800 Graded Kalshi BTC 15-Minute Windows Show. For how the two short-window Bitcoin venues compare on fees, read Polymarket 5-Minute vs Kalshi 15-Minute. For what a graded signal is and what we log on each one, see Kalshi signals, explained.
Per Kalshi's published schedule, checked October 4, 2026, the taker fee is 0.07 × contracts × price × (1 − price), rounded up to the next cent (price in dollars). On a one-contract order that comes to 2¢ at prices from 18¢ to 82¢ and 1¢ from 83¢ to 99¢. On a 100-contract order it falls to about 1.75¢ per contract at 50¢ and 0.07¢ per contract at 99¢. Kalshi's schedule says there is no settlement fee. Fee schedules change, so check Kalshi's current schedule.
The price you pay in cents, read as a percentage, plus the fee. A 90¢ contract needs a 90% win rate before fees and 91% after a 1¢ fee on a one-contract order. A 96¢ contract needs 96% before fees and 97% after.
Not on a one-contract order. A 99¢ contract pays $1.00 if it wins, a 1¢ profit, and the fee on a one-contract order rounds up to 1¢, so even a win nets nothing. On a 100-contract order the fee is about 0.07¢ per contract, so the breakeven win rate is about 99.07%. In our sample, 2,817 signals priced at 99¢ won 99.0% of the time.
Only relative to the price paid. Our own feed won 91.21% of 11,537 graded signals at a 91.74¢ average entry from March 24 to October 4, 2026, which is 0.53 points below breakeven before fees. A high win rate at a high price is the market pricing a likely outcome, not a sign of profit.
Kalshi's schedule applies its maker fee only to markets it lists for one, and the public series record for KXBTC15M shows the standard "quadratic" fee type with a multiplier of 1. Every price in this study is a taker price, the ask at the moment of the signal. Check Kalshi's current schedule before relying on maker pricing.
Not in the March 24 to October 4, 2026 sample. Its record was below breakeven before fees and further below after them, and no entry-price bucket beat its own price by more than chance. We sell the feed as fast, publicly graded coverage of the windows it signals, not as a source of returns.
Everything above can be recomputed from files you can download.
Every Kalshi KXBTC15M signal our feed sent from March 24 to October 4, 2026 that had a usable price and a settled result: 11,537 signals, the same set and count as our public results log. Graded rows that were never sent to subscribers (3,677), were priced at 100¢ with nothing left to buy (3,099), or failed to capture a price (26) are not counted; those that cleared the alert threshold still appear in the log with the reason noted.
The executable ask at the moment the signal fired: one minus the best bid on the opposite side of Kalshi's live order book, rounded to the whole cent. Signals fired with 104 to 235 seconds left in the window, median 170 seconds.
Graded against Kalshi's official settled result. In 167 signals (1.4%) the official result was not available within 30 seconds and the row was graded from our own exchange price feed. Fees use Kalshi's published taker formula, rounded up per single-fill order.
Venue rules and fees as published in each source, checked October 4, 2026. Venues change them; check the current version.
Citations welcome. If you write about these markets, the CSVs above are the primary source and this page is the canonical writeup. Dataset name for attribution: DegenHedge Kalshi Signal Dataset. Past performance does not guarantee future results.
The free, verifiable part comes first.
Every signal in this study keeps posting to a free public channel and a public results log, graded win or loss with nothing pulled. That is the main way to follow this work.
This is a free trial that becomes an automatically renewing subscription. A payment card is required to start, and you are not charged today. Your trial runs 30 days. At the end of it your card is charged $4.99 and then $4.99 every month until you cancel. Cancel any time before the trial ends and you pay nothing: use your account page, send /manage in the bot, or email [email protected]. Cancelling later stops future renewals at the end of the current paid month and does not refund it. No refunds except as required by law. By starting a trial you authorize these recurring charges and agree to the Terms and Privacy Policy. Signals are informational and for entertainment only, are not financial or betting advice, and any single signal can lose 100% of your stake. Must be 18+ (some platforms require 21+). If gambling is a problem for you, call 1-800-GAMBLER.
These are 15-minute binary prediction-market contracts. Any single window can lose 100% of your stake. At high prices a loss costs many times what a win pays, and Kalshi's fee raises the win rate you need above the price. No entry-price bucket in this sample beat its breakeven by more than chance. Nothing on this page is financial, investment, or betting advice.
Past performance does not guarantee future results. Trade only with money you can afford to lose. Must be 18+ or of legal age where you live. If gambling is a problem for you, call 1-800-GAMBLER.