Both markets ask the same question, whether Bitcoin ends the window at or above where it started, and both resolve a tie in favor of Up. Per each venue's published rules, checked October 4, 2026, they differ in three ways that matter: Polymarket settles on a 60-second Chainlink TWAP and Kalshi settles on a 60-second average of the CF Benchmarks BRTI; both charge takers 0.07 × price × (1 − price) per contract, but Kalshi rounds the charge up to the cent while Polymarket rounds to five decimal places; and Polymarket windows are a third as long.
Across 36,731 officially settled Polymarket 5-minute windows and 18,025 officially settled Kalshi 15-minute windows we logged, both venues landed Up 49.9% of the time. With two and a half to three minutes left, where Bitcoin sat relative to the open decided most windows, and the Kalshi ask already priced that in. A spot chart from Coinbase or Kraken pointed the wrong way at the close on 8.88% of Polymarket windows and 5.06% of Kalshi windows.
Read from each venue's own rules and fee documentation, checked October 4, 2026.
| Polymarket "Bitcoin Up or Down", 5 minutes | Kalshi KXBTC15M, 15 minutes | |
|---|---|---|
| New window | Every 5 minutes, around the clock | Every 15 minutes, around the clock except Kalshi's Thursday 3 to 5 AM ET maintenance |
| Settlement source | Chainlink BTC/USD time-weighted average price (TWAP) data stream | CF Benchmarks Bitcoin Real Time Index (BRTI) |
| Settlement value | 60-second Chainlink TWAP at the end of the window (since August 14, 2026) | Simple average of the 60 BRTI prints in the final minute, rounded to the cent |
| Price to beat | The same Chainlink TWAP at the start of the window | The same 60-second BRTI average before the window opens (the "strike") |
| Exact tie | Resolves Up ("greater than or equal to") | Resolves Yes, which is Up ("at least") |
| Taker fee per contract | 0.07 × p × (1 − p), rounded to 5 decimal places | 0.07 × P × (1 − P), with the order total rounded up to the next cent |
| Maker fee | Makers pay no fee; Polymarket lists a 20% maker rebate rate for crypto markets | Only on markets Kalshi lists for one; KXBTC15M is listed under the standard fee type |
| Official result | Marked closed 53 to 55 seconds after the window ended in the markets we checked | Available to our logger within about 30 seconds on 98.3% of graded windows |
The fee coefficient is the same on both venues, 0.07, and both scale by price × (1 − price). The difference is rounding. On one contract at 50¢ the formula gives 1.75¢: Polymarket charges 1.75¢ and Kalshi rounds the order up to 2¢. On one contract at 99¢ the formula gives 0.0693¢: Polymarket charges that, and Kalshi rounds the order up to a full cent, about 14 times as much, which is the whole profit a 99¢ contract can make. On a 100-contract Kalshi order the rounding is spread thin and the gap nearly closes. Our Kalshi fee and breakeven study works through every price.
Over tens of thousands of windows, neither venue leans.
Polymarket's 5-minute windows settled Up 49.87% of the time (95% range 49.35 to 50.38) and Kalshi's 15-minute windows settled Up 49.86% (49.13 to 50.59). Both rules hand an exact tie to Up, but a tie on a 60-second average quoted to the cent almost never happens, and no tilt shows in either sample. There is no direction to lean on before the window starts.
The question people ask most about these markets: if Bitcoin is up with a couple of minutes left, how often does it stay up?
| Bitcoin's move from the open so far | Polymarket 5-min: finished on that side | Kalshi 15-min: finished on that side | Kalshi windows with an ask: won, at average ask |
|---|---|---|---|
| Under 0.01% | 56.9% | 56.3% | 57.2% at 58.6¢ |
| 0.01% to 0.02% | 64.9% | 68.4% | 68.8% at 69.7¢ |
| 0.02% to 0.05% | 74.4% | 79.2% | 79.7% at 80.9¢ |
| 0.05% to 0.10% | 84.6% | 91.0% | 90.7% at 91.6¢ |
| 0.10% to 0.20% | 92.0% | 97.1% | 96.3% at 96.3¢ |
| 0.20% or more | 95.6% | 99.4% | 98.5% at 97.9¢ |
| All windows | 74.9% | 87.6% |
Two things stand out. First, the same size of move held up more often on Kalshi than on Polymarket. Time remaining was similar on both, so the likeliest reason is our measurement: the Kalshi move is measured against the official strike, while the Polymarket move is measured against our own spot read at the open, which adds noise to which side Bitcoin was really on. We cannot separate how much of the gap is the venue and how much is that noise. Second, look at the last column. In every bucket the Kalshi ask for the side Bitcoin was on sat within 1.4¢ of how often those same windows finished on that side, and above it in five of the six buckets. The market already knew. "Bitcoin is up with two minutes left" is information everyone watching has, and it is in the price.
Why there is no Polymarket price column. Our Polymarket logger recorded the market's displayed outcome price, not an executable ask. That field failed a basic check: windows logged at 48 to 52¢ finished on the logged side 70.1% of the time across 19,366 windows, which a live, tradeable price would not do. We treat the field as a stale display value and publish no Polymarket price, return or fee-adjusted figure from it. The outcomes above use Polymarket's official settled results only.
Neither venue settles on Coinbase, Binance or any single exchange. Close calls show it.
For every window we compared the direction our own spot feed (Coinbase and Kraken) showed at the close with each venue's official result. They disagreed on 8.88% of Polymarket windows (3,260 of 36,731) and 5.06% of Kalshi windows (912 of 18,023). Nearly all of the disagreement sits in the close calls:
| Distance between spot close and the window's reference | Polymarket: spot read disagreed | Kalshi: spot read disagreed |
|---|---|---|
| $0 to $5 | 36.4% of 3,977 | 38.6% of 852 |
| $5 to $10 | 21.8% of 3,300 | 23.2% of 923 |
| $10 to $25 | 10.4% of 7,768 | 10.1% of 2,392 |
| $25 to $50 | 2.9% of 8,332 | 2.8% of 3,463 |
| $50 or more | 0.3% of 13,354 | 0.3% of 10,393 ($50 to $100: 0.5%; $100+: 0.1%) |
Polymarket disagrees more often for two reasons we can see. Its windows are shorter, so more of them end close to the line: the median 5-minute move in our sample was $33.45 against $61.66 for 15 minutes. And our Polymarket reference is our own opening read rather than Chainlink's. The rate held between 7.2% and 9.4% across Polymarket's three settlement-rule periods, so the August switch to a TWAP did not change the picture. When Bitcoin is within $10 of the line at the close, a spot chart is a weak guide to the result on either venue.
A measurement, not a recommendation.
Related: how Kalshi BTC 15-minute markets settle, what 9,800 graded Kalshi BTC 15-minute windows show, and what a graded signal is.
Per Polymarket's market rules, checked October 4, 2026: Up if the Chainlink BTC/USD time-weighted average price at the end of the window is greater than or equal to the same TWAP at the start, otherwise Down. Since August 14, 2026 the 5-minute markets use a 60-second TWAP. They do not settle on any single exchange's price.
Yes (Up) if the simple average of the 60 CF Benchmarks BRTI prints in the final minute is at least the same 60-second average taken before the window opened. Averages are rounded to the cent, and an exact tie resolves Yes.
Per each venue's published docs, checked October 4, 2026, both charge takers 0.07 × price × (1 − price) per contract. Polymarket rounds the fee to five decimal places and charges makers nothing. Kalshi rounds a retail order's total up to the next cent, so a single 99¢ contract pays 1¢ on Kalshi against about 0.07¢ on Polymarket. On large Kalshi orders the difference mostly disappears.
Because neither venue settles on a single exchange. In our data a Coinbase/Kraken spot read at the close pointed the opposite way from the official result on 8.88% of Polymarket 5-minute windows and 5.06% of Kalshi 15-minute windows, and on more than a third of windows where spot was within $5 of the line.
It depends on how far up. In our data, with about two and a half to three minutes left, a move under 0.01% finished on the same side about 56 to 57% of the time on both venues, and a move of 0.20% or more finished on the same side 95.6% of the time on Polymarket and 99.4% on Kalshi. On Kalshi the ask for that side already sat within 1.4¢ of how often it won.
Neither. 49.87% of 36,731 Polymarket 5-minute windows and 49.86% of 18,025 Kalshi 15-minute windows settled Up.
No. Our paid Bitcoin feed covers Kalshi's 15-minute series only. In the March 24 to October 4, 2026 sample it won 91.21% at a 91.74¢ average entry, 0.53 points below breakeven before fees, and it is sold as fast, publicly graded coverage, not as a source of returns.
What we logged, what we left out, and why.
One logged observation per 5-minute window from our Polymarket logger, kept only where Polymarket's official result had been read back from its public API (36,731 windows, May 29 to October 4, 2026). Twelve rows whose market slug did not match the window they were logged in were dropped.
One logged observation per KXBTC15M window, kept where Kalshi's official result was read back (18,025 windows, March 24 to October 4, 2026; two rows with a corrupted strike value are dropped where a strike is needed). The Kalshi ask is one minus the best opposing bid in the live order book, rounded to the cent.
No Polymarket prices, returns or fees from our logs, for the reason given in Finding 2. No claim about which venue is better to trade: that depends on your order size and fills, which we did not measure.
Venue rules and fees as published in each source, checked October 4, 2026.
Citations welcome; the CSV above is the primary source and this page is the canonical writeup. Past performance does not guarantee future results.
The free, verifiable part comes first.
Our Bitcoin signals run on Kalshi's KXBTC15M series. Every one posts to a free public channel and a public results log, graded win or loss with nothing pulled.
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 5-minute and 15-minute binary prediction-market contracts. Any single window can lose 100% of your stake. Settlement follows an index or oracle average, not the exchange chart you may be watching, and close windows can settle against what your chart shows. 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.