USD composite · now
—
vs Kalshi implied
—
vs 1.5s lag
warming buffer
drift 200ms
—
HAIR
wiring venues
FLAT
WIRING THE TAPE
HAIR does not place trades. The line at the top is the bet it would take if it were allowed to.
model edge
0¢
Venues · last / bid-ask / age
| venue | last | bid | ask | age | w |
|---|---|---|---|---|---|
| coinbase · down | — | — | — | — | 0.42 |
| kraken · down | — | — | — | — | 0.16 |
| bitstamp · down | — | — | — | — | 0.12 |
| gemini · down | — | — | — | — | 0.10 |
| bitfinex · down | — | — | — | — | 0.08 |
| binance · down | — | — | — | — | 0.07 |
| okx · down | — | — | — | — | 0.05 |
Kalshi 15-minute · KXBTC15M
Waiting on the open 15-minute window.
Kalshi hourly ladder · KXBTCD
Waiting on KXBTCD
| strike | bid | ask | mid |
|---|
Tape · prints as they hit
Waiting for the first print.
How this is faster than Kalshi
HAIR opens the public wires to Coinbase, Kraken, Bitstamp, Gemini, Bitfinex, Binance, and OKX, then builds a Coinbase-weighted USD price with median outlier rejection. Kalshi hourly and 15-minute BTC contracts settle on a 60-second average of CF Benchmarks' BRTI — a USD composite of overlapping regulated venues — sampled once per second in the final minute. Their UI (and that index) lag the raw tape. HAIR does not place trades. Nobody sees tomorrow's price. The edge is milliseconds and a cleaner clock, not magic.
Settlement is a 60-second average of BRTI in the final minute — not a single Coinbase print. A spike through a strike that fades before the window dies is noise. Fees and spread eat most of a 1–2¢ gap. This board only speaks at about 3¢ after fee. Hourly strikes only get a vote in the last 12 minutes.