The 20 million coins as our ledger holds them (outputs of 0.01 BTC and up), split into thirteen age bands, every day since the first block: how many coins in each band, what they cost, and what the average coin in the band paid.
12,706,607 BTC coins that have not moved in over a year on Oct-22 2026, the last day in this build, measured and published by askthetape.com.
Every coin our ledger holds has an age: the days since it last moved. Sort them into thirteen buckets — under a day, a week, a month, on up to ten years and older — and you can watch supply walk from the young end to the old end through a bear market and back again. We show coins held over a year, all age bands, each band’s cost, and each band’s average last-move price in that order.
| leg | our own copy of the UTXO ledger (it was built once from a public chain archive and has been extended block by block from our own node since 2026-07-23): every output of 0.01 BTC and up, with the day it was created and the day it was spent (6,797,539 rows on this build, ledger to 2026-10-22) × exchanges' published daily closes (Bitstamp from Aug-18 2011; the Mt.Gox tape before) · bands by days since last move: 1, 7, 30, 90, 180, 365, 548, 730, 1095, 1825, 2555, 3650 · cost = coins × the close on the day they last moved |
|---|---|
| cadence | daily |
| since | 2009-01-03 |
| 🩸 what is missing on purpose | there is a fourth cut of this — the NUMBER of unspent outputs in each band — and it is not here. Our ledger builder drops the dust: it accounts for the coins to within half a percent but keeps only about a tenth of the outputs, so an output COUNT from it would be wrong by ten times. The count of unspent outputs on this site lives on «The network's daily vitals» — counted on our own node from 2025-11-03. We would rather have a hole than a wrong line. |
| 🩸 warm-up | the cost and average-price views rest on a price for every birthday, and our daily price starts 2010-07-17. Coins older than that carry a zero cost, so the oldest band's average price reads too low in the early years. |
| 🩸 measured against what comes next | Measured 2026-09-13: the share of coins older than a year has run with the next 90 days on our own data since 2013 (rank +0.25, three cycles — too few to call it more than a cycle marker); its 90-day change has not |
| policy | no backfill · read it here · export: whole history |
| hourly window · where it comes from | the events of this window are read from our own node; the ledger they are added to was built once from a public chain archive and has been extended block by block from our own node since 2026-07-23. The events are every output of 0.01 BTC and up that died and every output that was born, read block by block (getblock … 3 with prevout), with the hour it happened. The carrier holds 8,174 hours, 2025-11-16T00:00:00Z → 2026-10-23T18:00:00Z, 340 days deep from a fixed first day and growing every hour. |
| hourly window · how the hours fold | STATE: the LAST hour of a stretch is that stretch's value. A coin's BAND is decided in whole days, the way the daily page decides it — inside an hour it is the supply that moves, not the band edges. The dollars each band carries are not drawn: they are the band's coins × the band's average price, both of which are. «N hours» in the window's own header means BUCKETS WITH DATA, not the distance from the first point to the last — the header prints both. Selected spent-output flows are zero in verified no-block hours; ratios, states and other missing observations remain blank. Block counts are a separate series where present, and describe the native node segment only. |
| hourly window · what this ledger cannot see | the ledger does not see outputs below 0.01 BTC — that is 87,202 BTC, 0.434 % of the coins alive on 2026-10-22, measured against what our own node counts as mined (the scheduled subsidy of every block header it holds). It is a threshold, not a fault, and it is the same threshold the daily page carries. |
| hourly window · which price | the hour is priced with the Binance BTC perpetual of our own capture — the price of THAT HOUR. The daily page uses the Bitstamp daily close. The two were measured against each other on this build over 340 days: median 0.4 bp apart, mean 0.7 bp, worst 42.9 bp. Coins born before this window keep the daily close of their birthday, because the hour of their birth is not in the carrier. Second calibration, hour for hour against the public hourly closes of the spot venue (167 hours, measured 2026-09-19): median -0.05 bp, worst 24.46 bp — so the gap in the first number is the 23:00Z hour against a daily close, not one venue against another. A one-off check, not a captured feed. |
| hourly window · one line or two | one line, not two: for this row the hourly rule and the daily page's rule give the same number, and drawing a second line would promise a difference that does not exist. |
| hourly window · judged | This comparison covers the native node segment from 16 November 2025 only: 🩸 12 birth × spend cells on 3 of 341 closed days differ from the daily ledger of our own node — see the judge log; this window is not vouched for until that is explained. Judge run 2026-10-23, 13.3 hours ago. The judge runs daily against the daily ledger of our own node; a day the 35-day carrier no longer holds is not judged rather than counted as a difference. |
| this build | data to Oct-22 2026 |
| policy | no backfill · read it here · export: whole history |
The chart itself is drawn in the browser: turn JavaScript on to hover a day, pick a window and draw your line. The numbers above come from the last build, Oct-23 2026 21:24Z.
What the tape says today · what's new · all 111 metrics