Of everything that arrived at an exchange that day, the share that came in just the ten largest transactions at each venue — read off our own address map. Bold line = 30-day average.
0.472 the ten largest per exchange, as a share of the day's arrivals on Oct-22 2026, the last day in this build, measured and published by askthetape.com.
We keep our own map of which addresses belong to which exchange, and we keep every transaction that crossed that line since 2022 (a public copy of the chain to Sep-12 2026, our own node since). So we can ask a question a total cannot answer: was the day's arriving coin one crowd or ten wallets? Take each exchange, take its ten largest arrivals of that day, add them up, and divide by everything that arrived anywhere. Over the last year 51.5 % of all arriving coin sat in those ten transactions per venue; on Oct-22 2026 the number was 0.472. We measure concentration in the largest arrival transactions; this does not count the people or decisions behind them. The coin totals behind the share — what arrived, what left, and the balance — are on «Coins in and out of exchanges», and the twenty largest transactions of the last closed day are named there one by one. Thinner where we say so: the ratio is only as complete as the map — Coinbase is volume-only here, Upbit, Bithumb and Coinone are not mapped (Korbit is, thinly), and before 2025 the tape is thinner, so the early years read the shape of concentration rather than its exact height.
| leg | own address map v2.1 · transaction-level exchange arrivals, ranked per entity per UTC day · 1756 closed days, 2022-01-01 → 2026-10-22 |
|---|---|
| cadence | daily; the carrier is rebuilt by the map's owner each morning (05:50Z) and this site only copies it |
| since | 2022-01-01 |
| 🩸 the running day is left out | the carrier is written while today is still going, so its last line is a fraction of a day and is NOT drawn: 2026-10-23 carried 640 transactions, 0.16 of the previous thirty days' median. The rule is the calendar, not a threshold — the last day we draw is the day before the carrier was built (2026-10-23T05:50Z) — because arrivals have a weekly rhythm and the quietest REAL day of the 1,726 we could measure that way still carried 0.43 of its own 30-day median. A partial day would read as a record high, since ten transactions cover almost everything when there is almost nothing. |
| thinner here | Thinner here: the map is ours, so the ratio is only as complete as the map. Coinbase's hot wallets are selected statistically and its net is not readable, Bitstamp is volume-only, Upbit, Bithumb and Coinone are not mapped (Korbit is, thinly), and the coins an exchange shuffles between its own wallets are left out — the same rule as on the netflow page. Before 2025 the tape is thinner, so the early years read the SHAPE of concentration, not its exact height. The ratio is a share, so a thin day cuts both the top of the fraction and its bottom: that is why the level holds up while single days do not. |
| fund flow ratio, our way | Our own ruler with the definition written out: the coins that arrived at an exchange divided by every coin that moved on the chain that day (our ledger's own «coins that moved»). 🩸 The bottom of the fraction is our ledger's own «coins that moved», not an explorer's estimate of transaction volume — a different quantity, and it would make a different ratio. |
| policy | no backfill · read it here · export: whole history |
| hourly window | our own exchange tape, transaction level (rebuilt daily 06:07Z) · since 2026-07-13 00:00Z by data, 2,444 hours with data inside a 2,453-hour span → 2026-10-23 04:00Z (the Tape and Desk windows draw up to 400 days of that, never more than we hold) · captured 2026-10-23T21:21:56Z |
| 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