Colophon

How the numbers are made

This is a residual ledger, not a wallet census. If a number cannot be cited, it is not on the chart.

Formula

implied_retail =
  circulating
  − lost_assumption
  − corporate_treasuries
  − funds_floor
  − governments_incomplete

Sources this edition

• Price and circulating supply: CoinGecko, 2026-08-15T14:38:12.356Z
• Corporate file: CoinGecko public treasury API, 179 companies, 1,282,501 BTC
• Strategy: 840,447 BTC on strategy.com, 15 Aug 2026 (matches the API row)
• Funds floor: IBIT 802,823 BTC, cited 27 Apr 2026. Stale. Other ETFs not added.
• Governments: incomplete public figures, dated 2025. Not a live sovereign watch.
• Lost coins: 3,700,000 BTC — About 17.6% of the 21M cap. Common public range is 3–4M lost/unspendable. This is an assumption, not a measurement.

Flippening chart

Flippening = implied retail < 50% of circulating. History is a handful of public anchors (Strategy 10-Ks, GBTC conversion, CoinGecko company census Aug 2025, US spot ETF ~1.5M Apr 2026). Years in between are reconstructed. Forecasts assume a constant net institutional add and ~164k new BTC/year until the cap. They are scenarios, not a model of price or policy.

This chart uses the 1.5M ETF-complex print. The ledger’s funds row still uses the IBIT floor so we do not pretend the rest of the complex is live.

What we will not do

• Scrape bitcointreasuries.net or clone their mNAV table
• Call address-size heuristics “retail” without labeling them as a second estimate
• Deanonymize wallets
• Invent a daily ETF complex total

Governments in this file

• United States (seized / forfeiture) — 198,012 BTC

  Seized coins are not a treasury policy. Some have been sold. Treat as a ceiling, not a wallet we watch.

• El Salvador — 6,102 BTC

  Sovereign accumulation. Confirm on the national addresses.

• Bhutan — 13,029 BTC

  Mining-linked sovereign stack. Attribution is estimated.