Maxi Onti MAXI ONTI

Bitcoin Supply

Education · Part 2 · 5 October 2026 · block 970,059

Who holds the bitcoin?

The whole 21 million. The five largest named entities, then the cap.

20.09MBTC mined. 20,093,934 at block 970,059.
906kBTC left under the 21 million ceiling.
3.125BTC paid to miners each block, until 2028.
~450New BTC a day. About 0.82% a year.

Start here

The schedule is known. The holders are not.

The cap is near 21 million. This epoch pays 3.125 BTC a block, about 450 a day, until block 1,050,000 around 12 April 2028. Then 1.5625 BTC, about 225 a day.

Mined20.09M of 21M · 95.7%
This halving epochAbout 80,000 blocks left

Block 970,059, on the way to 1,050,000. Trackers put the remainder around 79,941–79,974 blocks. The date drifts if blocks run fast or slow.

Schedule rows
ItemValueAs ofSource
Mined supply20,093,934 BTC5 Oct 2026, block 970,059btbjb.com
Cross-check20,093,613.63 BTC5 Oct 2026, block 970,026Bitcoin.com
Left under 21,000,000~906,066 BTC5 Oct 2026btbjb.com
Block subsidy3.125 BTCSince block 840,000, 20 Apr 2024Protocol
Next halvingBlock 1,050,000 · est. 12 Apr 2028Estimatebtbjb.com, Bitcoin.com
Subsidy after that1.5625 BTCScheduledProtocol
Daily issuance~450 BTCCurrent epoch3.125 × 144
Issuance per year~164,250 BTCSubsidy only, fees ignored450 × 365
Issuance / circulating~0.82%5 Oct 2026Calculated
21Mcap
Outer band: the three groups, named on the ring. Inner band: the split inside each group. Select a piece to open its table.
All holder rows
WhoBTCShareAs ofConfidence
US spot ETFs, 11 funds1,272,4026.3%4–5 Oct 2026, bitcoinetfdbHigh, tracker reading
of which IBIT761,8023.8%Same tableHigh, conflicts below
IBIT, other tracker800,5354.0%5 Oct 2026, bitcointreasuries.netMedium. About 39,000 BTC apart.
US ETF AUM cross-check~$108.89B · ~1.28M BTC6.42% of market capWeek ending 2 Oct, SoSoValueHigh on dollars. Medium on the BTC conversion at ~$85,300.
Europe ETPs, ex-Xapo113,1060.6%5 Oct 2026Medium
Europe table with Xapo152,0370.8%5 Oct 2026Xapo is custody, so it stays out of the pie.
Public companies1,279,0336.4%5 Oct 2026, bitcointreasuries.netMedium-high
Strategy (MSTR)848,0004.2%4 Oct 2026, 8-K filed 5 OctHigh. SEC filing. 4.04% of 21 million.
Metaplanet44,0000.22%30 Sep 2026High. Company disclosure.
Twenty One Capital43,5140.22%5 Oct 2026Medium
MARA35,5770.18%5 Oct 2026Medium-high
Strive29,4620.15%2 Oct 2026High. Company disclosure.
United States328,3721.6%Unchanged on the tracker since at least Jun 2026Medium-high on labeled seizure wallets. This is an Arkham-style tag, not a Treasury balance sheet.
China190,0000.9%Tracker, PlusToken-era estimateLow. Other 2026 tallies put China near 15,000. Weakest line in the sovereign table.
United Kingdom61,2450.3%TrackerMedium. Seizure cases.
Ukraine, “public officials”46,3510.2%TrackerLow. Read this as tagged wallets, not a finance-ministry reserve.
El Salvador7,7950.04%30 Sep 2026High relative to peers. Disclosed purchases.
Bhutan518~0%15 Sep 2026Medium. Down from a peak near 13,000 in Oct 2024.

Private companies stay out of the pie. Block.one ~164,000, Tether ~97,141, Xapo ~38,931. Low to medium confidence.

Same coins, other lenses

How locked is the supply?

This split adds up. The bars under it each use the same 20.09 million, and they overlap. Hover a bar for the note.

66.6%illiquid

  • Entities that spend little of what they receive. Glassnode, 4 Oct 2026. This pie adds up. The bars under it do not.

  • Entities that spend a meaningful share of what comes in.

  • Entities that spend most of what comes in. Exchanges and active traders tend to land here. A different series from exchange reserves.

Glassnode, 4 Oct 2026. The three buckets sum to 20,093,484 BTC.

Each bar is a share of 20.09 million. They overlap.

Long-term holders, unmoved ~155 days~16.2M · ~81%

CryptoQuant, around 30 Sep 2026. This bucket already contains most ETFs, treasuries, seized coins, and dormant coins. Medium confidence.

Unmoved at least 6 months~81% · >16M

River Financial, late September 2026, via TokenPost. A looser clock than Glassnode’s 155-day rule. Medium confidence.

Exchange reserves2.68M · 13.3%

CryptoQuant, early October 2026. Down from about 3.2 million in early 2024, the lowest on that chart since 2023. An alternate 2026 reading is about 2.50 million, 12.4%. The gap is about 180,000 BTC, and one of them may include ETF custody. Medium confidence.

Binance alone~700,100 · 3.5%

CryptoQuant, 20 Sep 2026. A local high, against the drop in the all-exchange number. Medium confidence.

Miner-linked wallets1,192,766 · 5.9%

CryptoQuant, 26 Sep 2026, down 1,530 BTC on the week. Overlaps public miner treasuries. Glassnode’s map is different: a Patoshi bucket of 1,096,204 BTC inside a larger labeled miner set, as of 28 Sep 2026. Medium confidence.

Short-term holder supply was 14.2% of long-term holder supply in the week of 5 October, up from 13.7% a week earlier. Glassnode, via Bitcoin Insider. That is a ratio between the two age groups. It is not another slice of the 20.09 million.

Still, or gone

Dormant is observed. Lost is inferred.

Still is observed. Gone is inferred. Likely inaccessible: 1.1–1.8 million, inside a dormant stock of about 3.5–5.2 million.

3.56M

Unmoved more than 10 years, as of 16 August 2026. The note puts that near 17.7% of the 21 million cap, with about 14,000 BTC aging into the band over the prior 30 days. High confidence that they are unmoved. Low confidence that they are lost.

5.15M

Unmoved 7 years or more, as of 22 June 2026. That was 25.7% of a then-circulating 20.01 million, from a ReWallet address scan. The author says the split between hoarded and lost cannot be known.

~1.8M

Unmoved 15 years or more, in a Bank for International Settlements study through March 2026. The authors treat this as a lost-coin estimate. About 0.1% of coins older than 15 years later moved. About 24,000 BTC have moved after 10 years of silence, and about 3,000 after 15. Medium confidence as an upper bound on inaccessible early coins.

~1.1M

Patoshi-pattern coins. Bitquery’s audit through 1 September 2026 puts the broad estimate near 1.17 million BTC and finds 1,023,352 provably unmoved on a stricter reconstruction. Glassnode’s Patoshi bucket is 1,096,204 as of 28 September 2026. The last known Patoshi-pattern spend was December 2017. Medium-high confidence on the unmoved balance. Medium confidence that one person holds it.

Likely inaccessible, as a range1.1M–1.8M · about 5–9%

The bar marks the range against circulating supply. It is an inference from age and the Patoshi pattern. Permanently lost coins cannot be observed directly.

What changed

One strong month is not a run-rate.

Last 30 days of US ETF creations ran near 1,100 BTC a day, converted from dollars. Issuance is 450. That window is not the 2026 run-rate.

~450 BTC issued per day, this epoch
~1,100 BTC per day, US ETF net, last 30 days, converted from dollars
Flow rows

28 Sep–2 Oct was about +$241 million. 30 Sep was −$148.7 million. The week posted 26 Sep was +$2.39 billion. In the week ending 2 Oct, IBIT was +$450 million and FBTC was −$168 million.

WhatWindowChange
US spot ETF net flow28 Sep–2 Oct 2026+$241 million
US spot ETF net flow30 Sep / 1 Oct / 2 Oct−$148.7M, then +$102.7M, then +$189.9M. 5 Oct was not yet posted.
US spot ETF net flowWeek posted 26 Sep, and 21–25 Sep+$2.39 billion. 21 Sep alone was about +$999 million.
Cumulative US ETF net flowJan 2024–2 Oct 2026+$57.857 billion, Farside
Strategy28 Sep–4 Oct+334 BTC at $85,838.80. Total 848,000. Funded by an ATM share sale plus cash.
StrategyTrailing ~30 days+2,615 BTC
StriveWeek ending 2 Oct+2,000 BTC. Total 29,462.
MetaplanetQ3 2026Sold 10,000 and bought 11,000. Net +1,000, to 44,000.
Long-term holder supplyAround 30 SepToward 16.2 million. Weekly position change positive after the rebound from the $60k area.
Coins leaving older cohortsFirst half of 2026About 300,000 BTC
Retail, individualsH1 2026, then Q3 into late SepNet sold about 140,000 BTC, then net bought more than 107,000.
Exchange reservesEarly 2024 to early Oct 2026About 3.2 million to 2.68 million. The downtrend held through the 2025 rally and the 2026 pullback.
Mt. Gox remainder30 Sep 202634,387.51 BTC. Trustee deadline 31 Oct 2026.

Coins last moved 6–24 months ago are 31.94% of supply, with an average cost above price. CryptoQuant, 4 Oct 2026.

Both sides

Tightening is visible. A squeeze is earned.

Free float has to shrink, the bid has to stay ahead of 450 BTC a day, and illiquid supply has to stay high when price rises.

What is already true

  • Illiquid supply is 13.38 million BTC, 66.6% of the coins Glassnode sorts.
  • Coins unmoved for at least six months are about 81%.
  • Exchange reserves are 2.68 million BTC, the low on CryptoQuant’s chart since 2023.
  • Issuance is 450 BTC a day until about April 2028, then 225.
  • Over the last 30 days, US ETF net creations alone ran ahead of issuance. Strategy, Strive, and Metaplanet were also net buyers in their latest windows.

What can unlock

  • ETF creations reverse. 30 September was negative, and FBTC lost $168 million in the week ending 2 October.
  • Long-term holders distribute. About 300,000 BTC left older cohorts in the first half of 2026, and the short-term share rose in the latest week.
  • The US and UK figures are seized coins. Bhutan’s stack has already fallen from about 13,000 BTC to 518.
  • Mt. Gox still has 34,387 BTC, with a deadline of 31 October 2026.
  • Treasury companies buy with equity and preferred stock. Metaplanet both sold and bought in the third quarter. If the equity premium compresses, the bid can stop.
  • Dormancy is a choice until it is not. The historical rate of very old coins moving is small, and it is not zero.

Tightening is already in the numbers. A full squeeze is still a 2027 question, and the issuance schedule does not force it. Mt. Gox still has 34,387 BTC, deadline 31 Oct 2026.

The supply-shock post on X

Keep the reading honest

What would change this page.

  • A revised IBIT holding from BlackRock, or a 10-Q, that settles 762k against 801k.
  • CryptoQuant and CoinGlass exchange reserves both rising for several weeks.
  • Glassnode illiquid supply falling while long-term holder supply falls, rather than coins only changing age bands.
  • Farside posting an outflow streak larger than September’s inflow streak.
  • Arkham, or a court filing, showing the US or UK moving seized coins to an exchange or an auction.
  • The Mt. Gox remainder actually paid out, then deposited to exchanges after 31 October 2026.
  • Strategy, or another large treasury company, reporting a net sale.

Data gaps

  • No single primary source publishes US ETF holdings and flows in BTC on one page. Dollars are primary. BTC totals depend on the tracker.
  • IBIT differs by roughly 39,000 BTC between bitcoinetfdb and bitcointreasuries.net.
  • ETP coverage outside the US and the Europe table is incomplete here. Canada, Australia, Brazil, and Hong Kong are missing.
  • Private-company totals are not filing-complete. Tether and Block.one are tracker figures.
  • China’s 190,000 against roughly 15,000 BTC is unresolved. Ukraine’s line is not a sovereign reserve.
  • Exchange-reserve providers differ by about 180,000 BTC and may or may not include ETF custody.
  • Long-term holders at about 16.2 million, and “unmoved six months” at 81%, use different clocks.
  • Miner balances and public miner treasuries are not reconciled entity by entity.
  • Year-to-date 2026 ETF flows in BTC are not clean enough to publish.

Where the numbers come from

Sources

Figures use 20,093,934 BTC as circulating supply unless a row says otherwise. This is a reading of public data, not a forecast and not financial advice.

Back to Bitcoin