This puts the entire twenty-one million on one page: what has been issued, what has gone for good, what sits with identifiable owners, and what remains for everybody else. It is the supply question asked more than any other and answered carefully less than almost any other.

Three species of figure appear side by side and are never allowed to blur into one another: readings taken straight from feeds, attributed guesses at things that cannot be counted, and a single leftover that is whatever the others do not claim.

What it actually measures

Every row is tagged by which kind of number it is, so a reader can see at once whether they are looking at something counted, something estimated or something inferred. That tagging is what makes the total honest rather than merely tidy.

That leftover is what survives every claim anybody can make on the supply. A fat one puts plenty of coin in ordinary hands; a thin one means identifiable owners and guessed-at losses have swallowed most of it.

Double counting is the standard error in charts like this

The most common mistake is counting the same coins twice. Most published loss estimates already assume the earliest mined coins are gone, so adding them again inside a lost figure double counts roughly a million coins on its own.

Listed miners create the same trap from the other direction, because they already sit inside company treasury lists and adding them as a separate slice counts their holdings twice. A supply chart that does not say how it handles these two cases is almost certainly getting at least one of them wrong.

What it does not tell you

Nobody knows how much coin has gone for good, and the published attempts disagree considerably. Picking one and stating it plainly would be the quickest route to looking authoritative and the quickest route to being wrong.

Known holdings are also only the visible ones. Entities that do not disclose sit in the residual alongside ordinary holders, so the unattributed share is an upper bound on how much is held by individuals.

And the picture moves as disclosure improves rather than as ownership changes. An entity publishing its holdings for the first time shifts coin out of the residual without a single coin having moved on the chain. The chart records a change in what is known, and it looks identical to a change in what is held.

How to read it

Broadly held. Once every claim is settled, plenty of coin is left in ordinary hands.

Balanced. Identifiable owners and guessed-at losses leave about the customary amount unspoken for.

Heavily claimed. Between identifiable owners and guessed-at losses, very little coin is left unspoken for.

The Supply dashboard carries The 21M Distribution together with The Real Float, HODL Waves and STH Supply.

Common questions

What are the three kinds of number?

Readings taken straight from feeds, attributed guesses at things nobody can count, and a single leftover that is whatever the others do not claim. Each row carries its label.

Why does the lost figure swing about so much?

Because nobody knows, and the published attempts disagree considerably. Letting the reader watch the whole picture shift between scenarios is more honest than picking one and stating it plainly.

Why are the earliest coins handled separately?

Because the published loss attempts already treat those coins as gone. Adding them a second time inside the same estimate would count around a million of them twice over.

Why are miners carved out rather than added?

Because publicly traded miners are already inside the corporate treasury tally. Giving them a slice of their own would count identical coins twice.

Is the residual really held by individuals?

It is an upper bound. Entities that do not disclose sit in there too, alongside ordinary holders.