I said in an earlier article that I feed four cats, or nine, or fifteen, depending on what you mean by "feed". The number isn't a fact about the colony — it's a fact about the definition you picked.

"Who owns bitcoin" is the same kind of question, and almost every alarming statistic you've seen about it is really a statement about a definition. So before the numbers, the definitions — and before the definitions, an admission about what can actually be known.

1. Nobody Can Actually Count This

The ledger is public, which creates the illusion that ownership is measurable. It isn't.

What's visible is addresses, not people. One person can control thousands of addresses; one address can hold coins belonging to millions of people. Everything you read about "whales" or "the top 1% of holders" comes from clustering heuristics — educated inference about which addresses share an owner, built from spending patterns and labelled wallets.

That work is serious and it is also estimated, revised, and wrong at the edges. Any ownership statistic quoted without error bars is being presented with more confidence than it has.

2. The Categories That Actually Matter

More useful than a percentage is knowing which kind of holder you're looking at:

• Lost and dormant. A substantial quantity from the early years has not moved in well over a decade and is widely assumed gone — keys discarded, drives lost, owners dead. Functionally this reduces the supply, though nobody can say by how much, because "hasn't moved" and "can't move" look identical on-chain.
• Long-term holders. Coins that sit for years through full cycles. The most informative cohort there is, because the behaviour is expensive to fake.
• Exchanges and custodians. Enormous clusters that look like single whales and are actually millions of customers. This is where most "one entity holds X" headlines come from, and it's usually a misreading.
• Funds and ETFs. Growing, concentrated in a handful of custodians, held on behalf of many beneficiaries who hold claims rather than coins.
• Corporate treasuries. Visible, announced, and unusually reflexive — these holders have shareholders and debt, which means their coins can be sold for reasons that have nothing to do with bitcoin.
• Miners. Receiving newly issued coins continuously and selling a portion to pay electricity bills. Structural sellers, by business model.
• Governments. Mostly from seizures. Occasionally sold, sometimes with advance notice, which makes them the least predictable category.

Notice how differently each behaves under stress. A lost coin never sells. A long-term holder rarely does. A leveraged treasury company might have to.

3. The Float Is Smaller Than the Supply

The practically important point.

Of roughly 21 million eventual coins — with something like 95% already issued — only a fraction is actually available to trade at any moment. Subtract the lost, the long-dormant, the locked-up institutional holdings, and what's left to set the price is a far smaller pool.

This explains something confusing about the market: why relatively modest flows move the price so much. The marginal buyer and seller are transacting against a thin float, not against the whole supply. It's the same arithmetic as the market-cap article — price is set at the edge, and the edge is smaller than the headline number suggests.

It cuts both ways. A small float amplifies moves up and down equally, and "low float" has never been a guarantee of anything except volatility.

4. Does Concentration Matter?

Honestly: somewhat, and less than the headlines imply.

The case that it matters: a small number of custodians holding a large share recreates exactly the structure bitcoin was built to route around. Not through theft — through the ordinary gravity of custody, regulation, and whoever ends up holding the keys for everyone else.

The case that it's overstated: most of the scary clusters are exchanges and ETFs, which are many people wearing one address. "Ten entities control X%" usually means ten institutions holding other people's coins, which is a different claim from ten individuals.

The thing actually worth watching isn't the concentration figure. It's the direction: are coins moving toward self-custody over time, or toward custodians? That trend says something real about whether the thing is becoming what it claims to be. A snapshot percentage says almost nothing.

5. What to Do With Any of This

Very little, and that's the point.

Ownership distribution is context for understanding why the market behaves as it does — thin float, structural sellers, reflexive corporate holders. It is not a trading signal, and the people who quote it hardest are usually making an argument rather than a measurement.

The one practical consequence: you can move yourself out of a category. Coins in your own custody are, by definition, not part of anyone's float, not someone's claim, and not available to be sold by a third party under pressure. That's the only part of the distribution you control.

The Point

Four cats, or nine, or fifteen. All true, all answering different questions, and the one that sounds most impressive is the one that means least.

Treat every ownership statistic the same way: ask what was counted, who did the clustering, and what the person quoting it wants you to feel. Then go and check which category your own coins are in, because that's the only entry in the table you get a vote on.

Count the definition before you count the coins. 🐾⚡