Bitcoin velocity, and why the usual measure fits poorly

The first problem is what counts as a transaction. The ledger records every transfer, but not every transfer is a payment for goods. A change output returning to the sender is not spending. A transfer between two wallets owned by the same person is not spending. A consolidation of many small outputs into one is not spending. A transaction that moves coins to an exchange may be a sale, a deposit or an internal transfer, and the ledger does not say which.

The second problem is that the money supply itself is uncertain. The protocol supply is exact, but the effective supply is not, because some coins are unrecoverable. The lost bitcoin page explains why that total can only be estimated. A velocity calculated against the protocol supply will differ from one calculated against an estimated effective supply, and neither is obviously correct.

The third problem is that the numerator is not observable. Nominal spending in bitcoin requires knowing the price at which each transaction settled, and the ledger records quantities rather than prices. On-chain transfers carry no price. Off-chain activity, which includes exchange trading and layer-two payments, does not appear on the base chain at all. A velocity figure built from on-chain data therefore captures a fraction of the economic activity that the concept is meant to describe.

The fourth problem is that the measure is sensitive to how you count. Counting every output, every transaction or every coin-day gives different answers, and the choice is a methodological decision rather than a fact. Two analysts can compute two different velocities from the same ledger and both be defensible. That is why no single velocity figure for Bitcoin commands general agreement.

A low velocity figure is often read as proof that bitcoin holders are hoarding, and the inference does not follow. Velocity measures on-chain transfers relative to supply, and there are many reasons a transfer might not occur that have nothing to do with hoarding. A holder may be using a custodial service that settles internally. A holder may be transacting on a layer-two network that settles to the base chain only occasionally. A holder may simply have no reason to move coins this month.

There is also a definitional point. If an asset is held as a store of value, low velocity is the expected outcome rather than a surprising one. Gold has a very low velocity in the same sense, because most gold sits in vaults and does not change hands. Nobody treats that as evidence that gold holders are hoarding; it is what a store of value looks like. Applying a transactions-based measure to a store-of-value asset and then treating the low reading as a finding is a category error.

The measure is not useless, but it needs to be read for what it is. Rising on-chain velocity is evidence that more coins are moving, and that is worth knowing. Falling velocity is evidence that fewer coins are moving, which is consistent with holding but also with consolidation, custodial settlement and layer-two migration. The figure is a starting point for a question, not an answer to one.

A reader who wants to reason about how much bitcoin is actually available to trade should look at the supply side rather than the velocity side. The liquidity page covers order-book depth and why thin markets amplify price moves, and the trading volume page explains why reported volume is not the same as liquidity.