UTXO consolidation

The fee reason, stated plainly

Every input adds witness and outpoint data. Virtual size — and therefore the fee at a given sat/vB — grows with input count. A wallet that later needs to pay someone during congestion will pay more if it must attach fifty small UTXOs instead of one large one. Consolidating when the mempool is quiet (low sat/vB) is a common custody practice: pay a cheap fee now to reduce fee surface later.

That is an economic decision about future transaction cost, not a statement about spot price. An exchange sweeping cold-adjacent wallets, a miner paying itself, and an individual rotating storage can produce the same many-to-one shape. Size of the output does not tell you which of those happened.

What the chain actually shows

A consolidation you can defend from public data has a visible shape:

• Input count much larger than output count — often tens or hundreds of inputs and a single output.
• Output value close to the sum of inputs minus a modest fee (no large leftover “payment” split).
• Fee rate that may be low if the sender was not in a hurry.

What you still cannot see: whether every input was controlled by the same legal entity (clustering is a guess), whether the new output is “cold,” or whether a sale will follow. Our Brief classifies many-inputs → one-output as a possible consolidation and keeps certainty modest. See Methodology for the exact rule.

A sourced example

Transaction 61b5cc645177843e8514e18a19faf162a19247c8bb4f25519b57c166e9df302c (block 615825, 2020-02-03) spends 147 inputs into one output totaling about 22,024 BTC. An input matches a Bitfinex address in our open directory. The destination output is unlabeled. That is a textbook consolidation of coins already attributed — on the input side — to a known exchange. It is not proof of a customer withdrawal or a dump. Full write-up: Examples. Verify on mempool.space ↗.

What consolidation is not

• Not automatically “preparing to sell.” Custodians consolidate before withdrawals, before cold storage, and during routine wallet maintenance. The next hop may never be a market sale.
• Not CoinJoin. Privacy constructions often have many inputs and many similar-sized outputs. Consolidation collapses inward. See CoinJoin ambiguity.
• Not net new demand or supply. Coins already existed as UTXOs. The transaction reorganizes ownership records; it does not mint BTC.

How Radar and Brief treat it

The Radar sums output value. A consolidation’s BTC column is therefore the size of the new UTXO(s), including everything that remained under the same economic control. That number is an upper bound on “coins that changed hands,” not a payment amount.

Brief’s engine (see brief-pure.js) treats more than three inputs and exactly one output as a possible consolidation when no label rule fires first. If a sourced label is present and the output total is large, the labeled-touch rule can outrank the shape rule. Read the decision-logic list on every result — that list is the source of truth for that run, not this paragraph.

Reading checklist

1. Count inputs and outputs before reading any headline.
2. Note fee rate: a cheap many-input sweep often means “not urgent,” not “hidden whale.”
3. Check whether labels sit on inputs (sender already known) or outputs (destination known).
4. Paste the txid into Brief and compare its steps to your count.
5. Open the containing block on an explorer. One confirmation is inclusion, not a market event.

The privacy cost of sweeping

Every input in a consolidation is revealed as spendable by the same signing session. Addresses that never co-spent before become linked in public history. That is often acceptable for an exchange hot wallet; it is a serious trade for an individual who received coins on many unused addresses. CoinJoin-like constructions try to avoid that link. Consolidation embraces it. If you are reading someone else’s sweep, do not treat the new output as “more anonymous” — it is usually less ambiguous than the pile of inputs it ate.

Change vs consolidation

A 1→2 transaction is typically payment plus change, not a sweep. A 40→2 transaction might be a sweep that left a leftover output, or a payment funded by many coins. Brief’s rule fires on exactly one output for the consolidation heuristic. Two outputs keep you in high uncertainty unless a label rule fired first. That is deliberate. See Methodology.

Further reading

• Reading large transactions — shapes in one place
• Cold wallet movements — when the destination is custody
• How to read a whale move — pillar guide

Educational only. Not trading signals.