A Bitcoin balance is a sum of unspent outputs, not an account

Inherited archive entry · retained from the earlier edition. Its factual, legal and policy claims have not all been reverified for 19 September 2026. Research scope.

No balances, only outputs

Bitcoin's ledger does not store an account balance for any address. The developer guide's transactions chapter is explicit: when a wallet reports a balance, it really means that you have satoshis waiting in one or more UTXOs, unspent transaction outputs left over from earlier payments. An input then spends a previous output by naming its transaction identifier and output index, and that output is consumed in full; it cannot be partially spent. A wallet's displayed balance is therefore a sum computed by the software, not a figure the network keeps anywhere as a single record.

Why a change output exists

Because an output must be spent whole, paying an amount smaller than any single UTXO requires spending the whole UTXO and routing the surplus back to the spender. The guide describes this directly: change outputs are regular outputs which spend the surplus satoshis from the UTXOs back to the spender, and it recommends a fresh address for that change rather than the original one, to limit what a single address's history reveals. The whitepaper's ninth section, Combining and Splitting Value, frames the same design goal differently: transactions carry multiple inputs and outputs so value can be split and combined, normally to at most two outputs, one for the payment and one returning the change, if any, back to the sender.

What this means for reading a transaction

A transaction with two outputs is not necessarily a payment to two people; very often it is one payment and one change return to the same wallet. Fees are not a separate line item either: the guide notes that fees are based on the transaction's byte size, and the difference between total input value and total output value is what a miner collects. Treating every output as an intended recipient, or every multi-input transaction as evidence of pooled ownership beyond what the inputs themselves establish, overreads what the format actually records. The whitepaper's own caveat about fan-out is worth keeping: a transaction's dependency on earlier transactions does not require reconstructing a complete standalone history to verify it.

Questions to carry forward

• In a transaction with several outputs, is there a stated basis for which output is payment and which is change, or is that an assumption?
• Does a quoted balance for an address reflect a live UTXO sum, or a cached figure that may already be spent?
• Has a new address been used for change, or does a pattern of reused addresses make later linkage easier?

The UTXO model rewards precision: a balance is a sum, a change output is routine, and a transaction's shape says less about intent than it appears to.