Equal-Output CoinJoin

An equal-output CoinJoin is the most common privacy-preserving collaborative transaction structure on Bitcoin. Participants agree on a uniform output value — for example 0.1 BTC — and each contributes inputs worth at least that amount. The resulting transaction pays out many identical-value outputs, so an outside observer cannot tell which output belongs to which input. This indistinguishability is what creates a meaningful anonymity set and improves the fungibility of the coins: after the mix, each equal output could plausibly belong to any of the participants.

Why equal denominations are required

Amounts are the strongest deanonymization signal on Bitcoin's transparent ledger. If outputs had arbitrary values, a chain analyst could often match an input's amount (minus fees) to an output's amount and undo the mix with simple arithmetic — the basis of subset-sum analysis. By forcing every mixed output to the same denomination, the transaction produces genuine ambiguity: with N participants receiving N equal outputs, there are many plausible input-to-output mappings and no amount arithmetic to collapse them. The privacy compounds, too. Cascading several rounds multiplies the ambiguity, because each successive mix draws from a population that was already mixed, so coins gain privacy that accumulates over successive rounds rather than resetting.

The change-output trade-off

Because few real balances are an exact multiple of the chosen denomination, most participants also receive a non-equal change output. Change outputs trivially reveal themselves as the odd amounts, and they remain linked to the owner's pre-mix history — so the single most common way users destroy their own privacy is spending mixed coins and unmixed change together, re-linking in one transaction everything the mix separated. Disciplined coin control is therefore not optional: label change as toxic, keep it quarantined, and either re-mix it or spend it in contexts where the linkage does not matter. Newer protocol designs (notably WabiSabi-style rounds) allow variable input amounts while still producing standardized mixed outputs, shrinking the change problem, but the discipline around what touches what afterward — covered under post-mix spending — remains the part no protocol can automate for you.

Structure, trust, and limits

A CoinJoin is a single ordinary transaction signed by all participants, which means it is trustless with respect to funds: every participant verifies that their expected output is present before signing, and no coordinator or peer can steal anything — the worst a malicious party can do is refuse to sign and stall the round. Rounds are typically assembled by a CoinJoin coordinator who never takes custody, or by decentralized peer protocols. The honest caveats: equal-output CoinJoins are visually distinctive on-chain, so they provide privacy, not invisibility — some counterparties treat mixed UTXOs differently — and privacy is only as good as the behavior that follows. For a sovereign holder, the mix is one tool in a larger practice: it breaks the historical link between your acquisition footprint and your future spending, and it works exactly as well as your post-mix habits allow.

Costs and expectations

Going in with correct expectations makes the tool more useful. Mixing costs real money — on-chain fees for each round, plus coordinator fees where applicable — and meaningful privacy usually takes multiple rounds, so budget for the process rather than treating one pass as a checkbox. Timing matters as much as structure: consolidating all your freshly mixed outputs into one transaction five minutes after the round undoes much of what you paid for, since amount clusters and timing correlate. Well-designed wallets counter this with randomized delays and by encouraging spends directly from mixed outputs. Finally, remember what the technique protects: it severs on-chain history, not real-world metadata — the merchant still knows what you bought, and your network layer still needs Tor. CoinJoin is one strong layer in a privacy practice, not the practice itself.