Payment Reconciliation

Overview

How it works and why it matters

Payment reconciliation compares obligations and internal records with authoritative external evidence so every incoming or outgoing amount is accounted for exactly once. It handles missing records, duplicate callbacks, fees, conversion differences and timing boundaries.

Important technical and operational details

It proves completeness and accuracy while identifying unmatched, duplicated, delayed, or incorrectly valued items.

Teams should design for missing records, reused references, cutoff mismatches, duplicate matches, wrong currencies, hidden fees, unresolved suspense, forced balancing, partial refunds, late settlement changes, and corrections without approval evidence.

Controls should keep original source records immutable, use stable match keys, explain many-to-one or one-to-many relationships, and route unresolved differences to an aged exception queue.

Practical decision points

Reconciliation is easiest to audit when it follows three layers: commercial obligation, payment execution evidence and financial posting. Start from an immutable invoice or payout ID, match provider or chain transactions, then compare the net amount recognized by the ledger. Unmatched items must enter a queue with a reason code and aging policy, not disappear into a manually edited total. For token transfers, capture chain ID, contract address, base-unit amount and transaction log index where applicable. Daily totals can match while two customers' payments are misallocated, so item-level consistency matters as much as aggregate balancing.

Related concepts

Payment Reconciliation Engine; Payment Reconciliation Report

Further reading

• bis.org documentation
• ifrs.org documentation