Message Repair and STP in Legacy Flows

In payment operations, the most important number nobody outside the industry talks about is the STP rate — the share of payments that pass through a bank untouched by human hands. Straight-through processing is the goal; message repair is what happens when a payment falls short and someone has to intervene. The gap between the two drives cost, speed, and error rates across the entire back office.

What straight-through processing means

Straight-through processing (STP) is the automated handling of a payment from receipt to settlement without manual intervention. An STP payment arrives, passes validation, is enriched and routed automatically, clears compliance screening cleanly, and settles — all with no human touching it. High STP rates mean lower cost per transaction, faster processing, and fewer errors, which is why banks obsess over pushing the rate as close to 100% as possible.

When a payment cannot be processed automatically, it drops into an exception queue — sometimes called a repair queue or STP exception — where an operations analyst must inspect and fix it. Each of these manual touches is slow and expensive, and it introduces the risk of human error into an otherwise automated flow.

Why legacy flows break STP

Older message formats are especially prone to STP failures because they are loosely structured. Common causes include:

• Unstructured addresses and names. Legacy formats such as the SWIFT MT series often carry party details as free-text lines. A beneficiary name and address that a machine cannot reliably parse forces manual review.
• Missing or malformed routing data. An absent or incorrect BIC, an unrecognized clearing code, or an ambiguous intermediary institution stops automatic routing.
• Truncated fields. Fixed-length fields cut off long names or references, losing information needed downstream.
• Sanctions and compliance hits. A screening match — even a false positive from fuzzy name matching — pulls the payment out of the automated flow for investigation.
• Ambiguous purpose or remittance. Free-text remittance that cannot be matched to an invoice blocks automated reconciliation.

What message repair involves

Message repair is the operational process of correcting a payment so it can continue. An analyst in the repair queue might normalize a beneficiary address into structured fields, look up and insert a missing BIC, resolve an intermediary bank, correct a formatting error, or clear a false-positive sanctions alert after review. Banks maintain repair rules and reference data — BIC directories, clearing-code mappings, standing repair instructions for known counterparties — so that many repairs can be automated or semi-automated rather than done from scratch each time.

The distinction between fully manual repair and rules-based auto-repair matters. A well-run shop encodes common fixes as deterministic rules (for example, always mapping a known local clearing code to the correct BIC for a given corridor), which repairs the message automatically and preserves STP for that pattern going forward.

Why ISO 20022 raises STP rates

A central argument for migrating from MT to ISO 20022 is exactly this: structured data lifts STP. Where an MT103 carried a name and address as free text, ISO 20022's pacs.008 provides discrete elements for name, street, town, postcode, and country, plus structured references and purpose codes. Machines can validate and route on these fields without guessing. Richer, structured remittance information also improves automated reconciliation on the receiving side. The migration is, at its core, an STP program: fewer ambiguous fields means fewer exceptions and fewer costly manual repairs.

Measuring and improving STP

Operations teams track the STP rate as a core KPI and analyze the repair queue to find the biggest sources of exceptions. If a single corridor or counterparty generates a disproportionate share of repairs, that is a signal to add a repair rule, fix reference data, or push the counterparty to send cleaner messages. The economics are stark: manual repair can cost many times the marginal cost of an STP payment, so even a few percentage points of STP improvement across millions of payments is material. The strategic goal is to convert every recurring manual fix into a rule, so the exception queue shrinks toward genuinely novel cases.

Key takeaways

• STP is automated payment handling with no manual intervention; higher STP rates mean lower cost, faster processing, and fewer errors.
• Payments that fail STP drop into a repair/exception queue for manual correction.
• Legacy formats break STP through unstructured names/addresses, missing BICs, truncation, and sanctions false positives.
• Message repair fixes these; encoding common fixes as repair rules preserves STP going forward.
• ISO 20022's structured fields are a key driver of higher STP versus free-text MT messages.