Barritt Consulting

Transaction Monitoring Data and Lineage

A monitoring scenario can only work with the information that reaches it. A transaction monitoring data quality assessment follows required data from source systems through mappings and transformations to the records available when an alert is created.

Connect business risk to data coverage

The starting point is the business activity and risk the monitoring process is expected to cover. A defined review can map those needs to the products, accounts, transaction types, channels and customer attributes supplied to monitoring. It can ask whether the relevant source systems and populations are included, and how exclusions, rejected records, duplicates or late-arriving data are identified.

The FCA Handbook says applicable systems and controls should enable firms to identify, assess, monitor and manage money laundering risk in proportion to their activities. Its guidance also points to processes, systems, transaction complexity and volume as relevant factors. FATF standards call for countries to apply proportionate measures in light of assessed risk and national circumstances. These sources help frame whether monitoring data covers the firm's activities; specific requirements depend on applicable rules.

Follow each transformation to the alert

Lineage means being able to follow relevant information from its original source through interfaces, field mappings, cleansing, enrichment and aggregation into the monitoring process. The review can examine how identifiers link records to customers and accounts; whether units, currencies, dates and time zones are handled consistently; and how missing or conflicting values affect processing. It can also consider whether changes to mappings or transformations are documented and controlled.

Completeness and timeliness need evidence. Reconciliations can compare source and downstream counts or totals across stages and periods, with exceptions assigned for investigation. Trace testing can then follow selected records from the original source into the data used by monitoring and, where applicable, back from an alert to its underlying records. This helps distinguish a scenario or threshold question from a feed, mapping or transformation issue.

Trace the data that supports monitoring

A first conversation can define the jurisdiction, monitoring scope and known data concern. Useful inputs may include system and interface maps, field definitions, mapping specifications, reconciliation reports, exception logs and carefully governed sample records or alerts. The discussion can identify unclear coverage, weak traceability or unresolved reconciliation questions, then agree priorities and any written output in advance.

Data lineage review follows source records through interfaces and transformations to the data used by monitoring. Scenario redesign, alert tuning and case investigation address different questions and should be evaluated separately. A 2017 NICE Actimize webinar on transaction monitoring agility listed Trevor Barritt as a session leader; the dated event records his work in this field.

Common questions

What does transaction monitoring data lineage show?

It shows how relevant records and fields move from their original source through interfaces and transformations into the monitoring process, and how data used for an alert can be traced back to those sources.

Is a data lineage review the same as alert tuning?

No. Data-lineage work examines coverage, completeness, timeliness, mappings and transformations. Scenario design and threshold tuning concern how monitoring rules use that data.