Testing regulatory reporting

Prove the reported result is correct.

DORA, MaRisk and BCBS 239 free the budget. The real question stays a business one: did the processing behind the report produce the correct result? SQACE verifies that with rules and records the evidence defensibly.

Trigger Regulation as a budget trigger (DORA, MaRisk, BCBS 239), not as the category. Personas: CDO, Test and QA, CISO.

The problem

The report is filed. Is the result behind it correct?

Regulatory reports sit at the end of long processing chains. Aggregations, thresholds and classifications interlock. A misapplied rule does not change the format of the report, it quietly changes its content.

In an audit it is not enough to have tested a lot. The question is who checked what, when and on what basis. That reproducible, traceable evidence is exactly what most tools do not deliver: they show that data looks plausible, not that the processing computed the right result.

Approach

Set the reported result against the business rule.

SQACE computes, from the input data and your business rules, what must be reported, and sets it against the report result actually produced. Every deviation is classified, explained in business terms and recorded.

RUN REPORT-Q4-EXAMPLE, on-prem
RUN REPORT-Q4-EXAMPLE, on-prem COMPLETED
SOURCE MS SQL Server JDBC, LOCATION On-Prem/EU-Cloud, ACCESS read-only, REST /api/v1/runs
report_amount_eur Expected 2,430,000.00 Actual 2,430,000.00 verified
threshold_eur Expected 1,000,000.00 Actual 1,000,000.00 verified
reportable Expected yes Actual no deviation
Rule IF the report amount exceeds the threshold, THEN the position is reportable. In the actual run, it was flagged as not reportable despite exceeding the threshold.

Expected is the second, rule-based calculation from the input data. Classified as a rule deviation in the reportable field, tracked across runs.

Run reproducible, versioned, in the audit trail. Security and architecture
Illustrative example, synthetic values

CDO and Head of Data

Defensible evidence

Solid proof for the board and the regulator: classified, tracked across runs, reproducible. Not another dashboard, but evidence that holds up in an audit.

Test and QA

Secure the recurring run

Reporting runs recur with every cut-off date. Rule sets built once check every run again, scheduled or CI/CD-triggered via the REST API.

CISO and Compliance

Built in, not bolted on

Audit trail, versioning, permission concept and automatic deletion are part of the protected data space, on premises or in the EU cloud.

In practice

Regulatory testing in banking, documented.

At LBBW, SQACE has been the central platform for data testing since 2020. In the data-testing lifecycle, effort is documented publicly to have fallen by 40%, and by 50% for creating test cases.

An example of the scale: in a capital-gains-tax calculation, 15,847 positions were verified fully automatically against the rule-based expected result, rather than checked manually by sampling. At Commerzbank, an anti-money-laundering monitoring pipeline is verified fully automatically with SQACE.

At LBBW, SQACE is the central platform for data testing, with 40% less effort in the data-testing lifecycle. The figures shown are documented publicly on sqace.io.

LBBW, public reference (Jens Wauer)

The 15,847 verified positions are a method-framed example of the degree of automation, not a marketing number. Over 20 years of banking-domain depth (FINARIS, formerly RapidRep). Made in Europe, Frankfurt am Main.

Common questions

Regulatory reporting

Make your reporting run defensible.

In a demo we verify the result of one of your reporting runs against the business rule and show the reproducible evidence. Or start with a proof of concept.