Pillar Two Evidence and Controls: What Tax Teams Need to Retain for GloBE Reporting

Pillar Two reporting is not only a calculation exercise. Groups need to retain enough evidence to explain where the data came from, how GloBE and covered tax adjustments were made, who reviewed them, and why the final position is supportable.

Why Evidence Matters

Pillar Two calculations can look precise even when the underlying evidence is weak.

A group may be able to produce a jurisdictional effective tax rate, top-up tax position, or safe harbour assessment. The harder question is whether it can explain the result later, under review, using retained evidence rather than memory or recreated workpapers.

Evidence matters because Pillar Two reporting depends on several layers of judgement and transformation:

• source data pulled from finance, tax, consolidation, and local reporting systems
• mapping from accounting data to GloBE concepts
• adjustments to GloBE income or covered taxes
• jurisdictional aggregation and entity classification
• review and approval of positions before reporting

If those layers are not evidenced, the calculation may be difficult to defend even if the arithmetic is correct.

What Tax Teams Need to Retain

Evidence retention should be designed around the reporting chain, not around a generic folder structure.

At a minimum, tax teams should retain support for:

• source reports used for financial accounting income, covered taxes, deferred tax, and jurisdictional data
• data extracts and reporting packages received from finance, consolidation, and local teams
• mapping logic from source accounts or fields to GloBE reporting inputs
• adjustment workpapers and rationale for recurring treatments
• reconciliations between source data, adjusted data, and final reporting outputs
• review notes, approvals, and issue resolution decisions
• version history for material calculation files or reporting models

This does not mean retaining every possible file. It means retaining the files that explain the path from source data to reported output.

Controls Over Source Data

Source evidence is strongest when the source itself is controlled.

For Pillar Two, source data may come from trial balances, consolidation systems, tax provision files, local statutory accounts, deferred tax schedules, country-by-country reporting data, or manual local workpapers. Each source should have a clear owner and a defined extraction process.

Practical source controls include:

• defined source-of-record for each major input
• period and entity scope checks before data is used
• reconciliation to financial reporting or local reporting totals
• review of manual files received from local teams
• retention of data extracts used in the calculation, not only the final output

These controls help distinguish a repeatable process from a one-time data collection exercise.

Controls Over Adjustments

Adjustments are where evidence often becomes thin.

A recurring GloBE or covered tax adjustment should not be retained only as a number in a spreadsheet. It should have a documented rationale, source reference, owner, reviewer, and status.

This is where an adjustment ledger becomes useful. The ledger can track the adjustment category, amount, evidence reference, review status, and whether the treatment is manual, controlled, or ready for automation.

For material adjustments, the control question should be simple:

• can the adjustment be traced to source data?
• is the technical treatment documented?
• has the treatment been reviewed consistently across jurisdictions?
• is the evidence retained in a location that future reviewers can find?

Review and Approval Evidence

Review evidence is more than a final sign-off email.

A practical Pillar Two control framework should show who reviewed the key inputs, what they reviewed, what issues were identified, and how those issues were resolved.

Useful review evidence may include:

• review checklists for source data completeness and period accuracy
• approval records for material technical positions
• exception logs for missing data, late submissions, or local judgement areas
• evidence of reconciliation review
• documented approval of final jurisdictional results before reporting

Ownership matters here. As discussed in Pillar Two Data Ownership, tax cannot own every upstream input. But tax should be able to show that the right owners reviewed and approved the parts of the process they control.

How to Make Evidence Repeatable

Evidence retention becomes difficult when it is treated as an after-the-fact documentation exercise.

The better approach is to design evidence into the reporting calendar. For example:

• define required evidence by input category before data collection starts
• use standard file naming and storage conventions
• link evidence references directly to adjustment and calculation workpapers
• separate draft, reviewed, and approved versions
• retain issue logs and resolution notes with the period file
• review evidence completeness before final sign-off

This turns evidence retention from a scramble into a repeatable control step.

Conclusion

Pillar Two reporting needs a defensible audit trail.

For multinational groups, the practical question is not only whether the calculation can be performed. It is whether the group can show where the inputs came from, how adjustments were made, who reviewed them, and what evidence supports the final position.

Evidence and controls are therefore part of data readiness. They turn a calculation into a repeatable reporting process and help tax teams move from first-year compliance to a more sustainable GloBE operating model.