Loading...
Loading...
Borys Ulanenko
CEO of ArmsLength AI
Get the latest transfer pricing insights, AI benchmarking tips, and industry updates delivered straight to your inbox.
Pillar Two (GloBE) and Country-by-Country Reporting (CbCR) are separate regimes, but they’re operationally connected because Pillar Two reuses Action 13’s €750m “large MNE” scope and—during the transition period—allows groups to use CbCR + financial statement income tax expense to avoid full GloBE calculations in “low-risk” jurisdictions via the Transitional CbCR Safe Harbour.
Two practical guardrails matter in real-world operating models:
In practice, this means your CbCR dataset is no longer only a transfer pricing transparency deliverable; it becomes a core input into global minimum tax compliance controls and audit trails. Misalignment between CbCR, consolidated financial statements, and Pillar Two data can translate directly into safe harbour failures, rework, and controversy risk.
CbCR and Pillar Two were built for different jobs:
Yet “pillar two cbcr” has become a real workflow in large tax departments for one simple reason: Pillar Two intentionally piggybacks on CbCR’s scope and data availability to reduce transition burden.
Action 13 sets a consolidated revenue threshold of €750m, noting it excludes a large majority of groups while still covering most global corporate revenue and the highest BEPS risk profiles (OECD BEPS Action 13 Final Report (2015) ¶52–53).
Pillar Two adopts the same perimeter, but operationalizes it as a “2 out of 4 prior years” test (OECD GloBE Model Rules (2021) Art. 1.1).
The OECD’s Safe Harbours and Penalty Relief package (20 Dec 2022) created a Transitional CbCR Safe Harbour that allows in-scope groups to use:
to determine whether a jurisdiction’s Pillar Two top-up tax can be deemed to be zero for that year (OECD Safe Harbours (2022) Chapter 1, Transitional CbCR Safe Harbour rule text; see also ¶3).
If you already file CbCR, you likely have the only dataset that is (a) global, (b) standardized, and (c) consistently available across all jurisdictions. Pillar Two leverages that reality—especially for early years when GloBE data models are still being industrialized.
Search queries like “BEPS Action 13 Pillar Two” reflect a common confusion: Action 13 and Pillar Two sit under the OECD BEPS umbrella, but they are not the same regime and they do not use the same calculations.
| Dimension | CbCR (BEPS Action 13) | Pillar Two (GloBE / global minimum tax) |
|---|---|---|
| Policy objective | Tax transparency for risk assessment | Ensure minimum 15% tax on profits in each jurisdiction |
| Primary output | Annual CbC Report (tables + narrative) | Top-up tax computation + GloBE Information Return (GIR) |
| Core calculation | None required (reporting data) | Jurisdictional ETR, top-up %, Substance-based Income Exclusion (SBIE), allocation of top-up tax |
| Data granularity | High-level jurisdiction totals | Detailed adjustments to financial accounts and taxes, entity-level attributes |
| Audit posture | “Appropriate use” constraints; consistency checks | Liability determination; formal penalties and documentation expectations |
| Key risk | Inconsistent story vs TP files / Master file / Local files | Incorrect ETR/top-up; safe harbour misapplication; system/control failures |
Treat CbCR as a “tax authority-facing data product” and Pillar Two as a “tax liability engine.” Your governance needs to cover both, but your calculations, controls, and documentation are different.
CbCR: The OECD chose €750m to capture the largest groups while excluding most MNEs from the compliance burden (OECD BEPS Action 13 Final Report (2015) ¶52–53).
Pillar Two: The Model Rules adopt the same perimeter but apply it if consolidated revenue is ≥ €750m in at least 2 of the 4 preceding fiscal years (OECD GloBE Model Rules (2021) Art. 1.1). That “2/4” design matters for groups near the threshold or experiencing acquisitions/disposals.
| Regime | Threshold | Lookback | Measure | Practical implication |
|---|---|---|---|---|
| CbCR (OECD Action 13) | €750m | Typically prior year (local law varies) | Consolidated group revenue | Drives annual CbCR obligation and exchange |
| Pillar Two (GloBE Model Rules) | €750m | 2 of 4 preceding FYs | Consolidated group revenue | A “near-threshold” group can become in-scope due to temporary spikes |
| EU Public CbCR (Directive 2021/2101) | €750m | 2 consecutive financial years | Consolidated revenue | Public disclosure layer; often same dataset, different presentation and sign-offs |
| US CbCR (IRS) | $850m | Annual test | Consolidated revenue | US-headed groups may have CbCR even if Pillar Two not enacted in the US |
Don’t assume “below €750m this year” means “out of Pillar Two.” The Pillar Two test is multi-year (2/4), and some groups will be pulled in due to acquisitions or consolidation changes even if current-year revenue dips.
Even with the same revenue perimeter, CbCR and Pillar Two can diverge on:
This is why a “pillar two cbcr” program needs explicit, documented mapping rules (see the data hub section below).
The Transitional CbCR Safe Harbour is the centerpiece of the practical Pillar Two–CbCR connection. It exists to reduce compliance cost during early years by letting groups “screen out” low-risk jurisdictions.
The Transition Period is defined as fiscal years beginning on or before 31 Dec 2026, but not including any fiscal year ending after 30 Jun 2028 (OECD Safe Harbours (2022), Transitional CbCR Safe Harbour definitions).
That means:
Two compliance constraints are easy to miss when you’re building a “safe harbour-first” workflow:
Domestic-law dependency note: the Transitional CbCR Safe Harbour is an OECD design feature that jurisdictions may implement with local variations (and the broader GloBE framework is always mediated by domestic law). Treat availability, elections, and filing mechanics as jurisdiction-specific “must confirm” items.
A jurisdiction’s top-up tax is deemed zero for a year if any of the following is met (OECD Safe Harbours (2022), Transitional CbCR Safe Harbour rule text):
Here’s the tests in a working table.
| Test | Data inputs | Threshold | What it tells you |
|---|---|---|---|
| De minimis | CbCR Total Revenue; CbCR Profit (Loss) before tax | Revenue < €10m AND PBT < €1m | The jurisdiction is too small to justify full computation |
| Simplified ETR | Simplified Covered Taxes (from QFS); CbCR PBT | Simplified ETR ≥ 15% (FY 2023–2024), 16% (FY 2025), 17% (FY 2026) | The jurisdiction’s tax profile is high enough to presume no top-up |
| Routine profits | CbCR PBT; SBIE amount (GloBE rules) | PBT ≤ SBIE | Profits are no more than routine return on substance |
This is why “cash tax paid” lines in CbCR are usually not sufficient for Pillar Two screening. The Simplified ETR test is built on financial statement income tax expense (with specific eliminations), not CbCR cash tax paid/accrued lines. (OECD Safe Harbours (2022), Transitional CbCR Safe Harbour definitions; see also discussion in Chapter 1 “Covered Taxes” section.)
Loss jurisdictions (one practical nuance): when a jurisdiction has negative Profit (Loss) before Income Tax in the CbCR, the Simplified ETR can become non-intuitive (because you’re dividing by a loss). In practice, loss jurisdictions often get screened via de minimis (where the absolute footprint is small) or routine profits (where PBT is ≤ SBIE) rather than the ETR gateway. Either way, you still need clean entity/jurisdiction mapping and a defensible tax expense bridge.
Safe harbour eligibility depends on whether you have a Qualified CbC Report. The OECD definition is direct: a Qualified CbC Report is a CbCR “prepared and filed using Qualified Financial Statements” (OECD Safe Harbours (2022), Chapter 1 “Source of Information” definitions).
“Qualified Financial Statements” are also defined (OECD Safe Harbours (2022), Chapter 1 “Source of Information” definitions), and include:
Practically, this pushes you toward:
A frequent failure mode is assuming “we filed CbCR” equals “we have a Qualified CbC Report for safe harbour.” If your CbCR is built on local statutory accounts in some jurisdictions, or you blend multiple accounting bases without a documented policy, you may not meet the safe harbour’s qualifying conditions.
Transitional CbCR Safe Harbour: eligibility checklist (quick, practical)
To rely on the safe harbour in a jurisdiction, confirm you can evidence:
Most groups manage three parallel but linked tracks:
| Milestone | Regime | Date / period | Source |
|---|---|---|---|
| OECD recommends CbCR implementation for FYs beginning on/after | Action 13 | 1 Jan 2016 (recommended baseline; local effective dates vary) | OECD Action 13 Final Report (2015) (implementation language) |
| First broad CbCR exchanges begin | Action 13 | June 2018 | OECD CbCR program page |
| GloBE Model Rules released | Pillar Two | 20 Dec 2021 | OECD Model Rules (2021) |
| Safe Harbours & Penalty Relief released | Pillar Two | 20 Dec 2022 | OECD Safe Harbours (2022) |
| EU transposition deadline | Pillar Two (EU) | 31 Dec 2023 | Directive (EU) 2022/2523 (Art. 56) |
| EU IIR/DMTT generally applies for FYs beginning from | Pillar Two (EU) | 31 Dec 2023 (calendar-year groups: typically FY 2024) | Directive (EU) 2022/2523 (Art. 56), subject to Art. 50 election nuance |
| EU UTPR generally applies for FYs beginning from | Pillar Two (EU) | 31 Dec 2024 | Directive (EU) 2022/2523 (Art. 56) |
| EU Public CbCR begins for FYs starting on/after | Public CbCR (EU) | 22 Jun 2024 | Directive (EU) 2021/2101 (commencement provision) |
| OECD expects first GIRs/notifications due (many groups) | Pillar Two reporting | 30 Jun 2026 | OECD GIR requirements page (compilation updated as at 22 July 2025) |
| End of Transitional CbCR Safe Harbour window (outer limit) | Pillar Two | No FY ending after 30 Jun 2028 | OECD Safe Harbours (2022) |
Build your integrated calendar around “close + provision + CbCR + Pillar Two.” The bottleneck is rarely filing mechanics; it’s getting stable jurisdictional profit and tax expense numbers with reconciliations and approvals.
A “CbCR hub” isn’t a legal requirement; it’s a systems-and-controls pattern that makes both regimes auditable and repeatable.
You need a single, governed dataset for:
Deliverables (practical):
CbCR needs revenue and profit before tax in a standardized format; Pillar Two needs financial accounts plus adjustments and covered tax characterization.
A workable approach:
A durable pillar two cbcr process uses explicit reconciliations:
When these three reconciliations are built into close controls, safe harbour becomes a repeatable screening step rather than a year-end fire drill.
Even though CbCR is a reporting regime, Pillar Two turns it into a decision dataset. Expect scrutiny over:
CbCR filings typically use standardized XML schemas and exchange frameworks (see OECD CbCR program resources). Pillar Two reporting is now on the same trajectory: the GIR is a standardized information return, and OECD materials have been updated through 2025 (see OECD GIR requirements page and related OECD GIR materials).
A hub should produce:
Don’t wait for filing year one to build schema validation. Treat XML/GIR validation rules like payroll tax validation—run them at each close cycle on draft data.
A scalable workflow separates responsibilities but keeps one source of truth.
| Activity | Tax | Finance/Controllership | IT/Data | TP/Documentation |
|---|---|---|---|---|
| Entity/jurisdiction mapping governance | A | C | R | C |
| CbCR compilation | A/R | R | C | C |
| Simplified Covered Taxes computation | A/R | R | C | C |
| Safe harbour testing | A/R | C | C | C |
| Full GloBE calculation | A/R | C | C | C |
| Reconciliations & audit trail | A | A/R | R | C |
| Filing outputs (CbCR XML, GIR) | A | C | R | C |
(A = Accountable, R = Responsible, C = Consulted)
If “Tax owns everything,” you’ll miss critical controls embedded in close (tax provisioning, consolidation journals, uncertain tax positions). If “Finance owns everything,” you’ll miss Pillar Two technical judgments (covered taxes characterization, exclusions, elections). The hub is the handshake.
Facts (Jurisdiction A):
Step 1: Compute Simplified ETR
Step 2: Compare to the transition rate (FY beginning 2024 = 15%)
What to document (minimum):
Facts (Jurisdiction B):
Test 1: De minimis
Not met (jurisdiction is clearly above €10m revenue).
Test 2: Simplified ETR
Test 3: Routine profits (PBT ≤ SBIE?)
Result: No safe harbour. Proceed to full GloBE computation.
Illustrative top-up tax math (simplified, for intuition only)
Assume:
This is exactly where CbCR stops being “just a transparency report”—the same jurisdictional totals now drive whether you do (or avoid) a full Pillar Two calculation.
Facts (Jurisdiction C, FY 2026):
De minimis test (OECD Safe Harbours (2022)) requires:
Both conditions are met:
Result: Top-up tax deemed €0 for Jurisdiction C for FY 2026.
De minimis is often the fastest “win” in a pillar two cbcr program. But it depends on correct entity/jurisdiction mapping—small PEs and holding entities can swing totals above the thresholds if misassigned.
In the EU, many in-scope groups will face both:
Important EU nuance (deferral option): under Article 50 of the Minimum Tax Directive, certain Member States (where no more than 12 in-scope UPEs are located) may elect to delay application of the IIR and UTPR for six consecutive fiscal years beginning from 31 Dec 2023. This can affect local sequencing, but it does not eliminate group exposure elsewhere (see Directive (EU) 2022/2523, Art. 50; and EU Commission Notice on the Art. 50 election).
These regimes often reuse overlapping data, but they have different audiences and governance:
Public CbCR can expose inconsistencies that were previously “quiet” (e.g., high profit in low-tax jurisdictions). If your Pillar Two safe harbour narrative depends on certain profit/tax patterns, align communications early.
The UK implemented a multinational top-up tax regime effective for accounting periods beginning on/after 31 Dec 2023, and UTPR effective for periods beginning on/after 31 Dec 2024 (UK guidance and legislation updates). For groups with significant UK presence, UK compliance calendars can drive the internal sequencing for the entire pillar two cbcr program.
As of late 2025, the US has not enacted the OECD Model Rules as such, but:
Practical implication: CbCR may be the only globally consistent dataset available across the group, increasing the value (and scrutiny) of your CbCR build process.
Even though this is a pillar two cbcr hub page, most real implementations are shaped by:
For deeper dives, see:
/blog/qdmtt-guide/blog/utpr-guideA safe harbour-first approach is not “less work.” It’s different work:
This page is the hub for pillar two cbcr planning. Use these links to go deeper:
/blog/pillar-two-guide — Pillar Two fundamentals, mechanisms, and compliance roadmap/blog/cbcr-preparation-guide — how to prepare CbCR data, controls, and common errors/blog/form-8975-guide — US Form 8975 filing mechanics and practical considerations/blog/qdmtt-guide — QDMTT design, qualified status, and operational impacts/blog/utpr-guide — UTPR exposure, allocation logic, and readiness steps/blog/transfer-pricing-documentation-guide — documentation governance that often overlaps with Pillar Two data and narratives/blog/benchmarking-study-guide — benchmarking foundations (useful context for broader BEPS/TP workstreams)/glossary/pillar-two — definition and key Pillar Two concepts/glossary/country-by-country-report — CbCR definitions, tables, and reporting logic/glossary/qdmtt — Qualified Domestic Minimum Top-up Tax/glossary/utpr — Undertaxed Profits RuleThey share the €750m scope perimeter, and Pillar Two’s Transitional CbCR Safe Harbour uses CbCR revenue and profit before tax plus financial statement income tax expense (via “Simplified Covered Taxes”) to screen out low-risk jurisdictions (OECD Safe Harbours (2022), Transitional CbCR Safe Harbour rule text).
Not exactly. CbCR is mandated under Action 13 rules (as implemented locally). But for many in-scope groups, CbCR is a practical prerequisite for efficiently applying the Transitional CbCR Safe Harbour and supporting Pillar Two data governance.
Also note: applying the safe harbour in a jurisdiction doesn’t usually remove group-wide GloBE requirements (including the GIR), and skipping the safe harbour for a jurisdiction in an in-scope year can trigger “once out, always out” for that jurisdiction (OECD Safe Harbours (2022) ¶34–35).
Because Pillar Two deliberately aligns with Action 13’s “large MNE” perimeter to focus on the biggest groups and reduce administrative burden (Action 13 Final Report (2015) ¶52–53; GloBE Model Rules (2021) Art. 1.1).
Primarily:
The Simplified ETR must be at least:
For fiscal years beginning on or before 31 Dec 2026, but it does not apply to any fiscal year ending after 30 Jun 2028 (OECD Safe Harbours (2022), definition of “Transition Period”).
Also note the sequencing constraint: if you do not apply the safe harbour for a jurisdiction in an in-scope year, you generally can’t apply it later for that jurisdiction (“once out, always out”), subject to limited exceptions (OECD Safe Harbours (2022) ¶35–36).
Generally no. The Simplified ETR test uses income tax expense from Qualified Financial Statements (with eliminations to arrive at Simplified Covered Taxes), not CbCR cash tax paid/accrued lines (OECD Safe Harbours (2022), Transitional CbCR Safe Harbour definitions and “Covered Taxes” discussion).
Yes. You still need to evidence:
Also, qualifying on a jurisdictional basis does not discharge the group from group-wide requirements (including preparing/filing the GIR and including safe-harbour-related information where applicable) (OECD Safe Harbours (2022) ¶34; see also ¶12).
The OECD notes that the first GIRs/notifications are expected to be due on 30 Jun 2026 for many groups, depending on local adoption, fiscal year-ends, and filing frameworks (OECD GIR requirements page).
They are separate directives with different purposes—minimum tax computation vs public disclosure—but they share the €750m perimeter and often reuse overlapping jurisdictional profit and tax data.
Two nuances to keep in mind:
It’s a centralized data, controls, and reconciliation layer that supports:
Start by upgrading CbCR from a “filing-only” process to a “decision-grade dataset”: