Case Overview
In Skatteministeriet v. EET Group A/S (Østre Landsret, 19 June 2024), the Eastern High Court upheld the Danish National Tax Tribunal’s transfer pricing approach for income years 2010–2012. The High Court accepted a gross-margin (RPM-like) outcome test for the foreign sales companies and held that (i) the arm’s length range could be narrowed to the interquartile range (IQR) and (ii) if a result was outside that range, the adjustment could be made to the nearest point in the range (here, the upper quartile/third quartile) rather than the median. (BS-6035/2021-OLR; court summary: domstol.dk)
Key Facts
- Business: EET Group A/S (Denmark) purchased IT components from independent suppliers and resold them to affiliated foreign sales subsidiaries for local resale.
- Years: 2010–2012.
- Initial adjustment: SKAT increased EET’s taxable income by DKK 128,810,000 (2010: 17.72m; 2011: 65.95m; 2012: 45.14m) under Ligningsloven § 2 (arm’s length principle).
- Administrative appeal: The National Tax Tribunal reduced this to DKK 29,587,135 (2010: 0; 2011: 2,670,294; 2012: 26,916,841).
- Litigation posture: The Ministry sued seeking to restore SKAT’s higher adjustments (an additional DKK 99,222,865 vs the Tribunal’s outcome). EET counterclaimed for zero adjustments in 2011–2012.
The Dispute
1) Documentation and “discretionary assessment”
A key issue was whether EET’s transfer pricing documentation was so deficient that it could be treated as missing, allowing SKAT to apply a discretionary assessment under the documentation rules then in force (Tax Control Act provisions referenced by the courts). Østre Landsret held it was not.
2) Method: gross margin vs net margin
EET’s files described the method as TNMM, but the actual test compared gross margins of the sales subsidiaries to independent distributors (closer to the Resale Price Method outcome test). SKAT/Ministry argued a net-margin TNMM/EBIT test should be used and that EET’s approach was not reliable.
3) Range construction and “adjustment point”
A second (and heavily debated) issue was how to use statistical ranges from database comparables:
- Full range vs IQR: whether to accept the full spread of comparable results, or narrow to the IQR to mitigate unquantified comparability defects.
- Median vs nearest point: if an adjustment is needed, whether it should go to the median (a common Danish practice cited in commentary) or merely to the nearest boundary within the accepted range.
Holding (Østre Landsret, 19 June 2024)
Østre Landsret:
- Rejected the Ministry’s claim for further taxable-income increases and upheld the National Tax Tribunal’s result (meaning the remaining DKK 29,587,135 increase stayed in place).
- Found EET’s TP documentation not so defective that it could be equated with missing documentation, so SKAT could not rely on discretionary assessment powers on that basis.
- Accepted a gross-margin comparison as an appropriate basis for the arm’s length analysis, notwithstanding the documentation’s imprecise “TNMM” label.
- Found no basis to set aside the Tribunal’s discretion in:
- narrowing to the IQR, and
- adjusting out-of-range results to the nearest point in the arm’s length interval (here, the upper quartile/third quartile) rather than the median.
(See Østre Landsret summary: domstol.dk.)
Subsequent history (Højesteret, 21 May 2025)
The Supreme Court later fully acquitted EET and removed the remaining 2011–2012 income increases. Importantly for range mechanics, Højesteret held there was no legal basis for automatically limiting the arm’s length range to the IQR, and that being outside the IQR was not, by itself, sufficient to prove non-arm’s length pricing—especially with a limited number of comparables. (BS-35371/2024-HJR; summary: domstol.dk)
Takeaways for Practitioners
- Method consistency matters: align the profit level indicator with the business model and the controlled transaction (EET’s cost-plus pricing and a gross-margin outcome test for sales companies).
- Documentation quality can be decisive: documentation that is imperfect is not necessarily “missing”; that distinction affects SKAT’s ability to use discretionary assessment powers.
- Adjustment point is not automatic (at least at Østre Landsret stage): the High Court upheld a move to the nearest point in the accepted range rather than the median.
- Update after Højesteret (2025): don’t assume the IQR is the legally correct “default” arm’s length interval in Denmark; the tax authority must still prove non-arm’s length pricing on the evidence, and statistical tools require caution with small samples.
Related Content
For background on statistical narrowing and range construction, see: IQR vs full range. For core concepts referenced in this case, see Median and Interquartile range (IQR).
Frequently Asked Questions
Q1. What did Østre Landsret say about adjusting to the median?
Østre Landsret did not endorse a default median adjustment. It upheld the Tribunal’s approach: if a tested result was outside the accepted interval (here, the IQR), the correction could be to the nearest point within that interval (here, the upper quartile/third quartile).
Q2. Why did Østre Landsret accept the IQR instead of the full range?
It deferred to the Tribunal’s view that database comparables often contain residual comparability defects and that narrowing to the IQR could mitigate the impact of unidentified differences (referencing OECD guidance in the underlying reasoning). However, Højesteret (2025) later held there was no legal basis for automatically limiting the arm’s length range to the IQR.
Q3. Was EET’s transfer pricing documentation treated as “missing”?
No. Østre Landsret held it was not deficient to the extent it could be equated with missing documentation—restricting SKAT’s ability to use discretionary assessment powers. Højesteret (2025) likewise found no basis to treat the documentation as significantly deficient.
Q4. Did the courts require TNMM at net-margin level?
No. Østre Landsret accepted a gross-margin-based comparison. Højesteret (2025) also found the Ministry had not substantiated that a net-margin TNMM was required on these facts.