Loading content...
Loading...
Loading...
In Skatteministeriet v. Microsoft Danmark ApS (Supreme Court, 31 January 2019; SKM2019.136.HR), the Danish Supreme Court rejected a transfer pricing adjustment that would have increased Microsoft Denmark’s taxable income by about DKK 308 million for 2004–2007. The case is widely cited in Denmark for (i) the high threshold for making a discretionary/estimated assessment based on transfer pricing documentation issues, and (ii) the evidentiary burden when the tax authority alleges that local marketing created compensable value tied to a separate sales channel.
The Ministry argued that Microsoft Denmark’s marketing in Denmark increased demand among Danish consumers and therefore increased the group’s revenues from OEM licences embedded in PCs sold into Denmark by multinational PC manufacturers. In that view, an independent marketer would not accept a remuneration model that excluded those sales from the commission base.
A second (procedurally important) dispute concerned whether SKAT could make a discretionary/estimated assessment because Microsoft Denmark’s transfer pricing documentation was allegedly late and/or deficient—in particular, for not addressing the multinational-OEM channel.
The Supreme Court upheld Microsoft Denmark’s position (majority 3–2).
No discretionary assessment based on the documentation record.
Under the then-applicable rules (Tax Control Act, documentation provisions and the link to the discretionary-assessment rule), the Court held that the relevant point in time is the time of assessment. Discretionary assessment is only available if TP documentation is missing at that time, or so materially deficient that it must be equated with missing documentation. Here, the authorities had received documentation at the time of assessment, and it was not deficient to that level—especially because the exclusion of the multinational-OEM channel from the remuneration base was clear from the documentation.
The Ministry did not prove non-arm’s-length remuneration (Ligningslov § 2).
On the merits, the majority found it unproven that Microsoft Denmark’s remuneration was not arm’s length. Reading the MDA in light of the remuneration mechanism (commission on MIOL’s Denmark-invoiced sales), the majority understood the agreement as targeting products where sales were made directly in Denmark (package licences and OEM licences to Danish manufacturers).
While the Court did not exclude some “derived” marketing effect (notably around the 2007 launch of Windows Vista), the interaction between Microsoft Denmark’s marketing and other group companies’ arrangements (including subsidies/marketing support to OEMs) was not sufficiently clarified. The Court also noted it was unlikely that an independent marketing company would be able to obtain commission on OEM-licence sales concluded abroad without its involvement.
Dissent: Two judges interpreted the MDA more broadly (including multinational-OEM sales into Denmark), but they also criticised key assumptions in SKAT’s calculation approach.
For practical guidance on intangible-related pricing and substantiation, see Documentation for Intangibles. For the foundational legal standard applied throughout the case, see Arm's Length Principle.
Q1. What was the controlled transaction being tested?
Microsoft Denmark’s remuneration for its Danish marketing/sales-support functions under the MDA, calculated as a tiered commission on specified sales (with a cost-plus floor).
Q2. Was this a “royalty case”?
Not primarily. The dispute was about whether a Danish marketing entity’s service remuneration should be increased by reference to additional software-licence revenues embedded in a separate OEM sales channel.
Q3. When can Danish tax authorities make a discretionary TP assessment based on documentation issues (as in this case)?
Under the Supreme Court’s approach for the years at issue, only if TP documentation is missing at the assessment time, or is so deficient that it does not give the authorities a sufficient basis to assess arm’s-length compliance (i.e., it is effectively equivalent to missing documentation).
Q4. What did the Court say about “derived marketing effects”?
The majority accepted that indirect effects were possible but held the Ministry had not shown their extent or net impact—especially given other group entities’ arrangements with multinational OEMs that could also influence demand and channel outcomes.
Q5. How can software groups use this case in practice?
Use it to (i) stress-test whether intercompany agreements and documentation clearly reflect which licence channels drive remuneration, and (ii) resist adjustments grounded in broad “group benefit” theories unless the authority can present a coherent method and evidence quantifying incremental value attributable to the tested party.