2/ Apple was a huge upset and a major blow to Commission’s plan to use state aid to combat tax avoidance. I have argued for years that the EC’s legal reasoning in these cases is deficient, so of course I’m glad Apple won. But I’m still disappointed, you’ll see why.
3/ The most important legal issue: how to allocate profits to the Irish branches of the stateless subs. If “allocation by exclusion” was okay, then Ireland would get $14B in tax. If not, Apple would keep it tax (and pay tax to the US at low rates on a slow schedule).
4/ Ireland and Apple convinced the GCEU that, under Irish law, property could not be attributed to an Irish branch for purposes of calculating its income unless the property was “made available to that branch.”
5/ Since Apple never actually made available to the Irish branches its most precious IP, it couldn’t be attributed to those branches, and neither could the related profits. Blammo, billions move out of Ireland and back to nowhere.
6/ The GCEU accepted the Irish view on branch profits allocation. It rejected the Commission’s “allocation-by-exclusion” approach to attributing profits to the Irish branches of the stateless subs. There’s a lot more legal detail there.
7/ The GCEU is unclear on the applicability of ALS. Still! Does ALS apply in all state aid cases or when in domestic law? (and let’s agree: the view that ALS applies due to a domestic law “purpose to tax net income” is the same as saying it applies regardless of domestic law).
8/ So the whole part of the case about “benchmarks” and Forum 187 clears nothing up, and the court's statment that the EC doesn’t have power to determine what constitutes “normal” tax doesn’t help when it also talks about ALS as a benchmark.
9/ EC and EU courts must clear this up, using plain language.
10/ On the authorized OECD approach (AOA): the GCEU said that the EC was entitled to use it, as a “benchmark” or “tool” to check Ireland’s’ profit attribution because there was “essentially some overlap” between Irish domestic methods of determining branch profits and the AOA.
11/ The GCEU rejected the EC “attribution by exclusion” approach, which was permitted under neither Irish law nor the AOA. The EC failed its burden to prove that the Irish branches carried out the functions and assumed the risks the EC had presumed (by exclusion) for it.
12/ Basically, the EC’s approach in Apple involved too much assuming, and not enough proving.
13/ Remember how the EC decision slammed Ireland for simply accepting Apple’s “unsubstantiated assumptions” (using that term 16 times in its final decision)? Well, assumptions are even more problematic when, like the EC, you bear the burden of proof.
14/ More: the EC can use ALS and AOA to check the MSs’ work. But mere “regrettable methodological defects” in state work (e.g., poor contemporaneous doc, or basing alloc decisions on no “scientific grounds”) are not enough to prove state aid.
15/ Bottom line: The burden of proof and the flexibility of ALS will make it hard for the EC to win any transfer-pricing state-aid cases.
16/ But the real lesson here is that the Commission’s approach is all wrong.
17/ If the EC wants to use state aid to reduce state contributions to BEPS, then instead of arguing over the details of the AOA--Berry ratios and interquartile ranges—the Commissions should target the states’ allocation rules themselves.
18/ The Irish branch alloc rule—not its specific application to Apple—was the problem. That branch alloc rule was one-sided, outside OECD rules, and facilitated BEPS. For more, see my academic paper on state aid. ssrn.com/abstract=31914… 
19/ The U.S. Supreme Court deals with analogous issues (double tax, not failure of full tax), and it has a reliable solution that leads to clear and predictable results. It could be modified to fight BEPS without throwing the tax world into chaos. See my paper for how.
Here's the (modified for state aid) US Supreme Court approach visually.
Notice, on this approach, Ireland conveyed state aid to Apple.
Notice, on this approach, Ireland conveyed state aid to Apple.
This case will be appealed. CJEU: Check out how your fellow jurists handle these issues across the pond. This approach would promote private competition, combat BEPS, and improve certainty.
