Introduction
It has been a few weeks now since the Commission has made public its decisions in the FIAT and Starbucks cases.  I understand that the Commission, the countries involved and the taxpayers are now going through the decisions themselves, agreeing on what information is too confidential to be published, before the actual – blacked out – decisions will be made public.  During this relatively lull in between, I attended a public conference in Brussels about State Aid in direct taxation on 5 November. 

In between listening to professors, people from the Commission and lawyers, I wondered: are we getting ourselves lost in technicalities and missing the bigger picture? 

We focus on “selective advantage”, we focus on “the most appropriate method” and “comparability adjustments”, we argue about “whether the royalty was too high or there should have been a royalty at all” and whether “the price of the beans was too high”.  But, we miss one question:

what is the price that Starbucks would pay a third party to roast its coffee beans for it? 

All the other questions are merely tools to answer this one question.  But hitherto, that question never got asked, let alone answered: for there to be State Aid, there have to be a tax advantage which was obtained because the Starbucks Manufacturing EMEA BV (the roaster) got compensated less than a third party coffee roaster, e.g. me, would have been paid by Starbucks for roasting coffee.  Such lower compensation would lead to a lower taxable income, which would lead to Dutch state aid.

Procurement by multinationals

The reason for my concern is that I have worked for multinationals and I have seen how they treat their service providers.  The centralised procurement department of a multinational is an organ to be feared; it will not (and should not) hesitate to squeeze the last drop of profits from providers.  Due to volume discounts, strong legal teams, an overview of the global market and the possibility of taking their business elsewhere to companies who would love to be associated with them, global dominators have awesome bargaining powers and tend to dictate, rather than be dictated.  Which brings us back to Starbucks.

The Commission said it themselves: roasting coffee beans is not rocket science, and they could not find anyone paying a royalty for know-how on how to roast coffee beans.  That would make coffee roasting a routine function in TP terms.  Without a huge inventory risk or debtors’ risk, it would make coffee roasting a low risk routine function in TP terms.  And the question is, how much would Starbucks pay a third party for such a low risk routine function? 

The Commission mentioned a reduced Dutch tax burden of € 20 – 30 million.  At a 25% Dutch corporate income tax rate, that seems to mean that the Commission believes that Starbucks Manufacturing EMEA BV earned € 80 – € 120 million less in taxable earnings than Starbucks would have paid a third party coffee roaster. 

And that is my point:  I do not believe that Starbucks would ever have paid a third party coffee roaster so much money for roasting their coffee beans, that they could walk away with € 120 million in pre-tax profit.  I have no benchmark study to support me, but I would doubt that Starbucks would leave more than € 3 – 5 million per year.

Must the adjustment be at arm’s length

The question of course is whether the adjustment the Commission makes, must be at arm’s length.  Should the Commission not just throw out the wrong transactions (too high price for the coffee beans and too much royalties for the coffee roasting) and then conclude that the resulting calculation must be the arm’s length price?  Unfortunately, the answer is no.  A TP adjustment is a two-step process:

  1. adjusting the price used by the taxpayer/government giving the tax ruling AND
  2. setting a new ARM’s LENGTH PRICE.

What counts is the result, not the method.  If the resulting adjustment does not make sense because it is not at arm’s length, then the result must be rejected, no matter how sound the method.  I.e.  we cannot afford to get lost in the technicalities of the method, if the picture it paints looks nothing like reality. 

Proving that third parties would not have paid a royalty and proving that third parties would not have paid that much for coffee beans are the first steps that grant the Commission the license to make an adjustment.  However, they say nothing about the price that Starbucks would have paid a third party for roasting coffee beans, considering Starbuck’s considerable bargaining power.   It may not be hard for the Commission to prove that the beans were overpriced or that the royalty should be ignored, BUT it will be very hard to prove that Starbucks will pay € 120 million above cost to a third party coffee roaster.  I suspect that Starbucks has several internal CUPS on third party service providers showing just how little profits they do allow third parties in this, or other fields of services.  Such CUPS would ultimately prove the Commission wrong, even when they start out being right in rejecting the royalties and the bean prices.

I look forward to see if and how the Commission papers will address this point.  I have seen little of it so far.

image_pdfimage_print

2 comments

  1. Good post. Just dissected the Starbucks case last week with our 75 transfer pricing students at the business school, combing through all available public information, and fair to say – we came up with the same result more or less.

    The overage payment booked as royalty, brought up one of the China students, is actually quite normally applied against the China OEMs by the MNEs. I pushed back but there was general agreement within the class that it is an industry standard practice. OEMs just don’t have the negotiation strength.

Leave a Reply

Your email address will not be published. Required fields are marked *