Starbucks Manufacturing BV (SMBV), based in the Netherlands, is the only coffee roasting company in the Starbucks group in Europe. It sells and distributes roasted coffee and coffee-related products (e.g. cups, packaged food, pastries) to Starbucks outlets in Europe, the Middle East and Africa.
The EU Commission’s decision challenges the outcome of the Advanced Pricing Agreement (APA) between the Netherlands Tax Authority (Tax Authority) and SMBV. The Tax Authority responded that within the Dutch tax system profit is taxed where value is created. The Tax Authority concluded an Advance Pricing Agreement (APA) with SMBV which includes an arm’s length business remuneration for the roasting of coffee beans. The Tax Authority collects taxes on profit made by SMBV for roasting coffee beans. Because the intellectual property rights of Starbucks are not located in The Netherlands, the royalties for the use of these cannot be taxed in The Netherlands.
The Tax Authority, acting in accordance with the international OECD framework for transfer pricing, agreed with Starbucks that it may apply the Transactional Net Margin Method (TNMM) to determine an arm’s length result to attach to its Netherlands based activities. The TNMM requires that members of multinational enterprises be treated as independently operating national enterprises: profits are taxed wherever value is created, attaching to the specific enterprise of the activity creating the value.
In its decision, the Commission establishes a unique interpretation the OECD guidelines concerning the choice and application of the globally accepted transfer pricing methods. Based upon its interpretation, the Commission’s alleges that Starbucks should have applied the Comparable Uncontrolled Price (CUP) method to each activity of each enterprise instead of the TNMM. However, the Netherlands Tax Authority does not agree that the CUP method should have been applied in the Starbucks case in this fashion because of the absence of suitably similar, comparable data to the situation of Starbucks’ operations and value creating activities and assets.
After its misapplication of CUP to Starbucks’ operations, the Commission then creates a new criterion for profit calculation. While the methodologies and underlying criteria of application are not a closed universe for determining an arm’s length price, the Commission’s new criterion is incompatible with domestic regulations and the OECD framework. The Tax Authority will contend that the Commission does not adequately understand the nature and context of the value add of Starbucks’ myriad of activities.
The Commission states in its Starbucks decision that the arm’s length principle it has applied is not the same as the arm’s length principle stemming from Section 9 of the OECD treaty. The Commission’s application of a variant will cause confusion and uncertainty among tax authority of member states, among trade partners’ tax authorities, and the underlying enterprises subject to their audit authority. For a tax authority, such uncertainty relates to the question of what rules are to be applied and in which fashion. And for enterprises, such uncertainly relates to the proper application of rules in rulings. So as to obtain more clarity and jurisprudence in this matter, the Dutch Cabinet has appealed the Commission’s Starbucks decision.
The Commission alleges that the methodological choices in the transfer pricing report provided by the tax adviser for Starbucks to the Netherlands Tax Authority, and agreed to in the APA between Starbucks and the Tax Authority, are not a reliable approach to a market result and thereby do not fulfil the arm’s length principle. The Commission alleges that the transactional net margin method (TNMM) is not the most appropriate method to forecast a taxable profit because the OECD guidelines and the Transfer Pricing Decree show a preference for the Comparable Uncontrolled Price Method (CUP). The Commission determined that if the CUP had been applied to Starbucks’ coffee roasting of SMBV, the taxable profit would be substantially higher.
Most Appropriate Method?
The OECD adopted in 2010 a “most appropriate method” concept, similar to the U.S. “best method rule”. The most appropriate method concept replaced the previous OECD rule that transactional profit methods, profit split and TNMM were only to be leveraged as methods of last resort (with TNMM being in last spot). Regarding the “most appropriate method” the 2010 Guidelines states:
[T]he selection process should take account of the respective strengths and weaknesses of the OECD recognised methods; the appropriateness of the method considered in view of the nature of the controlled transaction, determined in particular through a functional analysis; the availability of reliable information (in particular on uncontrolled comparables) needed to apply the selected method and/or other methods; and the degree of comparability between controlled and uncontrolled transactions, including the reliability of comparability adjustments that may be needed to eliminate material differences between them. No one method is suitable in every possible situation, nor is it necessary to prove that a particular method is not suitable under the circumstances.
However, in spite of the foregoing, the 2010 Guidelines indicate a preference for traditional methods in applying the most appropriate method rule:
[W]here, taking account of the criteria described at paragraph 2.2, a traditional transaction method and a transactional profit method can be applied in an equally reliable manner, the traditional transaction method is preferable to the transactional profit method.
The 2010 OECD Guidelines for comparability analysis contains nine, non-linear, steps.
Step 1: Determination of years to be covered.
Step 2: Broad-based analysis of the taxpayer’s circumstances.
Step 3: Understanding the controlled transaction(s) under examination, based in particular on a functional analysis, in order to choose the tested party (where needed), the most appropriate transfer pricing method to the circumstances of the case, the financial indicator that will be tested (in the case of a transactional profit method), and to identify the significant comparability factors that should be taken into account.
Step 4: Review of existing internal comparables, if any.
Step 5: Determination of available sources of information on external comparables where such external comparables are needed taking into account their relative reliability.
Step 6: Selection of the most appropriate transfer pricing method and, depending on the method, determination of the relevant financial indicator (e.g. determination of the relevant net profit indicator in case of a transactional net margin method).
Step 7: Identification of potential comparables: determining the key characteristics to be met by any uncontrolled transaction in order to be regarded as potentially comparable, based on the relevant factors identified in Step 3 and in accordance with the comparability factors ….
Step 8: Determination of and making comparability adjustments where appropriate.
Step 9: Interpretation and use of data collected, determination of the arm’s length remuneration.
What Is the Value of Starbucks Roasting “Know How”?
The Commission alleges that the payment of royalties by SMBV to the Starbucks UK subsidiary (Alki) owning the “know-how” intellectual property rights does not provide a correct representation of the value of the intellectual property rights and therefore cannot be deemed to be arm’s length. This incorrect representation led Starbucks to exaggerate the value attaching to its coffee bean roasting “know-how”, in turn leading to an excessive royalty payment.
The royalty payment is based upon an “adjustment variable”, the level of which is determined by the accounting profits of SMBV subtracting the compensation agreed in the APA in the form of a fixed mark-up on the operational costs of SMBV. The APA does contain a fixed method of being able to assess the arm’s length nature of the level of the royalties.
The Commission alleges that, on the basis of its application of an arm’s length transaction price via a CUP test, SMBV would not have been willing to pay any royalty for know-how. The Commission’s allegation is based upon a comparison of Starbuck’s agreements for roasting coffee with other coffee roasters worldwide. Thus, Alki should not have been paid any royalties. Moreover, the Commission contends that the royalties, paid over for many years, cannot be arm’s length because SMBV does not appear to gain any business advantage from the use of the intellectual property in the area of roasting coffee. An independent company, argues the Commission, will not pay for a license if it is unable to earn back the royalties paid.
Additionally, the Commission contends that payment for royalties does not represent a payment for Alki taking upon itself the risks of SMBV. The Commission dismissed the Tax Authority argument that Alki bore the economic risk of SMBV’s loss of stock (wastage). The Commission points to Alki’s lack of employees as justification that Alki’s capacity is too limited to actually bear such risk. Finally, the Commission dismissed Alki’s payment for technology to Starbucks US as a justification of its royalty payment from SMBV.
What Is the Value of Starbucks Sourcing of Green Beans?
The Commission alleges that SMBV overpays Starbucks coffee sourcing operation in Switzerland (SCTC) for acquisition of ‘green beans’, which are then roasted by SMBV and distributed to Starbucks’ various national operations. The purchase price of green beans paid by SMBV to SCTC is abnormally high and therefore does not comply with the arm’s-length principle.
The Commission alleges that Starbucks did not investigate an arm’s length relationship for which the transactions between SCTC and SMBV, being the purchase and delivery of green coffee beans. Secondly, the Commission did not accept Starbucks’ underlying grounds for the justification of the significant increase from 2011 of the mark-up in the costs for the green beans supplied by SCTC. Starbucks’ contends that SCTC’s activities became increasingly important from 2011 partly due to the evolving “C.A.F.E. Practices” program (e.g. ‘fair-trade’). Comparing the costs of similar fair-trade programs, the figures provided by Starbucks in connection with its C.A.F.E. Practices program, argues the Commission, are problematic both in terms of consistency as well as the arm’s length nature. The Commission contends that the Tax Authorities should have rejected the additional deduction from the accounting profits. Moreover, the increased mark-up can be connected directly to the losses incurred by SMBV’s coffee roasting activities since 2010, which highlights the non arm’s length relationship of this mark-up.
Least complex function
The Commission posits a secondary argument that Starbucks misapplied the TNMM to its supply chain. Firstly, the Commission alleges that Starbucks incorrectly categorized SMBV as the “least complex function” of the Starbucks’ value added supply chain, basically as a contract manufacturer, in comparison with Starbucks’ UK subsidiary that owns the manufacturing and processing “know how”. This misapplication of the TNMM led Starbucks to incorrectly led Starbucks to select SMBV as the subsidiary to be the “tested party”. Secondly, the Commission posits that when SMBV is compared to other market participants in the coffee trade sector, SMBV incorrectly applied two upward adjustments to its cost base. Consequently, Starbucks inappropriately limited its Netherlands taxable basis.
Determining the least complex function takes place prior to the application of the TNMM as transfer price method. In order to determine the entity with the least complex function, a function comparison must be made. The outcome of the function comparison indicates an entity, to which the transfer price method can be applied in the most reliable manner and for which the most reliable comparison points can be found.
In its coffee roasting function, the Commission contends that SMBV does not only carry out routine activities. SMBV conducts market research reflected by its payments for market research. Also, SMBV holds significant intellectual property reflected by the amortisation of intangible assets in its accounts. Moreover, SMBV performs an important resale function. A routine producer is not involved in such activities. On the other hand, Alki activities are very limited. Alki does not have employees and it thus operates with limited capacity. The Commission contends that the financial capacity of Alki is not the total financial capacity of the worldwide Starbucks Group.
Starbucks released a statement: “The dispute between the European Commission and the Netherlands as to which OECD rules we and others should follow will require us to pay about €20m to €30m on top of the $3 billion in global taxes we have already paid over the seven years in question (2008-2014). Starbucks complies with all OECD rules, guidelines and laws and supports its tax reform process. Starbucks has paid an average global effective tax rate of roughly 33 percent, well above the 18.5 percent average rate paid by other large US companies.
Netherlands Government Reaction?
In October the European Commission has decided that the Netherlands provided State aid to Starbucks Manufacturing. The Commission decision is placed in the context of the fight against tax avoidance by multinationals. The Dutch government greatly values its practice of offering certainty in advance. The Dutch practice is lawful and compliant with the international system of the OECD. However the European Commission’s verdict in the Starbucks case deviates from national law and the OECD’s system. In the end this will cause a lot of uncertainty about how to enforce regulations.
In order to get certainty and case law on the application of certainty in advance by way of rulings, the government appeals the Commission decision in the Starbucks case. The government is of the opinion that the Commission does not convincingly demonstrate that the Tax Authority deviated from the statutory provisions. It follows that there is no State aid involved.
OECD rules for setting internal transfer prices constitute an international standard whereby double taxation is prevented. These rules require that each transaction is assessed on the basis of the most appropriate transfer pricing method. The TNMM method can be used to establish an at arm’s length remuneration for production activities, such as those of the Dutch coffee roaster Starbucks Manufacturing BV, and is widely used internationally.
“This decision is a staggering,” says Arjan van der Linde, Chairman of AmCham’s Tax Committee and fiscal spokesman for AmCham. “By disregarding OECD rules, the European Commission is creating considerable uncertainty about the tax implications for foreign investment in the Netherlands. This has a direct effect on new investments and future employment. Uncertainty about such a fundamental component of an investment is unacceptable for many companies,” predicts Van der Linde.
He also highlights the expertise of the Dutch tax authorities, “The Dutch tax authorities have years of experience with the application of OECD rules and work thorough and carefully in considering transfer pricing requests. A separate APA practice exists. In addition, the Dutch tax authorities are consistent in their approach, with all sorts of coordination groups looking over the shoulder of the inspector. This thorough approach cannot simply be cast aside.”
Professor William Byrnes’ Reaction?
Starbucks represents the first salvo by the EU Commission to establish that it has the authority, under a State Aid premise, to step into the shoes of the national revenue authority and re-allocate profits of an enterprise according to the EU Commission’s interpretation and analysis of the arm’s length concept. American attorneys will appreciate that this is a Marbury v Madison moment of Adam’s Federalists v. Jefferson’s Anti-Federalist.
The EU Commission’s finding of a range of two – three Euro million annual difference from its own assessment of the scenario versus the assessment of the Dutch revenue authority likely reflects its trepidation to venture into the area of interposing its own judgement call for that of a sovereign national revenue authority’s arm’s length determination, especially one memorialized in an advance pricing agreement (APA) with a taxpayer. The trepidation probably results from several causes, including weaknesses of the EU Commission’s choice and implementation of an arm’s length methodology, justification thereof, and even more so, from the geopolitical ramifications of its decision.
The trepidation is exemplified by the very low adjustments the EU Commission found, after its nearly year of investigation. The adjustments are enough to be noticed by the EU state authorities and the companies, but de minimis in the context of corporate annual profits, corporate profit accumulation over time (e.g. perpetual deferral), corporate tax reserves, and de minimis in the context of revenue collection for either The Netherlands or Luxembourg.
Starbucks’ potential 30 million Euro re-capture tax bill by The Netherlands (EU Commission required), dating back to accumulation from 2008, will, assuming the tax bill stands after Starbucks’ appeal and after Starbucks’ challenge the decision up through the EU Court Of Justice, be offset by a US tax credit of like amount. Consequently, the low adjustment is a wash out, albeit could require a cash flow payment in the nearer future than the perpetual one under U.S. tax deferral accounting. 30 million Euro is too small to be noticeable to Starbucks shareholders or to the U.S. Treasury, especially when the tax credits are applied. Viewed from an annual perspective though, the two to three million Euro per annum over 10-years finding against Starbucks annual three billion dollars paid in global taxes from a global effective tax rate of 33%, it is not even a rounding error.
Had the EU Commission found, as it alluded that it is able to, that the State Aid amounted to the hundreds of millions or even billions of Euro, the intensity of the EU Commission-National government conflict would have changed, and the EU Commission would have lost that battle with the stakes so high. Fiat would have drawn Italy into the fray, to align with Netherlands, Ireland and Luxembourg. As more advance pricing agreements are challenged, more national government would align against the EU Commission. At some tipping point, the EU Commission would have to withdraw from the fight or face a bloodied nose.
Yet, more so a danger for the EU Commission, had the EU Commission’s decision been an exaggerated amount, then the U.S. Treasury would have been forced to act as if a trade war had broken out. Treasury beating up on Starbucks for transfer pricing out of the U.S. tax base is OK because Starbucks in a U.S. company, as far as the U.S. Treasury is concerned. Starbucks represents potential U.S. deficit reduction tax dollars.
Had the EU Commission decided for a large amount well beyond any tax credit relief, thus which would have represented a significant subsidy from the U.S. to EU national budgets and/or a significant subsidy from US retirement system shareholders to EU budgets, one might imagine the joint-Republican Democratic Senate hearing called by Washington state’s two Democratic senators Patty Murray and Maria Cantwell. That hearing would conclude a joint statement to Treasury demanding it report back how it intends to implement a tit-for-tat strategy against EU companies to extract an equal amount to that the EU Commission pulled from the bowels of Starbucks reserves.
Throw in enough U.S. multinationals with HQs in the various states such as New York, Illinois, California and Texas, Congress may actually in rare bipartisan stature pass tit-for-tat legislation by year end requiring Treasury to act. Perhaps a $5 billion Section 482 adjustment against each of the top 50 European companies measured by revenues. The EU would respond, and the U.S. retort, to and fro, until the weight of taxation slowed cross border investment to a trickle.
But the EU Commission instead chose to bark very loudly and withhold its bite. Probably it has avoided the worst case scenarios of political warfare presented above. With such a small award, the various stakeholders will let the appeals and ECJ process run its course before acting. The US Congress and US Treasury may not understand the Marbury v Madison moment of the EU Commission’s decision – that the “perpetual deferral” reserves of U.S. MNEs such as Starbucks, Apple, Microsoft, Google, Amazon etc, may be put “up for grabs” by European revenue authorities to fill their bloated social spending expenditure gaps (instead of flowing into U.S. investment needs or back to U.S. institutional shareholders representing our collective national retirement savings). [But Treasury has now released the below response to the EU Commission decision].
US Treasury Response
Treasury has followed the state aid cases closely for a number of reasons. First, we are concerned that the EU Commission appears to be disproportionately targeting U.S. companies.
Second, these actions potentially undermine our rights under our tax treaties. The United States has a network of income tax treaties with the member states and has no income tax treaty with the EU because income tax is a matter of member state competence under EU law. While these cases are being billed as cases of illegal state subsidies under EU law (state aid), we are concerned that the EU Commission is in effect telling member states how they should have applied their own tax laws over a ten-year period. Plainly, the assertion of such broad power with respect to an income tax matter calls into question the finality of U.S. taxpayers’ dealings with member states, as well as the U.S. Government’s treaties with member states in the area of income taxation.
Third, the EU Commission is taking a novel approach to the state aid issue; yet, they have chosen to apply this new approach retroactively rather than only prospectively. While in the Starbucks case, the sums were relatively modest (20 to 30 million Euros), they maybe substantially larger – perhaps in the billions – in other cases. The retroactive application of a novel interpretation of EU law calls into question the basic fairness of the proceedings. Fourth, while the IRS and Treasury have not yet analyzed the equally novel foreign tax credit issues raised by these cases, it is possible that the settlement payments ultimately could be determined to give rise to creditable foreign taxes. If so, U.S. taxpayers would wind up footing the bill for these state aid settlements when the affected U.S. taxpayers either repatriate amounts voluntarily or Congress requires a deemed repatriation as part of tax reform (and less U.S. taxes are paid on the repatriated amounts as a result of the higher creditable foreign income taxes).
Finally, and this relates to the EU’s apparent substantive position in these cases, we are greatly concerned that the EU Commission is reaching out to tax income that no member state had the right to tax under internationally accepted standards. Rather, from all appearances they are seeking to tax the income of U.S. multinational enterprises that, under current U.S. tax rules, is deferred until such time as the amounts are repatriated to the United States. The mere fact that the U.S. system has left these amounts untaxed until repatriated does not provide under international tax standards a right for another jurisdiction to tax those amounts. We will continue to monitor these cases closely.
Professor William Byrnes is the primary author of Practical Guide to U.S. Transfer Pricing that is used extensively by multinationals to cope with the U.S. transfer pricing rules and procedures, taking into account the international norms established by the Organization for Economic Co-operation and Development (OECD).
EU State Aid – Starbucks Webpage