INSIGHT: Transfer Pricing Cases of 2019 (1)

Notwithstanding the continuous improvements to the model treaty, its commentary and the Transfer Pricing Guidelines, they are insufficiently detailed to prevent a large number of transfer pricing matters coming to the courts each year. Only there can the application of the arm’s-length principle in specific situations be resolved, often with the benefit of expert testimony, which may become a reference for practitioners in other jurisdictions. It is then for the guidance to catch up with these landmark decisions, if a sufficient measure of international agreement can be reached.

2019 featured an interesting mix of transfer pricing cases, both in national and international courts, and in relation to transfer pricing between entities and to the attribution of profits to permanent establishments. Thirteen cases are summarized in this article. The issues considered (sometimes repeatedly across the cases) included:

• the commercial rationality of changes to commercial arrangements,

• changes in transfer prices linked to changes in commercial arrangements,

• the significance of similar arrangements being observable or not observable in the open market,

• the recharacterization of transactions,

• incidental benefits,

• the possible existence of an unrecognized transaction,

• the potential influence of ‘economic circumstances’ comparability differences,

• transfer pricing for treasury companies,

• the choice of simpler party for benchmarking purposes,

• the allowable extent of simplification and standardization when applying the transfer pricing guidelines,

• short term and longer term losses,

• the submission of inconsistent evidence to the court,

• the allowable differences between comparables and the tested party,

• the separation of transactions into elements to be priced separately,

• the relevance of transfer pricing reports for related parties in other jurisdictions,

• the attribution of a different capital structure to a permanent establishment,

• the inclusion of stock options costs in cost sharing agreements (again), and

• whether the definition of intangibles should be limited to assets which could be transferred separately.

Some conclusions are included at the end of this article.

Australia

The decision in the case of Glencore Investments, in the Federal Court of Australia, was made on Sept. 3, 2019. The issue was that the taxpayer had switched to a new commercial relationship with its parent company in 2007 which the tax administration argued was an irrational thing for it to have done given its characteristics, and notwithstanding that the new terms were ones which could be found in contracts between independent parties.

The court held that there was no authority in Australian law or in the applicable Organization for Economic Cooperation and Development (OECD) Transfer Pricing Guidelines (of 1995) to recharacterize a transaction, and that the motive or commercial prudence for altering a commercial relationship is irrelevant as long as it matches a real world transaction (which did not need to be very closely comparable). It therefore set aside the income adjustments.

The Australian law and the OECD Guidelines have changed since 2007, and it is likely that a different decision would be reached if such a change in the commercial arrangement were to be made now. However, the decision confirms the benefit of matching related party arrangements to ones which can be found between independent parties, and of producing industry experts who can confirm this in court.

Denmark

The decision in the case of Microsoft Danmark ApS, in the Danish Supreme Court, was made on Jan. 31, 2019. The issue was not the level of commission earned by the local sales subsidiary, but whether this commission should also have been earned on a second set of sales for which the subsidiary had no contractual responsibility. Specifically, the taxpayer was responsible for marketing the software to the public, while another subsidiary was responsible for marketing the software to computer manufacturers. The tax administration argued that marketing the software to the public made them more likely to purchase computers that came loaded with the software, so that the taxpayer should receive an additional commission on those sales.

The court ruled that any effect on computer sales of the marketing of the software to the public was an ‘incidental’ benefit and as such the guidance of the OECD Transfer Pricing Guidelines was that no charge should be made.

This decision confirmed the benefit of a functional analysis which clearly defines the role of the taxpayer, and the potential benefit of reference to certain brief discussions in the OECD Transfer Pricing Guidelines.

The July 4, 2019, decision in the case of Anon, Eastern High Court, was published on Oct. 28, 2019. The issue was whether the increase in the royalty rate to be paid by the licensee taxpayer was supportable.

The tax administration challenged the increase because it resulted in a loss for the taxpayer. The court concluded that the local market was price-driven and self-contained so that the IP could have little value there. It also took the high marketing expenditure of the taxpayer to be a sign that it received little value from the licensed intangibles. Furthermore, the court noted that the group’s IP-related expenditure had actually fallen after the royalty rate had been increased. Moreover, there appeared to be no commercial reason for the losses because the revenue of the taxpayer had remained high. The court suspected that the taxpayer was kept going to satisfy the requirements of the group’s global clients. For all of these reasons, the court found in favor of the tax administration.

This decision confirms the willingness of tax administrations around the world to run the standard argument that there must be another unrecognized transaction when a subsidiary continues to report losses, despite the frequency with which this argument has failed in the past. Here the taxpayer lost because it failed to recognize the obvious conclusions of its own market and functional analysis for its transfer pricing, it failed to link the change in transfer price to any real commercial change, and it did not take advantage of the possibility to present ‘economic circumstances’ comparability explanations for its losses.

European Union

The decision in the case of Fiat Finance and Trade, in the EU General Court, was published on Sept. 24, 2019. The issue was whether a Luxembourg tax ruling permitted a non-arm’s-length calculation of the taxable profit of the taxpayer. The taxpayer was a group treasury company and the ruling allowed it to calculate its profit for tax purposes on the following assumptions: that the amount of equity, which it would put at risk, could be determined by reference to the regulatory capital ratio for banks; that this equity could be segmented into three parts corresponding to three types of activity of the taxpayer which involved different amounts of risk (namely lending, other treasury-related activities and participations in foreign treasury company subsidiaries), and that the return on this equity could be calculated through the Capital Asset Pricing Method (CAPM) in the case of lending, external comparable uncontrolled prices (CUPs) in the case of other treasury activities, and by reference to the dividends received from the participations (which, having been taxed in the jurisdictions of the subsidiaries, should not be taxed again in Luxembourg).

The General Court ruled that the capital of the taxpayer could not be segmented because it is “fungible,” that it was insufficiently precise to refer to the regulatory capital ratio, and that the benchmarks used in the CAPM calculation were not sufficiently comparable. It concluded that a single rate of return should have been earned on all the taxpayer’s equity and this should be a return based on an article about the banking sector.

The decision in the case of StarbucksManufacturing EMEA BV, in the EU General Court, was also published on Sept. 24, 2019. The issue was whether a Netherlands tax ruling permitted a non-arm’s-length calculation of the taxable profit of the company. The taxpayer was the producer of roasted coffee beans for Starbucks coffee shops throughout the region. It purchased the beans, roasted them using group technical IP (for which it paid a license fee), and sold them to coffee shops. The concern of the Commission was that the license fee was varied so as to extract all but a routine margin of profit from the taxpayer, on the basis that it was the simpler party. The Commission concluded that, by licensing the IP and by being at the heart of the business, the taxpayer was the entrepreneur not the simpler party, and that there was no evidence of license agreements between independent parties involving a similar way of calculating the license fee. For these reasons, the Commission concluded that the tax ruling had granted illegal state aid.

The approach of the Commission in these two cases has the following aggressive and debatable features:

• the restriction on state aid can be applied to taxation as well as to expenditure;

• the separate, ‘normal’ tax law of each member state must always have included the arm’s-length principle by implication;

• this arm’s-length principle can be applied by international transfer pricing guidance, even if it has no legal authority in a member state;

• today’s international guidance should be referred to, not that in existence at the time of the tax ruling; and

• all tax rulings or advance pricing agreements which include non-arm’s-length pricing must involve the ‘selectivity’ that is necessary to sustain a state aid challenge.

The decision of the General Court in the Fiat case is unfortunate in that it failed to recognize the need for simplifications in the application of the arm’s-length standard to high volumes of transactions, particularly to complex financing activities. The taxpayer had applied a simplified version of the Luxembourg financial intermediary transfer pricing Circular, which was itself a state-of-the-art document at the time, and in a way which was standard among transfer pricing advisers. The discussion of “fungibility” and the choice of the key rate of return benchmark were also flawed, and it is fortunate that the decision has been appealed by the taxpayer to the European Court of Justice.

However, the decision of the General Court in the Starbucks case (which was against the Commission) confirms that the onus is on the Commission to show that the outcome of a transfer pricing calculation differs significantly from the arm’s-length result, rather than merely asserting that a different method of calculation should have been used. It also confirms the importance of the existence or otherwise of similar open market commercial arrangements on the decisions in transfer pricing disputes, and the need to confirm rather than merely assert that one party is the simpler one and the only one to be tested for transfer pricing purposes.

Italy

The decision in the case of Anon, in the Regional Tax Commission of Lombardy (judgment 928/20/2019) was made on Sept. 5, 2019. The issue was the continuing losses of the taxpayer, which the tax administration argued were caused by the failure of the rest of its group (which was profitable) to pay for local marketing services which the taxpayer was performing for it on its behalf. The court rejected this argument because there was no firm evidence for any such service being provided. It also accepted the taxpayer’s explanations of the external economic factors behind the losses.

This decision again confirms the willingness of tax administrations to run the argument that there is another unrecognized transaction when a taxpayer continues to report a loss, and also the continued inconsistency of the courts as to how to respond to this argument. In this case the court used the (relatively) sophisticated tests of whether there were economic circumstances comparability factors to explain the losses and whether there was sufficient evidence of the alleged unrecognized transaction, rather than it being merely asserted.

Luxembourg

The decision in the case of Anon, in the Administrative Court (No. 42043C), was made on July 17, 2019. The issue was the interest rate on a shareholder loan. As there was no support for the interest rate, the tax administration imputed a much lower one. The taxpayer then produced two ex post transfer pricing reports, the second of which supported the interest rate which had been used.

The Appeals Tribunal had ruled earlier that, as the two reports used different methods and reached different conclusions, it was not possible to accept either one of them. Furthermore, the taxpayer had not explained why the tax administration’s calculations were incorrect, and indeed the first report showed that the tax administration’s interest rate was in the arm’s-length range.

The Appeals Court ruled that the second report was unreliable because of comparability differences relating to the different size and simplicity of the taxpayer compared to the benchmark borrowers. It felt that the high interest rate was not consistent with the low risk presented by the taxpayers’ real estate property.

This case illustrates the need to carry out benchmarking exercises before setting related party prices, to be consistent in what is presented to the court, and to provide an effective technical critique of the tax administration’s benchmarks and calculations.

Norway

The decision in the case of Dynamic Rock Support AS, in the Borgarting Court of Appeal, was made on March 19, 2019. The issue was the allocation of a purchase price between the intellectual property (IP) of an acquired company and the shares in its operating companies. Although the shares were purchased after the IP, the tax administration argued that the two transactions should be viewed as being part of a single transaction, i.e., the acquisition of a company, and that the prices agreed between the parties for the two elements did not need to be respected for tax purposes.

The taxpayer determined a price for the IP by taking a present value of the royalties which were agreed for similar IP in the open market. The tax administration observed that this value was low in relation to the price paid for the whole company and that this greatly reduced the tax payable on the transaction. The correct approach was to allocate the purchase price rather than to price an element of the purchase without regard to the value of the whole company. On this basis, the correct approach was to value the simpler element (the operating companies) and to allocate the residual to the IP.

The court found for the tax administration, deciding that this was an example of transactions being so closely linked that they cannot be priced adequately on a separate basis, as noted in the OECD Transfer Pricing Guidelines. Again, reference to a short discussion in the OECD Transfer Pricing Guidelines had a powerful influence in court.

Spain

The decision in the case of IKEA Distribution Services, in the National Court of Appeal, was made on March 6, 2019. The issue was that the net margin of the company fell below the taxpayer’s own calculation of the arm’s-length range in a particular year. Notwithstanding that the average margin for the taxpayer was still in the arm’s-length over a run of years, the court ruled that an income adjustment should be made because the margin should have been in the range in every year. However, the court accepted that the arm’s-length range could be determined by reference to the average results of benchmark companies over a run of years, and could then be applied for a run of years.

A secondary issue was the size of the income adjustment. The tax administration argued that the taxpayer’s range was not reliable because of a comparability difference between the size of the benchmark companies and the size of the taxpayer—presumably, a suggestion that there were economies of scale in its distribution activities. In local law, this unreliability could mean that the income adjustment should be to the median rather than to the bottom of the range. The court ruled that the size differences were not sufficient to make the taxpayer’s arm’s-length range unreliable, so that the adjustment should only be to the bottom of the range.

This decision confirms the willingness of courts to consider economic circumstances comparability differences, and the need to test sets of comparables for any patterns linked to size, etc. We also note that IKEA is the subject of a Netherlands transfer pricing ruling state aid investigation by the European Commission, which may lead to an EC decision and an appeal to the General Court of the EU in 2020.

Sweden

The decision in the case of Absolut Company AB, in the Swedish Supreme Administrative Court, was published on June 19, 2019. In this case, the issue was that the net margin of the taxpayer’s U.S. subsidiary was above the arm’s-length range in one year. The tax administration therefore adjusted the taxpayer’s income upwards.

The court accepted that the subsidiary’s margin had been affected by unusual market conditions (i.e. ‘economic circumstances’) and that it was significant that the subsidiary’s margin had been reduced in the following year to ensure that its average margin was in the arm’s-length range across the two years. On this basis the court decided that it was not appropriate to adjust the taxpayer’s margin.

This case illustrates the danger of having transfer pricing reports in a group which state that the pricing is not consistent with the arm’s-length principle, even if it results in a higher margin for the local taxpayer—such documents may well be requested and taken into account in audits and disputes in other jurisdictions.

United Kingdom

The decision in the case of Irish Bank Resolution Corporation Limited and Irish Nationwide Building Society, in the Upper Tribunal, [2019] UKUT 0277 (TCC), was published on Oct. 9, 2019. In this case, the issue was the amount which the branches of two banks could deduct for their notional payment of interest to their head offices. The tax administration argued that U.K. law required the equity and loan capital of the permanent establishments to be assumed to be what they would be if they were standalone entities with the same credit rating as their head offices, and this was different from the amounts in their books of accounts. The taxpayers argued that the relevant paragraphs of the model treaty referred to the permanent establishment as it is and not to a hypothetical permanent establishment with a different capital structure.

The First-tier Tribunal had decided that there was no inconsistency between the domestic law and the treaty.

The Upper Tribunal ruled that the treaty permitted the rectification of a permanent establishment’s book of accounts in the manner proposed by the tax administration (i.e. with less debt and hence less interest to pay).

United States

The decision in the case of Altera Corp., in the U.S. Court of Appeals for the Ninth Circuit, was issued on June 7, 2019. The issue was whether stock options costs should be included in a cost-sharing agreement. The U.S. Tax Court had found for the taxpayer on procedural grounds. The Court of Appeals overturned this decision, confirming that stock options costs should be included as costs for transfer pricing purposes (the situation having been rather confused by the earlier withdrawal of a similar judgment by the same court in 2018). This brings the US position into line with that in Israel, where the jurisdiction’s first transfer pricing decisions, beginning with Kontera Technologies Ltd. and Finisar Israel Ltd. (Supreme Court, April 22 2018), were quick to establish that the use of stock options in the Israeli software development sector should be treated in the same way. However, the fact remains that there is little evidence in either jurisdiction of independent parties wishing to enter into such agreements.

The decision in the case of Amazon.com, Inc. in the Ninth Circuit, was issued on Aug. 16, 2019. The issue was the buy-in payment made by a foreign related party for pre-existing intangibles in the context of a cost sharing arrangement. The IRS had allocated all non-routine profit to the intangibles, assuming thereby that goodwill and going concern were automatically transferred alongside the intellectual property. However, the court ruled that only assets which could be transferred separately could be intangible assets, and therefore that the buy-in payment should rightly have been limited to specific intangibles. This decision indicates the willingness of courts to refer to specific definitions in the OECD Transfer Pricing Guidelines as opposed to general concepts in national statute.

Themes Arising

The decisions in the courts in 2019 indicate the following approaches to the transfer pricing issues involved:

• The commercial rationality of changes to commercial arrangements appears to be raised more often than in earlier years, and while reference to even roughly similar arrangements between independent parties can be persuasive in court, there is a risk of a follow-up challenge that it was not commercially rational or necessary for the taxpayer to change to the new arrangement. The issue of the relative bargaining power of the taxpayer can become relevant at this point.

• It may be proposed that additional income should be imputed to a taxpayer because of ‘incidental’ benefits which it has provided to a related party, but such a challenge can be headed off by pointing to the scope of the related party agreement, the presence of another party which is formally charged with providing the services which are an accidental outcome of the taxpayer’s role, and to relevant text in the OECD Transfer Pricing Guidelines.

• Particularly where losses have been reported for a run of years, it is common for an unrecognized transaction to be imputed, especially a marketing service, or possibly filling a gap in service coverage to allow contracts to be won from global customers. In such cases the taxpayer’s functional analysis could be crucial, and may well not be helpful.

• Royalty rates seem to be more prone to challenge on the basis that the IP has little value in the local market, as evidenced by the need for the taxpayer to devote a great deal to marketing, and (unfortunately) by the description of the local market in the taxpayer’s own transfer pricing report. There needs to be consistency between global royalty rate setting and local transfer pricing reports.

• Transfer pricing reports from other jurisdictions may be used to show that the local taxpayer has not earned a sufficient income. It should be assumed that reports which happily conclude that a taxpayer has been undercharged will be read in other jurisdictions and are not ‘good’ reports.

• Transfer pricing for treasury companies and their loans, guarantees and cash pooling policies continues to be challenged and is harder to defend in the absence of detailed guidance in this area (at the time of writing). The courts do not yet seem to appreciate the complexity and volume of these transactions and the need for simplifying approaches.

• As would be expected after the BEPS project, the choice of simpler party for benchmarking purposes is now very open to challenge, and the contribution of all the members of an MNE group will be reviewed. It should be remembered that if the MNE’s marketing literature talks of the group’s uniquely valuable IP, that contribution will not be treated as routine. The same issue arises in valuation matters.
• Along the same lines, optimistic attempts to separate transactions into separately priced elements are unlikely to be sustained in court.

• Courts are willing to accept evidence of comparability differences between the taxpayer and potential benchmarks, but the impact of these differences needs to be quantified.

• Courts continue to struggle with the attribution of profit to permanent establishments, especially the attribution of a different capital structure to a permanent establishment in the financial sector. This may be made more difficult by inconsistencies between local statute and the relevant treaty.

This column does not necessarily reflect the opinion of The Bureau of National Affairs, Inc. or its owners.

Author Information

Danny Beeton is of counsel in the London and Luxembourg offices of Arendt & Medernach. He can be contacted at: daniel.beeton@arendt.com.