Transfer pricing in a post-Pillar Two world
A session report from the IBA’s 24th Annual US and Europe Tax Practice Trends Conference in Munich
Thursday 11 April 2024
Co-Chairs
Jenny A Austin, Mayer Brown, Chicago
Mark van Casteren, Huygens Quantitative Tax Consulting, Amsterdam
Speakers
Thomas Eisgruber, Faculty of Economics, University of Stralsund, Stralsund
Nils Harbeke, Pestalozzi Law, Zürich
Aurelio Massimiano, Maisto e Associati, Milan
Yuval Navot, Herzog Fox & Neeman, Tel Aviv
Loren Ponds, Miller & Chevalier, Washington, DC
Ocka Stumm, Gleiss Lutz, Frankfurt
Reporter
Francesco Ricci, Maisto e Associati, Rome
Introduction
The panel discussed the role of transfer pricing (TP) in the post-Pillar Two tax world, focusing in particular on the potential impact of the Organisation for Economic Cooperation and Development’s (OECD) Pillar Two rules on TP (section 1); recent developments at European Union level involving the proposed EU Directive on Transfer Pricing (commonly known as the ‘TP Directive’), recent intellectual property (IP) fact patterns (section 3) and the current trends in intercompany financing (section 4).
Panel discussion
The potential impact of Pillar Two on TP
Jenny A Austin highlighted that there are different views on the relevance of TP after the introduction of Pillar Two. Some multinational enterprises (MNEs) in the United States, for example, argue that Pillar Two, by providing a floor (minimum tax level), will make TP less relevant in the future in terms of strategic planning and, thus, it will become a compliance item no different from other tax issues. Others argue that TP will become even more important and relevant in the future, as the difference between the minimum taxation required by Pillar Two and the level of taxation in high tax jurisdictions is still large. Moreover, TP mismatches have not yet been resolved in an international context.
On this point, Thomas Eisgruber argued that there will not be many changes arising from Pillar Two, considering that MNEs generally prefer to resolve TP disputes locally. However, some issues may arise from the large amount of information that will be provided pursuant to Pillar Two.
Loren Ponds took another view. She said that any TP adjustment will have an impact on the affected constituent entities and, therefore, on the effective tax rate (ETR) calculation pursuant to Pillar Two. For example, in the US, self-initiated adjustments are allowed in regard to a taxpayer’s tax return and, therefore, TP self-adjustments may have an impact on the OECD’s Global Anti-Base Erosion (GlobE) rules, otherwise known as Pillar Two.
For Nils Harbeke, TP will remain relevant in a post-Pillar Two world because, even in cases where there are countries with the same tax rates, the tax administrations in those countries will in any case have an interest in applying TP rules to increase the profits attributable to their country.
Recent developments at EU level in regard to the proposed TP Directive
Mark van Casteren asked Thomas Eisgruber and Aurelio Massimiano to provide an update on the status of the EU’s proposed TP Directive.
According to Eisgruber, the proposed TP Directive is only a proposal at this stage, which would not in practice result in many changes to the current framework.
From a German perspective, there is not much interest in the proposed TP Directive, even in a scenario where its adoption would result in Germany having to change some of its domestic rules, such as the rules on the exchange of information. Some other additional rules could be useful, added Eisgruber, such as those relating to the treatment of compensating adjustments associated with transfer pricing.
The EU’s proposal is part of the Business in Europe: Framework for Income Taxation (BEFIT) project, which provides common rules for determining the tax base of groups of companies operating in the EU that have an annual combined revenue of at least €750m. Eisgruber expressed some concerns that the BEFIT initiative could be seen as harmful to EU Member States, as it raises the same issues already raised in the context of the common consolidated corporate tax base and would create different types of taxpayers operating within the same tax system, depending on their size.
Massimiano reported that the European Parliament adopted its opinion on the proposed TP Directive on 10 April 2024, which urges swifter implementation. However, the EU Parliament’s opinion is now to be considered by the Council of the European Union. Given that the Council of the EU takes decisions unanimously and that some EU Member States have already expressed their doubts about the measure, it is unlikely that adoption will be quick.
The European Commission’s proposal for harmonised transfer pricing rules within the EU is the result of a number of decisions by the Court of Justice of the European Union (CJEU), such as Fiat Chrysler Finance Europe v European Commission (Case C-885/19), in which the CJEU dismissed the revision made by the European Commission of a tax ruling on transfer pricing in the light of state aid rules, based, inter alia, on the absence of harmonisation in regard to the arm’s length principle (ALP) within the EU.
In terms of provisions, Massimiano agreed with Eisgruber’s opinion that the proposed TP Directive does not add much to the current framework. Even if the provisions concerning compensating and corresponding adjustments seem to be a good step forward, some concerns arise from the definition of associated entities, where the proposed TP Directive sets a threshold of 25 per cent, which may lead to a potential undue extension of the notion of control to cases where there is no significant influence on the determination of transfer prices.
Finally, the harmonisation of transfer pricing rules within the EU could lead to the resolution of transfer pricing disputes before the CJEU.
Recent IP fact patterns
Case one: IPCo acquisition and the post-closing intra-group transfer of IP
Navot described a scenario involving the acquisition of a small IP company by an MNE, where only 60 per cent of the consideration due to the founders is payable at closing; the remaining 40 per cent is contingent on the founders’ continued work for a period of four years post-closing. After closing, the target company transfers its IP to an affiliated company and has no other material assets.
Firstly, one of the main TP issues arising from this particular scenario concerns the determination of the value for the purpose of the post-closing of an intra-group transaction. Navot mentioned that there are a number of court cases that have commented on the appropriateness of the acquisition price method, as detailed by the examples in Annex I to Chapter VI of the OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (otherwise known as the ‘TP Guidelines’).
Another issue is whether the holdback consideration to the founders should be considered as part of the stock price (ie, part of the capital gain) or as remuneration for the founder for continuing to develop the IP.
According to Ocka Stumm, there are similar discussions in Germany, typically arising from merger and acquisition (M&A) deals in which startups are acquired. In these deals, the acquirer usually wants the founders to stay on board for at a least a couple of years, as they have the most important know-how and have developed the products. If a deferred purchase is contingent on the founder’s continued service at the company, a discussion must take place as to whether this relates to the purchase price (relevant for the determination of the capital gain) or is a form of compensation. In Germany, there is no clear guidance on this issue, but the worst-case scenario would be that the deferred purchase price is considered to be a form of compensation to the founder (subject to ordinary income taxation) and, from the buyer’s perspective, the payment is considered an acquisition cost, which is not tax deductible. In addition, if the transfer of the IP takes place in a short period of time after the acquisition of the target company, such a payment would be considered relevant for the valuation of the IP.
Finally, Harbeke highlighted that the main question arising from this particular scenario is what constitutes a transfer of IP from a TP perspective; also, considering that when we talk about the transfer of IP, we are mainly referring to income allocation rules that look at the entities that actually perform the development, enhancement, maintenance, protection and exploitation (DEMPE) functions associated with such IP. In this regard, if we look at the particular facts from a TP perspective, it makes sense to consider this income as part of the sale compensation.
Case two: USCo acquirer of Dutch IPCo
Ponds presented the case concerning an acquisition made by Dutch IPCo 1, a wholly owned subsidiary of USCo 1, of all the shares in Dutch IPCo 2. The total consideration for the shares in IPCo 2 is payable at closing. At the date of the acquisition, IPCo 1 conducts research and development (R&D) in the Netherlands in respect of the IP of IPCo 2. Post-closing, at the beginning of year one, USCo 1 announces a re-organisation and plans to discontinue conducting R&D through IPCo 1 and elects to treat IPCo 1 as a controlled foreign corporation (CFC) (it was previously treated as a disregarded entity for US federal income tax purposes). On 30 June of year one, IPCo 1’s lead researcher resigns and all R&D activity within IPCo 1 stops.
As of 1 July of year one, the lead researcher becomes a consultant to USCo 1, all R&D activities were performed within USCo 1, and IPCo 1 was charged for these services for the second half of year one. No formal R&D agreement was put in place between USCo 1 and IPCo 1. In October of year one, USCo 1 merges with an unrelated US entity, USCo 2, and the purchase price allocation (PPA) assigns value to the IP. USCo 2 is the surviving entity from the merger. An R&D agreement is entered into between USCo 2 and IPCo 1, reflecting the fact that USCo 2 conducts R&D and IPCo 1 continues to be charged for those activities. At the end of year two, USCo 2 enters a joint venture and contributes rights to the IP for territory X (neither in the US nor in the Netherlands). In so doing, USCo 2 recognises that it must get the IP out of the Netherlands and undertakes a valuation.
Based on the above, the Dutch tax authorities concluded that an exit tax was due at the level of IPCo 1.
The first question arising from this scenario is: when is the exit charge date: 30 June of year one or December of year two?
Ponds observed that from a US perspective, the exit charge date is deemed to be the second date, even though it could be argued that there are two different exits in this case: one from the Netherlands to the US and the other from the US to territory X.
Mark van Casteren pointed out that the Netherlands, like many other jurisdictions, could take two different approaches in this case in regard to how this IP would be considered to have been transferred.
The simplest approach would be to consider only the legal ownership of the IP and to what extent the owner is performing the DEMPE functions associated with the IP. Based on this approach, it seems that the exit date should be December of year two, considering the former value attributed to the IP by the group at arm’s length.
The other approach that is currently followed by the Dutch authorities is based on a restrictive interpretation of paragraph B of Chapter VI of the OECD’s TP Guidelines, according to which the transfer of the IP takes place at the time of the transfer of the DEMPE function. Therefore, in the case of a transfer of IP to an empty entity, the latter would be entitled to no more than a risk-free return, while the ultimate right to retain the excess profits derived from the exploitation of the IP should be attributed to the entity performing the DEMPE functions.
In any case, van Casteren concluded that tax administrations would adopt an opportunistic approach and choose the date of the transfer of the DEMPE functions or December of year two as the exit charge date, depending on which has the highest value in the specific case.
Finally, Ponds agreed with van Casteren, concluding that from a transfer pricing perspective, paying for something doesn’t mean that you are the owner of the IP.
Case three: Israeli Broadcom ruling
Navot introduced an Israeli court case concerning the use of a tax planning strategy commonly deployed to avoid corporate taxes on the sale of IP. In this case Broadcom acquired 100 per cent of the shares of a target company. Immediately after the acquisition, the Broadcom group entered into several agreements: (1) the first agreement concluded between the target and Broadcom affiliates concerned the licensing of legacy IP (royalty rate based on affiliates’ sales), R&D services regarding the development of new IP (cost plus) and marketing and support services; and (2) a second intercompany IP sale agreement, according to which the target sold its IP for less than the acquisition cost.
Based on these facts, the Israeli tax authorities recharacterised the transactions, arguing that the transfer of functions, assets and risks from the target to other group entities constituted a business restructuring under Chapter IX of the OECD’s TP Guidelines, representing a sale at the closing date, which should be compensated based on the acquisition price method and subject to capital gains tax.
The target, on the other hand, argued that full validity should be given to the agreements between the parties under which it did not sell anything, but rather granted a licence in exchange for royalties and switched to a model of providing services on a cost-plus basis. In addition, one of the main challenges for the taxpayer was to demonstrate that there was a clear separation between ‘legacy’ and ‘new’ IP.
The Court accepted Broadcom’s defence and ruled in its favour.
Referring to this particular scenario, Stumm noted that similar discussions are taking place in Germany as to whether any transfer of IP constitutes a transfer of a business function triggering exit taxation. In this regard, recent case law has stated that a transfer occurs to the extent that there is a transfer of a business function, including the associated opportunities and risks, as well as assets and other benefits, to the transferee, so that it can exercise that function or expand an existing function.
However, this will not normally be the case if you simply transfer ownership of the IP without transferring the associated DEMPE functions.
Current trends on intercompany financing
Case one: intercompany financing
Van Casteren introduced the issue of determining the arm’s length interest rate for a loan granted by a group entity (in this case ‘Group FinCO’) to an associated entity, without any guarantee or other form of support (comfort letter, etc) from the parent entity. In such a case, the first question is whether implicit support should be taken into account when determining the interest rate on the loan from Group FinCo. The position of the Dutch tax authorities is to attribute to the borrower the same credit rating as the parent entity solely by virtue of group affiliation (ie, there is implicit support).
Stumm added that there have been a number of court decisions in Germany over the past decade on the question of how implicit support should be taken into account in determining the arm’s length interest rate. In previous years, some decisions have taken the view that the collateralisation of the loan claim already lies in the possibility of influence that a controlling shareholder has over the corporation; therefore, the mere group affiliation itself constitutes a form of collateral. More recently, some court decisions have taken a different view, stating that, even if the absence of loan collateral is unusual, an unsecured intragroup loan may be considered at arm’s length if, having regard to the facts and circumstances of the case, an unrelated party would have granted the loan under the same conditions, potentially considering risk-related compensation.
With reference to the US perspective, Austin reported that the position of the Internal Revenue Service (IRS) is similar to that of the Dutch tax authorities, without taking into account any comparability factors that may have an impact on credit ratings.
Case two: intercompany non-interest bearing loan
Massimiano illustrated a recent Italian Supreme Court decision (judgment No 7361/2024), in which the Italian parent companies, Alfa, Beta and Gamma, granted non-interest bearing intra-group loans to their subsidiaries, Delta and Alfa (which had solvency problems). The Italian tax authority challenged the inconsistency of the zero interest rate applied in terms of the arm’s length rate, arguing that the loans should bear interest set at the market price.
Based on these facts, the Italian Supreme Court ruled, consistently with recent CJEU decisions (C-382/16 Hornbach-Baumarkt and C-558/19 Impresa Pizzarotti), that in the case of cross-border non-interest bearing intra-group loans (or, more generally, in the case of intragroup loans), taxpayers may alternatively prove that: (1) the intra-group conditions do not differ from those that would be agreed between independent enterprises or (2) if this is not the case, that there are ‘commercial reasons’ behind the absence of an interest charge (in line with previous Supreme Court decisions, such as judgment Nos 998/2024 and 13850/2021).
In the case at hand, an important aspect is the role assumed by the ‘commercial reasons’ surrounding the decision. From the first point of view, ‘commercial reasons’ should be interpreted in line with the arm’s length principle and, therefore, interpreted as a useful argument to support and justify the consistency of the intercompany price applied with the arm’s length principle (eg, to recharacterise a purported loan as a contribution to equity capital).
However, the Italian Supreme Court seems to take a different view, interpreting ‘commercial reasons’ as outside of the scope of the arm’s length principle. In particular, in line with the CJEU’s ruling in the Hornbach-Baumarkt and Impresa Pizzarotti cases, the Italian Supreme Court seems to grant to the taxpayer the right to demonstrate that the disapplication of the arm’s length principle was determined by the existence of a prevailing group interest (as in the case of subsidiaries with solvency problems).