INSIGHT: The MLI—Impact on Luxembourg: a Case Study

Luxembourg and its MLI Choices

Over the last decades, Luxembourg has developed and cemented its position as a prime holding location and a major financial center within Europe. International investors and multinational enterprises alike use Luxembourg as a platform to manage their investments and business activities in and through Europe. It is important that Luxembourg has an extensive tax treaty network that puts restrictions on the contracting states’ taxing rights.

Flexibility

The MLI provided a lot of flexibility, allowing parties:

• to choose the tax treaties that should come within the scope of the MLI;
• to opt out of (part of) provisions; and
• to choose to apply optional and alternative provisions.

While Luxembourg broadly chose not to adopt any changes in regard to its tax treaty network, the Principal Purpose Test (“PPT”) had to be adopted as a so-called minimum standard measure. In this regard, the 2017 version of the Commentary to the OECD Model is of significant importance for the interpretation of the PPT.

The PPT in General

The PPT would deny a treaty benefit (the term “benefits” includes all limitations (e.g. a tax reduction, exemption, deferral or refund) on taxation imposed on the State of source under Article 6 through 22 of the Convention, the relief from double taxation provided by Article 23 and the protection afforded to residents and nationals of a Contracting State under Article 24 or any other similar limitations; see Paragraph 175 of the Commentary on Article 29 of the OECD Model) where it is reasonable to conclude that obtaining this treaty benefit was “one of the principal purposes” (emphasis added) of any arrangement or transaction unless the taxpayer is able to establish that granting the benefit would be “in accordance with the object and purpose” of the relevant treaty provisions.

The contradictory message of the PPT is that treaty benefits are available to qualifying taxpayers unless taxpayers intend to gain from those benefits. Obviously, this injects a subjective element into every aspect of determining whether treaty benefits are available. Moreover, the PPT imposes a significant burden on the taxpayer (“establish that the granting of tax benefit would be in accordance with the object and purpose of provision in the convention”), whereas the onus on the tax administration is set at a relatively low level (“reasonable to conclude,” “one of the main purposes,” “directly or indirectly”).

The Commentary emphasizes, however, that it is important to undertake an objective analysis of the aims and objectives of all persons involved in putting that arrangement or transaction in place or being a party to it. It is interesting to note the paradox contained in this undertaking: an objective analysis is made seeking a conclusion on the subjective aims and objects of various persons.

It is further stated that tax authorities should not lightly assume that obtaining a benefit under a tax treaty was one of the principal purposes of an arrangement or a transaction. Furthermore, merely reviewing the effects of an arrangement will not usually enable tax authorities to draw a conclusion about its purposes.

Overall, the PPT creates significant uncertainty for taxpayers (and their advisors) because of the unpredictable outcomes and causes serious concerns for bona fide businesses. Holding, financing, IP management and other investment activities are all legitimate and genuine business activities that may fall within the scope of the PPT. In the view of the author, a PPT-like rule should be designed to tackle only clear-cut cases of treaty abuse in arrangements that are set up for the predominant purpose of obtaining treaty benefits.

Crucially, the Commentary limits the scope of the PPT through the statement that a purpose will not be a principal purpose when it is reasonable to conclude that obtaining the benefit was not a principal consideration and would not have justified entering into an arrangement or a transaction that has resulted in the benefit. This limitation in its first part suffers from being somewhat circular: a purpose is not a principal purpose if its consideration was not a principal consideration. However, the second part of the limitation, i.e. that if the treaty benefit would not justify entering into the arrangement, then the purpose of obtaining a treaty benefit is not a principal purpose, is clearer and may be helpful.

When it comes to the interpretation of the examples in the Commentary, it is explicitly stated that the examples are “purely illustrative” and should not be interpreted as providing conditions or requirements that similar transactions must meet in order to avoid the application of the PPT. Therefore, it cannot be construed that the PPT should apply if a particular aspect described in the examples is missing. Instead, it has to be determined on a case-by-case basis whether one of the principal purposes of an arrangement or a transaction was obtaining treaty benefits. Nevertheless, it is interesting to note that the examples in the Commentary conclude that it would not be reasonable to deny the benefit of tax treaties in case of cross-border investments unless specific facts and circumstances can be established that prompt the application of the PPT.

The PPT in an EU Context

A large part of the investments structured via Luxembourg companies is made in other EU member states. In an EU context, however, it is not only the guidance included in the Commentary to the OECD Model that is decisive for the application of the PPT. Instead, the application of the PPT should be subject to a stricter standard, which is determined by the jurisprudence of the Court of Justice of the European Union (“CJEU”).

As a rule, the fundamental EU freedoms of establishment and free movement of capital, as interpreted in CJEU case law, provide that a given structure may only be disregarded if it is proven to be a “wholly artificial arrangement” which does not reflect economic reality and the purpose of which is to unduly obtain a tax advantage. Such a purely artificial structure may be present in the case of “letterbox companies.” As early as 2006, the CJEU acknowledged in the Cadbury Schweppes case that a taxpayer is free to rely on its EU freedoms for tax planning purposes as long as the underlying contractual arrangements are not “purely artificial.”

The right of a member state to protect its tax base against abusive arrangements is limited by the fundamental freedoms. It follows that “tax jurisdiction shopping” is a legitimate activity in an internal market, even if the choice of the jurisdiction is principally based on tax considerations. Why should an investor be obliged to choose a high-tax jurisdiction or arrange its affairs in such a way as to be liable to more tax than necessary? Nevertheless, EU member states are free to protect their tax bases by way of anti-abuse rules which are exclusively directed at “wholly artificial arrangements.”

An abusive situation does not depend only on the intention of the taxpayer to obtain tax advantages (i.e. a motive test), but also requires the existence (or absence) of certain objective factors. Amongst these objective elements, the CJEU emphasized the importance of the existence of an “actual establishment” in the host state (for example, premises, staff, facilities and equipment) and a “genuine economic activity” performed by the foreign company. Here, a company may even rely on staff and premises of affiliated companies resident in the same jurisdiction.

The notion of “genuine economic activity” should be understood in a very broad manner and may include the mere exploitation of assets such as shareholdings, receivables and intangibles for the purpose of deriving what is often described as “passive” income. The nature of the activity should not be compromised if such passive income is principally sourced outside the host state of the entity.

In addition, no specific ties or connections between the economic activity assigned to the foreign entity and the territory of the host state of that entity can be required by domestic anti-abuse provisions. Therefore, insofar as the EU internal market is concerned, the mere fact that an intermediary company is “active” in conducting the functions and assets allocated to it (rather than being a mere letterbox company) should suffice to be out of the scope of domestic anti-abuse rules or the PPT in tax treaties concluded between EU member states.

It is interesting to note that, until now, national courts have not deviated from the “wholly artificial arrangement” doctrine laid down by the CJEU. While the CJEU does not seem to require an extensive level of substance, from a risk management perspective, it may nevertheless be wise to exceed the minimum standard of substance in order to limit foreign tax risks.

Serious Compatibility Issues with EU Law

As such, the PPT poses significant compatibility issues with EU Law. In fact, the PPT may deny treaty benefits on the sole ground that one of the main purposes was to obtain treaty benefits. Accordingly, even companies having economic substance in their state of residence and performing bona fide business activities may not be entitled to treaty benefits.

However, within the EU, restrictions can only be justified by the need to prevent tax avoidance when a specific anti-avoidance rule targets “wholly artificial arrangements” aimed solely at escaping national tax normally due. Considering that the PPT imposes a lower “abuse” threshold than the standard set by the CJEU, serious doubts can be raised on the compatibility of the PPT with EU law.

This concern has been confirmed by a more recent case law of the CJEU that may be helpful when analyzing the potential scope of the PPT in an EU context. On September 7, 2017, the CJEU issued its decision in the French case C-6/16 (http://src.bna.com/ArL) regarding the application of a (former) French anti-abuse provision that automatically denied the withholding tax exemption on dividends under the EU Parent-Subsidiary Directive when dividends were paid to an EU parent company that was controlled by one or more entities established in non-EU countries. Under this provision, the recipient of the dividends only qualified for the exemption if it could prove that benefiting from the exemption was not the main purpose or one of the main purposes of the structure. Accordingly, this provision was broadly similar to the PPT in the 2017 version of the OECD Model.

The CJEU ruled that the French anti-abuse provision infringed both the EU Parent-Subsidiary Directive and the freedom of establishment, as it only took into account the taxpayer’s motive for the structure. It did not, however, make an individual examination of the whole operation and it did not contain an “economic activity” (or “substance”) test as required under EU Law. In addition, the burden of proof automatically rested with the taxpayer, whereas the French tax authorities did not even have to evidence tax avoidance when denying the dividend withholding tax exemption.

German tax law also provides for an anti-abuse provision that denies withholding tax exemptions (or reductions) granted under domestic tax law or tax treaties (Para. 50d (3) of the German Income Tax Law) unless the recipient of the income complies with certain (excessive) substance requirements. On December 20, 2017, the CJEU gave its decision in two German cases (Cases C-504/16, http://src.bna.com/ArM and C-613/16, http://src.bna.com/ArN) which were joined for the purposes of the judgment.

In line with its decision in the French case, the CJEU ruled that the German anti-abuse provision infringed both the EU Parent-Subsidiary Directive and the freedom of establishment, re-emphasizing its “wholly artificial arrangement” standard. Indeed, a general presumption of fraud or abuse cannot justify either a fiscal measure which compromises the objectives of the Parent-Subsidiary Directive or a fiscal measure which prejudices the enjoyment of the fundamental freedoms guaranteed by the treaty.

Then again, on June 14, 2018, the CJEU decided on another case involving German anti-abuse legislation in its 2012 version (GS v Bundeszentralamt für Steuern, Case C-440/17, http://src.bna.com/ArO). Here, the Court held that the amended version of the German anti-abuse provision is also incompatible with EU Law. Nevertheless, as a reaction to the first decision of the CJEU, the German Ministry of Finance released in April 2018 a Circular that eliminates most of the compatibility issues highlighted by the CJEU, evidencing the importance of the Court’s case law.

Impact Analysis

The effect of the MLI on Luxembourg’s tax treaty network will broadly be limited to the adoption of the PPT. The question arises as to what will be the impact on investments structured via Luxembourg. Here, it seems reasonable to conclude that where such investments are made for legitimate commercial purposes (generating regular income, maximization of value, etc.) the PPT should generally not apply, regardless of the fact that tax implications cannot be completely neglected when investments are made. Likewise, multinational groups that implemented an investment platform in Luxembourg to manage their subsidiaries in and beyond Europe should commonly not be affected by this.

Planning Points

There remains, however, some uncertainty as to when foreign tax authorities may deny tax treaty benefits. Given the current diverging attitudes of foreign tax authorities in regard to the application of anti-abuse provisions, it might be expected that the PPT will be interpreted differently by different tax authorities. Therefore, it would be wise for taxpayers to establish the reasoning of their choice to invest via a Luxembourg company so as to be prepared for potential questions from foreign tax authorities.

In an EU context, the jurisprudence of the CJEU is particularly helpful for the interpretation of anti-abuse provisions such as the PPT. Apart from established CJEU case law, the more recent decisions on French and German anti-abuse rules confirm the Court’s adherence to its longstanding “wholly artificial arrangement” tenet which puts strict limitations on the scope of anti-abuse legislation. Ultimately, the more recent decisions of the CJEU should have a major impact on EU lawmakers and tax authorities alike, paving the way for greater legal certainty in the post-BEPS era.

Oliver R. Hoor is a Tax Partner (Head of Transfer Pricing and the German Desk) with ATOZ Tax Advisers (Taxand Luxembourg).
The author wishes to thank Samantha Schmitz (Chief Knowledge Officer) for her assistance.

The author may be contacted at: oliver.hoor@atoz.lu
www.atoz.lu