EU anti-abuse rules and beneficial ownership (IBA Finance and Capital Markets Tax Conference, 2021)
Report on session at 10th Annual IBA Finance & Capital Markets Tax Virtual Conference
Albert Collado J&A Garrigues, Madrid
Luana Foffo Ciucci Procter & Gamble, Rome
Thierry Lesage Arendt & Medernach, Luxembourg City
Jakob Schilder-Knudsen Lundgrens, Hellerup
Pieternel Verhoeven - van den Brink NautaDutilh, Amsterdam
Benjamin Twardosz Cerha Hempel, Vienna
Introduction and overview
Albert Collado introduced the audience to the relevant issues behind navigating anti-abuse rules. There is an overlap of anti-abuse rules, because of domestic anti-abuse rules - both general (GAAR) and specific (SAAR) - various tax treaties, the EU Parent Subsidiary Directive (PSD) on dividends and the EU Interest and Royalties Directive (IRD). Because of this preponderance of regulations, navigating anti-abuse rules is becoming extremely complex.
Through the introduction of base erosion and profit shifting (BEPS), tax transparency has significantly evolved because of the way in which we exchange information with tax authorities. On the other hand, tax certainty has not evolved to the same extent. Although we have the same international rules, different interpretations still occur, mainly because tax authorities have their own bias and different perspectives in each country. In this context, Collado mentioned the art word ‘VUCA’, which describes the volatility, uncertainty, complexity and ambiguity of the current tax environment.
The focus of the session was on the concept of beneficial ownership. Significant concern has been raised about the application of the concept of beneficial ownership in the Parent/Subsidiary Directive to dividends, because of the sentences handed down by the Court of Justice of the European Union (CJEU) in the so-called 'Danish cases'.
Brief reference to Danish cases and beneficial ownership
Jakob Schilder-Knudsen then briefly outlined the most important points from the aforementioned Danish cases at CJEU level (cases C-115/16, C-116/16, C-117/16 C-118/16, C-119/16 and C-299/16). In 2016 the Danish Supreme Court asked for a preliminary ruling on whether dividends and interest payments were exempt from the Danish withholding tax (WHT) when paid from a Danish company to an EU company, if these payments are passed on to a parent company that is resident outside of the EU, and whether the Danish companies are allowed to rely on the directives.
Back then, the Danish tax rules provided for a 28 per cent WHT unless the dividend was reduced under the EU PSD or a double tax treaty (DTT) with the country in question was applicable. There were no beneficial ownership requirements and no general anti-abuse rules in Danish law. The PSD also had no beneficial ownership requirement while the IRD provides that the recipient of the interest has to be the beneficial owner in order to benefit from the IRD.
The CJEU was asked under which circumstances Denmark may impose a WHT on dividend distributions and interest payments, based on the beneficial ownership concept. It was also asked if there was a need for specific domestic or DTT anti-abuse rules and if the Member State is required to determine the beneficial owner of the payments if it suspects that the recipient is not the beneficial owner.
The CJEU stated, as a general rule for both interest and dividend payments, that abuse of law is prohibited under EU law. Based on this general rule, in case of fraud, Denmark has to deny the benefits of the Directive, even if domestic law does not provide for any anti-abuse rules. Furthermore, the CJEU stated that the freedoms of the EU Treaty cannot be relied upon in case of an abuse of law. First, the proof of abuse requires that the purpose of the EU rules has not been achieved despite that they are perhaps (formally) fulfilled and secondly it requires an intention of artificially creating these conditions. Although the Member State has the burden of proof to demonstrate that an arrangement is abusive, there is no need to determine the person/company that is the beneficial owner in case of an abusive structure. The CJEU concluded as a result that the entity that economically benefits and has the freedom to use and enjoy the payment is considered the beneficial owner. The cases were then referred back to the Danish High Court, which has not yet decided on the matter but hopefully will do so in 2021.
In summary, Schilder-Knudsen said that from a Danish point of view the Danish tax authorities have applied a very strict approach, even in cases that were not similar and especially not such extreme cases as the ones referred to the CJEU.
Thierry Lesage stated that some of the Danish cases were certainly extreme and bad facts lead to bad law. If, for example, a Cypriot company without any substance is interposed just a few days before a dividend payment, this arrangement would not correspond to industry standards for Alternative Investments Funds (AIFs) at all. There are certain industry standards and a regulated environment in this sector. Since 2013, AIFs are managed by authorised and supervised managers as a result of European legislation. Typically, European AIFs correspond to a pan-European or multi-jurisdictional platform where substance is present in terms of investment decisions and management and they are therefore very different from the above fact pattern.
Lesage then continued with the main concerns of fund managers, starting with the question of whether there are safe-harbour situations. In regard to this concern, the Organisation for Economic Cooperation and Development (OECD) commentaries on the Principal Purpose Test (PPT) provide good and useful guidance and largely recognise the legitimacy of having a master holding company or holding company (Master HoldCo or HoldCo) below the fund because of the absolute prohibition on the fund manager creating tax filing obligations for the investors in the source countries.
The other question is how the exit shall be performed by the fund in view of the influence of the PPT. Instead of receiving dividends, the fund could rely on the exemption of capital gains. This used to be the case in the past. Since the Multilateral Instrument (MLI) has entered into force this is not necessarily an option any longer, as the MLI provides for the PPT irrespective of the type of income. The tax certainty is also a recurring issue. In this regard, a tax ruling in the source country would be an option. From the speaker’s point of view, however, such tax rulings are very rare. Regarding tax opinions given by practitioners, the reliability of such opinions are usually very remote due to reservations and qualifications. One could also note the increasing use of third-party insurance to cover potential tax risks.
Additionally, Lesage shared his personal experience of other countries regarding the taxation of funds: in Denmark, to his opinion, investing is complex. Representatives of the Danish government have stated that a fund would not receive the right to a source relief and investors would have to apply for a treaty refund on a look-through basis if they recognise the same category of income in the same tax year. This strict approach is completely contrary to the standards of the AIF industry and impossible to manage.
Regarding Italy, the situation is very complex but a general withholding tax exemption has been introduced for dividends and capital gains realised by European AIFs, which will certainly simplify the situation in the future in the opinion of the speaker. The Dutch jurisdiction appears manageable, because of sophisticated tax authorities, who are using industry benchmarks. In Spain, the situation tends to be comparable to the Netherlands but with a greater level of uncertainty. France is quite demanding but still manageable for those funds that respect industry standards. Germany is, in the view of the speaker, a very difficult country in a private equity environment and will become even more difficult in the future in view of pending legislation: it will be impossible for a HoldCo to get a relief of a German WHT because this option will be reserved for companies with a genuine economic activity.
Pieternel Verhoeven-van den Brink noted that the Danish cases have an effect on Dutch tax law as well. Following the publication of the Danish cases, the Dutch State Secretary for Finance published a question-and-answer document on the cases in June 2019. In the view of the State Secretary, it does not follow directly from the Danish cases whether and to what extent they have an effect on the application of the Dutch participation exemption. In addition, it cannot be concluded from the Danish cases that taxpayers are no longer entitled to (similar) benefits from bilateral tax treaties. Furthermore, the State Secretary stated that the domestic beneficial ownership concept is in line with the Danish cases and is to be interpreted in accordance with the most recent OECD Commentary.
Verhoeven-van den Brink continued by describing the effect on the Dutch safe harbour rules. Safe harbour rules come into play when there is, for example, a Dutch company with an intermediary HoldCo shareholder, which improves the dividend withholding tax position of its foreign shareholder. In such cases, the domestic dividend withholding tax exemption is granted only if the so-called ‘objective test’ is met. To meet this test, a taxpayer has to prove that the HoldCo was set up for business reasons reflecting the economic reality. In order to avoid lengthy discussions on when this would be the case, ‘relevant substance requirements’ were introduced, which would function as a safe harbour. This meant that if these requirements were met at the level of intermediary HoldCo, the objective test was considered met and the dividend withholding tax exemption was thus available. Examples of relevant substance requirements include having a wage of at least EUR 100,000 at the level of the intermediary HoldCo, at least 50 per cent locally resident directors and an office space. After the Danish cases, the State Secretary of Finance said that it could not generally be excluded that an abusive situation would exist even if the relevant substance requirements are met. Therefore, the policy needed to be changed. Currently, the substance requirements no longer function as a safe harbour, but only shift the burden of proof regarding the objective test to the Dutch tax authorities in case the ‘relevant substance requirements’ are fulfilled.
There have already been several Dutch court cases referencing the Danish CJEU cases. Verhoeven-van den Brink introduced (in a simplified manner) a Dutch lower court case that is interesting because in this case the suspicion of abuse was refuted. A Belgian HoldCo had several subsidiaries, mostly in Belgium and the Netherlands. The case revolved around the applicability of the domestic dividend withholding tax exemption in relation to a dividend paid by a Dutch subsidiary to the HoldCo. The HoldCo was considered to improve the WHT situation for the Belgian shareholders. This resulted in the suspicion of abuse. In light of the objective test, amongst other things, the following facts were considered relevant. The HoldCo was incorporated in 1991 and its shareholders lived in Belgium (that is, in the same jurisdiction). The HoldCo acquired, held and sold several Dutch and Belgian subsidiaries which could, directly or indirectly, be considered active businesses. The HoldCo also had actual influence on several subsidiaries. In addition, the HoldCo had a separate office space (in the home of its director) and incurred a significant amount of costs, for example, management and administrative costs (EUR 650,000). Therefore, the HoldCo made a plausible case that its existence had real meaning and there was an economic and commercial justification for its existence. As a result, the objective test was not considered met and the suspicion of abuse was refuted. Since the case is currently pending at the Court of Appeal, it is still to be seen whether further guidance will be provided in this grey area.
Luana Foffo Ciucci then presented a specific case of the Italian Supreme Court of 10 July 2020, in which the revenue authority challenged the WHT exemption for a ‘cascading loan’. The Italian Supreme Court referred to the Commentary on the OECD Model Convention and established the principle that the beneficial owner is the one who has the right to use and enjoy the income without any legal or contractual obligation to pass it on to another person. This specific right has to be referred to the specific income, has to be assessed on a case-by-case basis and cannot be generic. Therefore, the real autonomy of holding companies has to be validated. The Court referred to the Danish cases, according to which a conduit company cannot be a beneficial owner of the respective income. In the case at hand, the company was effectively running the financing business and was therefore considered the beneficial owner.
Beside the beneficial ownership concept, the Court also introduced the ‘Subject to Tax’ concept. In several Italian court decisions, the exemption of WHT was granted only if the dividend was subject to tax at the level of the parent company, which is clearly against the PSD. A number of other decisions referred to tax treaties. These only require that the parent company is liable to tax in the country of its residence, whereas the specific tax regime for the dividends in this country is irrelevant, which is in line with the aim of tax treaties to avoid double taxation.
This once again raises the concern regarding tax certainty. In Italy, regional sections of a court are not bound by previous rulings of the same court. Only if a decision was issued by the united section of the Supreme Court do the single sections have to comply with the published principles. Also the fact that a taxpayer cannot escalate over the Supreme Court and trigger a decision from the CJEU means there is significant uncertainty in Italy if a local decision is not in line with EU or international law. The taxpayer can only initiate infringement proceedings at EU level through the European Commission or appeal to the European Court of Human Rights. At a domestic level, a civil action for compensation of damages can be brought or, alternatively, the taxpayer can file a preventive ruling request to the domestic revenue agency regarding the correct interpretation of the EU directive.
Collado then introduced the case study regarding the beneficial ownership concept within a private equity structure. In the case presented, a fund has both a Master HoldCo and a HoldCo within the EU as well as several operating companies (OpCos) in different European countries such as Denmark, Italy, the Netherlands and Spain. In the case study, the following questions were to be answered: how are substance requirements and valid business reasons approached by source countries in this structure? How is the beneficial ownership applied in source countries in this structure? Which aspects may help in assessing the taxation of dividends, interest or capital gains in the source country?
Commentaries to the case study
Schilder-Knudsen began by describing aspects of the legal situation in Denmark. The challenge in Denmark is that since 2015, both the domestic GAAR and the beneficial ownership concept have to be considered. As already mentioned, the Danish authorities have a very strict approach in this respect. Even if the HoldCo is the beneficial owner, it would not meet the relevant substance criteria and therefore would fail to meet the requirements established by the GAAR. In case the HoldCo has substance, it still might not be considered as a beneficial owner due to the fact that either all or some of the funds are passed on in the structure. If only a part of the income is redistributed, the HoldCo may not be considered as the beneficial owner of the income. Another opinion states that the HoldCo would be considered as a beneficial owner at least of that part of the income that is not re-distributed.
Lesage continued by assessing the situation from a Luxembourg point of view and pointed out that funds like the one in the case study usually have a lifespan of ten to 12 years. Therefore, all net proceeds have to be ultimately repatriated to the fund but it is an industry standard that either the Master-HoldCo or the HoldCo have the capacity to reinvest payments received by the OpCo. Furthermore, external financing usually takes place at the HoldCo level because securities also have to be granted by the HoldCo, which directly holds the investments. Another comment is that the substance should be examined at the global level because usually the Master HoldCo will concentrate the substance of all fund activities. It may be counterintuitive or artificial to create substance at a HoldCo level. A general recommendation, therefore, would be to keep it simple and not tax-engineered. A fund structure should always be easy to explain and understand.
Verhoeven-van den Brink explained two options in the Netherlands in relation to the domestic dividend withholding tax exemption, assuming that the fund is located in the Cayman Islands. If an active business exists at Master-HoldCo level, it would, in her opinion, not be necessary to increase substance at HoldCo level, because the HoldCo would not improve the dividend withholding tax position of the Master-HoldCo in that situation. As a result, the interposition of the HoldCo should not be considered as abusive. If there would be insufficient substance at the Master-HoldCo level, you would normally advise to at least meet the relevant substance requirements at HoldCo level. Meeting these requirements at HoldCo level would shift the burden of proof regarding the objective test to the tax authorities. As a ‘worst-case scenario’, Verhoeven-van den Brink mentioned a situation in which the substance is split between the Master-HoldCo and the HoldCo in a way that neither option mentioned above would be viable.
Collado closed the panel by noting that we are living in a changing world and that today there is not yet a single version of the same truth in connection with the application of some common international tax rules, mainly because of the widely varying interpretation of different legal terms. The application of the beneficial ownership concept in connection with intra-EU dividends is just an example. At this stage, it is of paramount importance for tax lawyers to understand how countries interpret similar situations ruled out by common regulations (that is, the Parent Subsidiary Directive on dividends), potentially leading to different results.