Concept and Implementation of CFC Legislation

S. 2321. Introduction

1.1. Topic

The objective of this thesis is to define the concept of taxable income for the purposes of the CFC regime. The ATAD Directive devises, among others, a specific methodology aiming to identify the taxable income of a CFC (entity or PE). This methodology is based on the classification of predefined types of income and is therefore called the “categorical approach”. The purpose of the thesis is also to explore whether the specific classification of income contemplated by the categorical approach is consistent with the aims of the directive. Each individual class of income may, in fact, be defined and taxed differently in each Member State, thus giving rise to tax planning phenomena. Consequently, the classification of CFC taxable income is a very critical issue for taxpayers, tax authorities, and tax advisors.

1.2. Specific purpose and scope

This thesis attempts to provide an answer to the following research questions:

1.3. Approach and research method

The approach of this thesis is primarily that of a legal study. However, being a master thesis, it also has a practical and operational slant. It also aims to assess how the categorical approach has been implemented in some significant Member States, such as Germany, Spain, Austria, Denmark, Finland, and the Netherlands. In addition, it intends to analyze the categorical approach in light of the tax policy goals of these selected Member States. These countries have provided particularly significant examples of the implementation of the categorical approach. Anyway, this thesis does not adopt a pure comparative method, and reference to the above-mentioned countries is made as practical exemplifications. A substantial part of this thesis is based upon literature research.

S. 2331.4. Outline of the thesis

The following chapters try to address the issues above. Chapter 1 will set out the theoretical rationale for the introduction of a specific taxation aimed at affecting the income of CFCs. Chapter 2 will analyze the legal genesis and the international background of the categorical approach. Chapter 3 delves into the income classification of the categorical approach and focuses on each single class of income. Chapter 4 will explore the relationship with the primary law of the European Union and its implications.

2. CFC taxable income under ATAD

2.1. The definition of taxable income

According to Preamble number 12 of the ATAD, the main objective of the CFC regime consists in “re-attributing the income of a low-taxed controlled subsidiary to its parent company”. In order to efficiently achieve this aim, it is essential to define what is meant by “income” for CFC purposes, how to delimit this concept so as not to be so broad as to entail a risk of double taxation that undermines economic competition, and whether there are possible justifications for such constraints.

The ATAD definition of taxable income for CFC purposes stands as an attempt to combine BEPS recommendations on one hand, the legacy of the ECJ settled case law and the EU primary legislation on the other hand. In addition, political priorities of the Member States also play an important role. In order to meet various requirements, the ATAD provides for three different methodologies to address the definition of income. As briefly represented in Preamble number 12 of the directive, CFC rules may target:

2.2. Taxable income and CFC rationale

The three methodologies that the ATAD uses to determine income as stated in Preamble number 12 underlie the three rationales of the CFC regime as highlighted by the relevant doctrine on ATAD.

2.2.1. Anti-deferral rationale

The option of including the entire low-taxed entity’s (or PE’s) income in the taxpayer’s tax base is consistent with the anti-deferral rationale. The system of full inclusion of CFC income (also called the entity approach) has been adopted in those countries where the anti-deferral motive is prioritized. As an anti-deferral measure, a CFC regime favours capital export neutrality (CEN) by placing foreign investments on the same level of taxation as those existing in the parent company’s jurisdiction. However, capital import neutrality (CIN) which considers the tax treatment of investments from the point of view of the controlled entity’s jurisdiction may be undermined since CFC income could be taxed in both jurisdictions.

The “entity approach” (full income inclusion) was adopted to determine CFC income under the first proposal of Article 8 ATAD.

The tax base of a taxpayer shall include the non-distributed income of an entity when the following conditions are met:

The rule provided for a subjective requisite and an objective requisite that triggered the CFC regime. The approach based on specific income categories represented the third step of the objective requisite. So, if those requisites were met, the entity approach applied, and the entire non-distributed income of the entity (or PE) was subject to taxation in the hands of the controlling company. According to this approach, also called “piercing the corporate veil”, the taxpayer and the subsidiary are deemed to be one sole entity from a tax point of view, and the income earned by the CFC is taxed directly in the hands of the shareholders. However, this structure was not considered efficient to pursue the objectives foreseen in ATAD Preamble number 12 and was subsequently set aside.

2.2.2. Anti-abuse rationale

From a CIN perspective, in order to not undermine the tax competitiveness of investments abroad, in some cases, tax jurisdictions of parent companies may permit tax deferral. Many countries have reached a compromise solution between pursuing CEN without undermining tax competition on investments abroad (CIN protection). In this perspective, CFC rules as anti-deferral rules are adopted as long as they are limited to specific categories of passive income while active income is taxed by the parent jurisdiction only in the case of distribution. Hence, tax deferral is allowed only for active income so as to not impair the CIN in the foreign countries of which the subsidiaries are resident, and passive income is taxed under the CFC regime.

2.2.3. Anti-tax avoidance rationale

The ATAD enriches the CFC rationale with a third perspective: the “fight against tax avoidance and aggressive tax planning, both at the global and EU levels”. So, S. 236CFC income definition is also supposed to be consistent against tax avoidance that is generally achieved through the creation of “artificial” economic activities in the jurisdiction of the entity or the permanent establishment within the Union and in third countries as specified in Preamble number 12.

3. The boundaries to the concept of income

3.1. Two flexible options

In accordance with the objectives outlined above, the concept of income is designed as a flexible one so as to address the priorities of each individual Member State. Moreover, income should be an easy-to-determine concept in order to limit “administrative burden and compliance costs” while maintaining business competitiveness. To this end, the EU legislator has designed Article 7(a) ATAD on CFC income according to two approaches:

The categorical approach consists of allocating the entity’s (or PE’s) income to the taxpayer provided that income falls into certain categories set out in a given list and regardless of any actual distribution. It is based on the assumption that certain types of income, named “passive income”, lend themselves as an instrument for tax base stripping, especially parent stripping or foreign base stripping and time deferral of taxation. Passive income is produced without an intense rooting of capital, assets, risks, or functions in the foreign jurisdiction. In other words, it has no or low “tax nexus” with the foreign jurisdiction. For those reasons, passive income can be easily displaced from one jurisdiction to another without significant economic barriers or economic side-effects. Passive income stems from functions, assets or risks for which legal ownership can be legally shifted from one jurisdiction to another. Thus, they can be legally relocated for taxation purposes. Passive income is not the antithesis of active income as they may be interconnected and, in any case, passive income comes from economical activities. Moreover, many CFC rules may also target active income for mere tax deferral purposes.

S. 237According to the transactional approach, the simplest units of the tax base analysis are the single transactions or arrangements between the parent company and the controlled entity or permanent establishment. If certain characteristics or requisites arise, those items trigger CFC income.

Table 1

S. 2383.2. The background to the categorical approach

The idea of using the categorical approach as the fish-net to catch “passive income” within CFC rules goes back many years. The United States was the first country to attempt to elaborate on “passive income”. In 1962, it introduced the CFC rules in their Internal Revenue Code, Subpart F. A previous tax rule, dated 1937, aiming at taxing foreign personal holding companies where individuals could avoid US tax by getting passive income through controlled foreign companies inspired CFC rules. However, provisions have never specified any definition of passive income. They have only provided a classification for it since the United States considered that the mere listing of income categories was a very effective “tool for ensuring current taxation in the US of certain types of highly mobile income, including passive income earned by foreign corporations with majority US ownership” and to counter tax deferral. As an example, on the subject of sheltering income with non-cash losses from businesses in which the taxpayers do not participate (IRC, Sec. 469(c), 1986 and subsequent), the tax code provides three types of gross income:

According to the given definition, “Income is passive activity gross income if generated by activities that involve the conduct of a trade or business in which the taxpayer does not materially participate or by rental activities”. Income from intangible assets, such as patents, copyrights, and literary compositions, is excluded from passive income if the taxpayer has spent significant personal efforts in the creation of those assets. Moreover,

On the subject of CFC taxable income, Subpart F rules include the following categories:

The item FPHCI includes eight categories: (a) dividends, interest, rents, royalties and annuities, (b) gains from certain property transactions (e.g. investment assets, stock and partnership interests), (c) gains from commodities transactions, (d) foreign currency gains, (e) income equivalent to interest, (f) income from notional principal contracts, (g) payments in lieu of dividends, and (h) income from certain personal service contracts. Capital gains should not be considered passive income if they are derived from the disposal of assets producing active income.

The OECD did not fundamentally criticize the outcome of the categorical approach as applied in the United States as a tool to reattribute income of a low-taxed controlled subsidiary to its parent company and it has even supported it in some respects. Thus, the US experience served as a litmus test for other experiences in other countries and, in some respects, also for the BEPS Project and, consequently, for the elaboration of the categorical approach in the ATAD Directive.

3.3. Pros and cons of the categorical approach

ATAD lawmakers revealed a bias in favour of the categorical approach for the following considerations. Firstly, it represents a fair compromise between the safeguarding of CEN (capital export neutrality) and CIN (capital import neutrality). Secondly, it directly targets those types of income that international doctrine and established ECJ jurisprudence would consider as “passive” income. Thirdly, it allows for a relatively simple and cost-effective implementation. The great advantage of the categorical approach based on a legal classification and the reason why it has been chosen in many jurisdictions, although applied in different ways, is the simplicity in its assessment and application. However, the categorical approach has the disadvantage that it targets some items of income irrespective of the fact that they can derive from sources that have genuinely produced that income from a substantive or active economic activity performed in the other state. Thence, the risk of double taxation arises although the relief methods might be applied.

S. 2404. The categorical approach: a legal classification

Article 7(2)(a) adopted the categorical approach as one of the two best methods to reattribute the undistributed CFC income. Article 7(2)(a) mirrors BEPS Action 3 which outlined certain items as “passive income” in accordance with international practice. The BEPS recommendations propose a classification based on the legal definition of income as further elaborated by the ATAD. While BEPS only provides a “recommendation” for which items should be included in the list for the categorical approach, the ATAD provides a list of income items that constitute a “minimum standard” from which member countries may deviate and set provisions that are more rigorous in order to achieve the objectives of the directive. Anyway, in both cases, the list of income items is not exhaustive and is left open so that member countries can select or remove items as long as they are characterized by the following hallmarks:

4.1. Article 8(2)(a) in the “presidency compromise”

The intermediate elaboration of the ATAD, named the “presidency compromise”, revised the previous framework for Article 8(2)(a) on CFC income. The entity approach was abandoned in favour of the categorical approach which is elevated as one of the two preferred methodologies aiming at “re-attributing the income of a low-taxed controlled subsidiary to its parent company”. Article 8(2)(a) S. 241featured the list of income categories as follows which was thereby confirmed in the last final version as Article 7(2)(a) ATAD.

The list of categories does not consider income from immovable property and real estate anymore which had appeared in the previous version. In the initial ATAD proposal, non-distributed income from immovable property was included as a CFC tax base unless the Member State of the parent company would not be entitled to tax the income on the basis of a tax treaty. Income from immovable property may have specific characteristics such as the absence of active economic activities involving the participation of the tax-payer that make it suitable for inclusion in the list of passive CFC income. However, the real estate income may not raise concerns of tax avoidance and profit stripping from a low-taxed controlled subsidiary to its parent company due to the immovability of the tax base. Moreover, it was argued that the management of real estate is per se an active and genuine business. That is why immovable property and real estate income was removed from the CFC income list under the categorical approach. However, Spain has taken a different approach. Article 100.3 a) Spanish corporate income tax (CIT) provides that income derived from the “entitlement to rustic and urban real estate or real rights that fall upon them” is included in CFC income as the first category of tainted income unless real-estate income derives from a business activity (unless it is assigned to a business activity). Since Article 7(2)(a) is a “minimum standard” allowing the Member States to further its requirements, Spanish CFC domestic law will not cause compatibility problems with ATAD rules. According to the German approach based on an explicit list of active income so that passive income is determined by exclusion from active income, rental income from S. 242movable and immovable property is considered passive income unless it is earned in the context of an activity carried out on a commercial basis under section 8(1) number 6 AStG.

4.2. Article 7(2)(a): the final version

In the final version of Article 7(2)(a), the ATAD adopted the enumerative technique so that this provision includes a list of categories of income that are allegedly considered as passive income and, as such, targeted to be considered within the scope of the CFC rules. Based on this approach, passive income can be divided into four main areas according to the specific nature of the enumerated items (see Table 2). The four areas are the following:

If we compare the BEPS version to the ATAD legal income classification, the following table comes out. Article 7(2)(a) expands the formal wording and adds additional items such as income from the disposal of shares, income generated from financial assets, income from banking and other financial activities, income from financial leasing for which inclusion was already suggested by the BEPS Project, and income from invoicing companies.

Table 2: Legal classification

4.3. Dividends

According to the ATAD itemization, dividends are considered passive or tainted income and therefore included per se in CFC income. “Dividend in its economic sense is the return on an equity investment in a corporate entity. The directive does not offer a definition of dividends, the scope of which remains very general and thus very flexible. It requires the inclusion of dividends irrespective of whether they are portfolio or non-portfolio dividends nor does it consider any other kind of characterization. Non-portfolio dividends and PE profits usually benefit from tax exemptions in the parent company’s residence state and thence can raise concerns about double non-taxation or very long deferral from domestic taxation which ultimately turns into avoidance of domestic tax on income. On the other hand, the directive does not transpose certain concrete situations that do not give rise to any concerns according to the BEPS Project and would therefore be worthy of exclusion from a CFC scope. The cases are the following ones:

As it is rather easy to shift equity, dividends could be consequently shifted as purely “passive” income (i.e. income that does not arise from any underlying activity). The assumption that dividends are passive income has been interpreted differently among the EU countries. Germany has adopted a unique legislative techS. 244nique consisting of deriving passive income from a negative assumption that passive income is any type of income that is not explicitly stated to be active income. Section 8(1) AStG contains the catalogue of active income and also selective exceptions. Germany considers dividends as active income provided that some requisites occur: the subsidiary is not located in a low-tax country (tax rate of less than 25 per cent), dividend derives from designated positive activities, and it is not a hybrid payment. In detail, Germany has adopted a “reverse” entity approach by treating all income of the subsidiary company as harmless for CFC purposes, provided that income derives from specified activities listed in section 8(1) AStG as “active” income. On the other hand, if income derives from activities that are not included in the active catalogue, then it triggers CFC legislation. Besides, in some countries, dividend income from fourth- or lower-tier companies automatically qualifies as passive income even if the dividends originated from companies that are engaged in active trade or business. The inclusion of dividends among targeted passive income for CFC treatment has at least two undesirable consequences especially when it comes to holding companies that regularly make dividend distributions. Assume that a group consists of a parent company in one state and a holding company in another state that, in turn, controls operating companies. Firstly, the same holding company would be subject to CFC rules for the years in which it distributes dividends to the parent company and not for the other years. So, a holding can be treated as a CFC in one year but not in another year with a consequent increase in administrative burdens. Moreover, those dividends would be subject to the CFC regime irrespective of whether they result from genuine economic activities but would be taxed in the hands of the parent company in ''mass'' terms. “This will result in classification and tracking issues (i.e., contaminated and non-contaminated income need to be distinguished)”. Secondly, in the case of multiple tiers' companies for which a dividend is distributed by the third-tier and then by the second-tier company, there is a sound risk that dividend is taxed twice as a case of double taxation, which is a negative effect that should be eliminated. A solution to this issue could be the implementation of a “switch-over” clause or the exclusion from taxation of distributions by lower-tier subsidiaries so that only the ultimate recipient (the parent company) can be taxed on dividends.

S. 2454.4. Income from the disposal of shares

Capital gains deriving from the disposal of shares for which dividends are qualified as passive income usually will remain qualified as passive income. However, capital gains arising from the sale or disposal of shares in companies that primarily conduct an active business may be considered as active income. An issue to be addressed is the relationship between the CFC regime and the capital gains exemption regime. In many countries such as Germany, Sweden, Austria, Spain, and Romania, income from the disposal of active shares is exempt according to domestic taxation. Therefore, in the implementation of a CFC rule, some countries such as Austria have added the remark: “to the extent that these (income) would be taxable for the participating corporation”. In other words, the CFC regime applies if the capital gain income is taxed in the state of the parent company. If, on the other hand, the jurisdiction of the parent company considers such income as exempt, then there is no reason to subject it to the CFC regime, otherwise it would be taxed under the CFC regime even though it is entitled to exemption. In some other jurisdictions, if the seller is a credit institution or financial service institution that does not qualify for the domestic participation exemption, then the income from the disposal of shares is considered passive income and subjected to CFC rules. In these circumstances, double taxation can arise. On the other hand, as a tool to avoid double non-taxation, a CFC regime also covers income from the capital gain of the disposal of shares provided that the capital gain is taxed in the jurisdiction of the parent company that acknowledges the tax credit, if any, for the tax paid in the CFC country. In the United Kingdom, there is a general exclusion for all types of capital gains whereas an increasing number of countries, such as Denmark, exempt capital gains arising from the disposition of an IP.

4.5. Interest or any other income generated by financial assets

Most of the countries that have adopted the categorical approach of Model A (Italy, Portugal, Spain, Greece, Croatia, Slovenia, Austria, Romania, Poland, Czech Republic, Germany, Denmark, Lithuania, and Sweden) have included financial income as an item of the passive income list. Interest and financing income is assumed to be passive income as the parent company can easily shift it to the low-taxed CFC by providing it the necessary financial resources in order for the CFC to grant back loans or other financing services (like letters of patronage, S. 246guarantees, etc.) worthy of remuneration. In general, interest and financial income is treated as passive income if financial services do not represent genuine financial activities in the CFC jurisdiction. However, the approach is different for banking and insurance activities performed by financial institutions. In general, these activities are genuine economic activities, bound up with local regulations being location-based activities. Therefore, they generally enjoy the non-application clause provided for in Article 7(2)(a) which contains an exclusion from CFC regime for substantially economic activities. According to the Spanish implementation, income from lending, financial, and insurance activities between related parties is included in the CFC regime if it generates a deductible cost for the resident parent company. Additionally, income from derivative instruments is included in CFC income except from those derivatives covering risk deriving from an economic activity. According to German CFC rules, the rising and lending of funds is active income if the taxpayers “prove that the CFC raises the capital exclusively on foreign capital markets and not from the taxpayer or an associated person“. In order to protect intra-group treasury activities or group finance centres, the CFC rules may provide for interest exclusions. As an example, the Spanish regulations provide that the positive income derived from the transfer of capital to third parties shall be understood as coming from the carrying out of credit and financial activities. This is valid unless the transferor and the transferee belong to a group of companies irrespective of residence and the obligation to draw up consolidated annual accounts as long as the revenue of the transferee derives at least by 85% from the exercise of economic activities. This behaviour would possibly lead to the overleveraging of the parent and overcapitalization of the CFC without an economic justification in both jurisdictions. It was noted that the categorical approach on interest could also be combined with a “look-through rule” that would consider interest to be active finance income if deductible by the payor against its active business income. Accordingly, this rule would treat inS. 247terest paid out of active earnings as active in order to exclude that interest from CFC rules. However, it was also noted that such a rule could raise foreign-to-foreign base stripping issues.

4.6. Income from insurance, banking, and other financial activities

The analysis of this item involves distinguishing between entities engaged in financial activities in a structured manner or regulated entities (such as banks, financial institutions, etc.) and other entities or non-regulated entities. In general, taxable income from financial, insurance, mortgage, or banking activities is included in passive income. If such income is exempt, then it should not be included in the CFC income. Income from insurance, banking, and other financial activities is considered per se as CFC income on the assumption that the income arising from the insurance of risks can be shifted away from jurisdictions in which those risks are located and into a low-tax jurisdiction. In cases of organized financial services, such as banks or insurance companies, income may not raise the same concerns because of the regulatory environment settings in terms of risks, capital, and remuneration. According to the Danish CFC approach, subsidiaries that qualify as regulated entities in terms of financial institutions under Article 2(5) ATAD are excluded from the CFC regime if no more than one third of the CFC income stems from operations with group related parties or associated persons. Additionally, CFC rules apply if the value of the subsidiary’s financial assets exceeds 10% of the value of the subsidiary’s total assets (asset test). In other words, they qualify for an exemption from CFC regime if they collect financial income in the context of a structured activity that has commercial substance as demonstrated by exceeding certain parameters. Interestingly, the list of Danish passive income also includes taxable gains and deductible losses on CO₂ quotas and CO₂ credits as well as taxable gains and deductible losses on debt claims, debt, and financial contracts. According to some thinking, if derivative contracts (forward contracts, etc.) serve as hedging instruments for active transactions, they should not be included in passive income.

As to regulated entities (such as banks) that are subject to capitalization regulations and other requirements, rules designed to attribute CFC income should recognize that any rules on overcapitalization should take account of such requirements and should not attribute CFC income just because an entity is required to maintain a certain level of capital for non-tax purposes (such as for credit, market, and operational risks).

S. 2484.7. Income from financial leasing

Many operating lessees take “ownership” of the asset via finance leases (rather than outright purchase). Contracts of finance leases can be legally placed in low-tax countries so that, under the dislocated finance leases, the operating lessor will record income that will be regarded as lease rentals receipts. On the other hand, the parent company might transfer assets to the CFC via loan agreements at low rental fees. To avoid disruptions in the leasing industry and to ensure that the lessor is both the legal and also the economical owner of the assets under financial leasing, income from financial leasing is regarded as CFC passive income. Countries such as Austria, Croatia, and Denmark have included financial leasing and subleasing (Spain) including gains and losses from the sale of assets used in connection with financial leasing in the list of passive income.

4.8. Royalties or any other income generated from an intellectual property

Income from royalties or any other income generated from intellectual property is considered per se as CFC income on the assumption that IP assets are highly mobile. The income from these assets can easily be shifted from the location where the value of the assets was created to low-tax jurisdictions. According to BEPS Action 3, IP income raises several CFC concerns because the IP income is particularly easy to manipulate. An IP can be exploited, and related income can be distributed in many different forms, all of which may have different formalistic classifications under the CFC rules of different countries. For instance, income from the IP could be embedded in income from sales and therefore be treated as active sales income under the CFC rules. Moreover, IP assets are often hard-to-value assets because of the difficulty in finding comparable assets, and their cost base may not be accurate for measuring the income they can generate. In addition, in the case of an IP that is embedded in other goods or services, it is often difficult to separate the income directly earned from the underlying IP asset from the income that is earned from associated services or products. CFC rules that use a categorical analysis based on legal classification often attempt to address the concerns raised by IP income by separating royalties and treating them as attributable using OECD transS. 249fer pricing guidelines (2017). BEPS Action 3 makes reference to income generated from some sort of intellectual property (IP) as underlying the income from digital goods and services. As shown in Pillar 1 and Pillar 2, income from digital goods escapes the classification under income from intellectual property and is to be detected using completely different tools. A very broad definition of income from IP can be seen in the list provided by Denmark which indicates “payments of any kind received as compensation for the use of or the right to use intangible assets as well as gains and losses connected to the transfer of intangible assets” as passive income for ATAD purposes. This is a very broad definition covering any income derived from the exploitation of an IP at the subsidiary level. Some jurisdictions provide for the exclusion of royalties from passive income if intangibles have been developed at the entity (or PE) level. Spain does not include royalties in passive income as does the United States which considers royalties as active income depending on the organizational setup related to the local exploitation of the IP. According to the German approach, royalties are considered as active income if the IP is developed at the subsidiary level. In this case, they are excluded from the CFC provisions pursuant to section 8(1)(6) AStG.

4.9. Income from invoicing companies that earn sales and services income from goods and services purchased from “and” sold to associated enterprises “and” add no or little economic value

This item is one of the most intricate and controversial legal definitions both because of its wording and its complex content. The international tax debate has often focused on the need to avoid that, through the transfer of sideline activities, income from sales and services attributable to the parent company’s jurisdiction is shifted to a CFC’s low-tax jurisdiction. Article 7(2)(a) picks the “income from invoicing companies that earn sales and services income from goods and services purchased from and sold to” other entities if the following two requisites occur:

The text of the article states that companies that purchase from “and” sell to their associated companies fall under the scope of the provision whereas companies that purchase from “third parties” and sell to group affiliates, or the opposite, should not generate a passive income. In other words, CFC income applies if the transactions have both a “cross-border” element (buying from abroad or selling abroad) in a transaction of goods or services “and” both legs of the transaction with a related party (so called “invoicing companies”). As illustrated below, in case 1, the proceeds fall under the scope of CFC income; in case 2, proceeds cannot be considered as CFC income.

Table 3

Moreover, the provision refers to invoicing companies that “add no or little economic value“. The question arises as to how to identify the extent of the “value added in a transaction”. The EU promoted a Joint Transfer Pricing Forum (JTPF) to analyse intercompany services that are routinary in nature, i.e. do not generate added value for the parties involved. The EU Forum stated that the concept of “added value” is to be interpreted in relation to the nature of the services rendered, S. 251the provider, and the beneficiary. The purpose of the EUJTPF analysis was to verify the criterion for determining the compensation agreed upon for the services included in the service agreement, the advantage obtained by the company receiving the service, the fairness of the compensation, and the effectiveness of the services. However, the EUJTPF list of services is very extensive and far-reaching. Rather, for the purposes of Article 7(2)(a), one can resort to the OECD Transfer Pricing Guidelines (TP Guidelines) (2017) which identifies low (or no) value-adding trading activities under the application of the arm’s length principle. In detail, TP Guidelines contain the definition for what the international best practice defines as “low value-adding services”, namely:

• Auxiliary services;

• Services that are not part of the multinational entity’s core business;

• Services that need no intangible asset and do not contribute to their creation; and

• Services that do not involve an assumption or control of significant risk by or give rise to significant risks for the service provider.

The above list has the advantage of being concise, practice-oriented, and straightforward to apply. Although the application of the criteria above might be useful to the purpose, real cases still imply subjective evaluations. Some authors have described some particular cases which I find particularly interesting as they well represent the difficulties for the implementation of this income category. If we think of the operations performed by the service providers, distributors, contract, and toll manufacturers, then “it could be disputable whether a service or activity can be deemed “auxiliary”, “non-core” or “non-risk bearing”. In other words, services or activities may or may not be deemed low value-adding depending on the specific context and circumstances”. Let us consider the example in Table 3: the parent company controls a Hong Kong trading company (HKCFC) that:

Under these facts, it could be reasonably concluded that HKCFC performs a low value-adding activity falling within the scope of the CFC. On the other hand, the conclusion may be different if:

In the latter case, since HKCo’s services add a high value, it follows that the income does not fall within the scope of CFC income. In any case, when the CFC effectively adds no or little economic value, it should achieve little remuneration based on transfer pricing principles and, therefore, the income to be taxed in the hands of the parent entity, according to the CFC rules, should not be significant. Furthermore, ATAD Article 7(2)(a)(vi) says that passive income is an “earned” one, meaning that income should be not only invoiced but also cashed-in before being taxed in the hands of the parent company. Anyway, this aspect has not yet been clarified. Spain preferred not to introduce this type of tainted income because it was considered unnecessary in light of the general design of CFCs. Germany considers trading activities as generally active unless the business activities take place between related parties. However, income remains active if evidence is provided that the business is commercially organized with appropriate documentation. Portugal has adopted a restrictive implementation that considers trading income as tainted income even if it arises from transactions with unrelated parties.

4.10. Deductions and treatment of losses

Article 7(2)(a) ATAD does not specify how to deduct the respective direct and indirect costs from each item of passive income to determine the taxable income. Some countries such as Denmark and the Netherlands have chosen to explicitly state that deductible expenses are an integral part of passive income items and, as such, should be deducted according to local rules. Thus, costs relating to passive income should be deducted from passive income. In the case of losses deriving from one or more of the categories defined as CFC passive income (passive exS. 253penses higher than passive revenues), Article 8 ATAD does not clearly state whether those losses should offset future CFC taxable income deriving exactly from the same categories or if those losses could be used to offset taxable income whatever category to which the income belongs. In general, if the CFC generates losses, there is no attribution to the shareholder. Vertical loss compensation with other income of the taxpayer is therefore not possible. Instead, the loss is carried forward and can offset future CFC income. If a CFC generates negative passive income in one area and positive passive income in another, these shall be offset in accordance with general principles. Alternatively, losses should be stored and tracked until the same kind of passive income arises, possibly within a certain time frame.

4.11. Conclusions

After analysing the types of income envisaged by the categorical approach under Article 7(2)(a) (so called Model A), it can be said that the directive has essentially achieved its aim of providing a flexible guidance list for re-attributing the income allocated in low-tax entities (and PE) to the parent company. Model A is a “minimum standard” as individual Member States may deviate from the provision and may introduce requirements that are more rigorous and more restrictive “to safeguard a higher level of protection of national corporate tax bases”, pursuant to Article 3 ATAD. Member States adhering to Model A have implemented this methodology with substantial uniformity, albeit adopting variations in tune with their own tax competition policy.

5. The EU Primary Law and the “substance carve-out” rule

5.1. Pursuing compliance with the fundamental freedoms

The categorical approach lends itself to reflections of whether a priori taxation based on legal categories entails a restriction to the fundamental freedoms protecting the EU internal market: the freedom of establishment (Article 49 TFEU), and the free movement of capital (Article 63 TFEU). The question is whether a formulaic listing of given income items triggering taxation in the hands of a taxpayer is compliant with the aims of the EU market functioning notwithstanding S. 254the limitations developed by the ECJ on measures preventing tax abuse. Reservations on this model as a challenge to the EU fundamental freedoms have been reported since the first proposal for a directive against tax avoidance. Consequently, a disapplication clause (also named exemption clause) was added. It was engineered as a “substance carve-out”. In Article 8(2) of the first proposal for a council directive, dated 28 January 2016, the disapplication clause reads as follows:

In this respect, the CFC regime would have applied if the entity had been a wholly artificial one or an entity engaged in non-genuine arrangements that had the essential purpose of obtaining a tax advantage. Several concerns arose around this version because of its complexity. A later draft version, the so called “presidency compromise”, dated 24 May 2016, reported the further concept of the “valid commercial reasons“ which replaced the approach based on wholly artificial entities. This concept was also mentioned in Preamble number 13 which stated that: “(…) Tocomply with the fundamental freedoms, the income categories should be combined with a substance carve-out aimed to limit, within the Union, the impact of the rules to cases where the CFC has not been established for valid commercial reasons“. Additionally, Article 8(2)(a), the penultimate paragraph of the presidency compromise stated that: the categorical approach shall not apply where the taxpayer can establish that the controlled foreign company has been set up for valid commercial reasons and carries on an economic activity supported by commensurate staff, equipment, assets and premises which justify the income attributed to it“. Several delegations brought up a few concerns to delimit the scope of intra-EU/EEA CFC, which should have been limited to wholly artificial entiS. 255ties as the ECJ had extensively represented the (subjective and objective) requirements needed to enforce the CFC regime to wholly artificial entities in its case law. Moreover, the verb “commensurate” and the wording “which justify the income attributed to it” were removed since they do not appear in the ECJ case law, and it would not be easy to apply them. Finally, the motive test (“valid commercial reasons”) was removed since it was based on subjective considerations of the economic reasons underlying the incorporation of a CFC. On the contrary, objective factors were introduced both to emphasize that the test was of a “substantive” nature and to reach a political compromise between the nexus principle of the economic activity and the wholly artificial entity’s principle. In the final version, Article 7(2)(a) ATAD states that the categorical approach “shall not apply where the controlled foreign company carries on a substantive economic activity supported by staff, equipment, assets and premises, as evidenced by relevant facts and circumstances“ . The so-called “substance carve-out” acts as a “non-application clause” of the CFC categorical approach. Preamble number 12 of the directive explains it as follows “To comply with the fundamental freedoms, the income categories should be combined with a substance carve-out aimed to limit, within the Union, the impact of the rules to cases where the CFC does not carry on a substantive economic activity.” It is worth noting that the angle of the “substantive economic activity” underlying this exception is purely objective. The taxpayer must prove some objective requirements related to the existence of a real and actual economic activity supported by staff, equipment, assets, and premises and evidenced by relevant facts and circumstances. The “substance carve-out” of Article 7(2)(a) appears to leave aside the notion of “wholly artificial arrangements” already addressed by the ECJ in Cadbury Schweppes’ case law which still represents a cornerstone in the application of anti-abuse provisions. However, although the two formulations appear to differ significantly, it is considered that, for the purposes of proving the new exempS. 256tion, reference can still be made to the elaboration of both expressions of community derivation (see Court of Justice and, in particular, judgment C-196/04, Cadbury-Schweppes). It must, however, be considered that the new expression “substantive economic activity” would appear to be broader than the previous wording. It would enable the disapplication of the CFC rules even where the activity of the foreign entity does not consist in the industrial or commercial activity, but it only consists in collecting income such as dividends, royalties or interest (e.g. in the case of holding companies). In these cases, a lean structure would be consistent with its activity. The objective test at issue seems to be more restrictive than that of the artificial entity since the parent company not only has to prove the absence of artificial economic activity (negative evidence) but also has to provide positive evidence that the entity conducts a substantial economic activity to which staff, equipment, assets, and premises are devoted whereby all are substantiated by facts and circumstances. So, uncertainties arise with respect to companies with a minimal structure such as holding and finance companies but which nevertheless conduct substantial business activity. According to part of the doctrine, for an intra-EU CFC, the substantive economic test can still be interpreted in light of the case law concerning “wholly artificial arrangements”. In order to avoid uncertainties, the Netherlands has introduced a specific model in which the substance carve-out is conceived as a safe harbor provision in which, if certain requirements related to corporate governance and the labour force are met, substantive economic activity is deemed to exist. Differently, Denmark has decided not to implement the substance carve-out as it is considered that the provision of a non-application clause may undermine the objectives of the CFC regime as outlined in the directive.

S. 2575.2. Third Countries

Pursuant to Article 7(2)(a), last paragraph, if the CFC is a resident of a third country that is not a party to the EEA agreement, a Member State may refrain from applying the “substance carve-out” exemption. In other words, if a Member State does not grant the substance carve-out exception to the CFCs that are resident of an extra-EEA country, taxpayers may not disapply the CFC regime even though they can give evidence that the CFC exercises a substantive economic activity. In this regard, a parent company whose CFC is located in a third counties would not be enabled to invoke the disapplication clause even in the evidence of an active business thereof. Consequently, the income of the entity (or PE) would be taxed in the hands of the parent company under the categorical approach. In this context, the question arises as to whether the different treatment between intra EU/EEA and third countries’ CFC disapplication clause entails an unjustified restriction of the fundamental freedoms protected by EU primary law. The question presupposes a characterization issue and, precisely, whether such a restriction is framed within the free movement of capital under Article 63 TFEU or as a restriction of the freedom of establishment under Article 49 TFEU in which its scope of protection is covered within the European Union and does not extend to third countries. In this second perspective, a difference in treatment between intra-EU/EEA and extra-EEA application of CFC rules could be compatible with EU primary law. Moreover, as CFC ownership rules require the control of at least 50% of the shares or profit of the entity, thus assuming the possibility of exercising a significant influence on the CFC so as to determine its activities, a restriction of the freedom of establishment would be involved. In the first perspective, a difference in the application of the substance carve-out between intra-EU/EEA and extra-EEA CFCs also entails an unjustified restriction to the free movement of capital in light of which different treatment between intra-EU and extra-EU CFCs would be unjustified and against the fundamental freedoms of the European Union. However, these restrictions to fundamental freedoms could find justifcation in the context of the effectiveness of fiscal supervision of extra-EEA CFCs which are commonly marked by a lack of exchange of information, and the need to avoid profit stripping and tax abuse. The Netherlands extended the substance carve-out to third countries mainly to reduce the administrative burden on S. 258taxpayers and streamlining business operations but, above all, to consolidate the attractiveness as a tax residence of companies acting as group holding companies. Finland has also extended the substance carve-out to third countries but under two types of requirements: the first one is the presence of an agreement on the exchange of information; the second one is the conduct of specific production activities such as industrial and comparable production or service activities, shipping activities, sales, and marketing activities related to such activities. Finally, Finland does not apply the substance carve-out to jurisdictions considered as non-cooperative and therefore included in the EU's black list. These procedures also serve the purpose of allowing those third countries to be removed from blacklists created for the purposes of the non-application clause of the substance carve-out. With a technique opposite to that envisaged by Finland, Austrian legislation provides a list of non-substantive activities whose exercise by the CFC prevents it from benefiting of the substance carve-out. These activities include those of holding companies and conduit companies. As an alternative, the Netherlands states that, if the foreign entity carries out substantive activity that could be demonstrated, this could lead to an exclusion of the CFC status even in low-tax jurisdictions.

5.3. The Burden of proof

The effective application of the “substance carve-out” requires streamlining the burden of proof, otherwise, any possible restriction on fundamental freedoms would risk not being proportional to the aims of the directive. In the first draft, the burden of proof was placed on the tax authorities in consideration of the anti-abuse nature of the CFC rule. The presidency compromise provided a reversal S. 259of the burden of proof on the taxpayer who was required to give evidence of the valid commercial reasons that justify the set-up of the CFC carrying on economic activity supported by consistent and appropriate staff, equipment, assets, and premises. Moreover, the adoption of white-, grey-, or black-lists of third countries was introduced so as to graduate the burden of proof according to the location of the CFC within the EU/EEA or in third countries. The reversed onus on the taxpayer was assessed as legally sound. Furthermore, this approach was consistent with the recommendations of the BEPS report which appears to propose a reversed burden of proof to demonstrate a sufficient level of substance, and that the CFC under analysis be not wholly artificial. The burden of proof was ultimately revised by the council whose major aim was to promote the effective cooperation between taxpayers and financial authorities in the awareness that only transparency in the exchange of information could ultimately enable the pursuit of the objectives targeted by the categorical approach, namely, the anti-deferral, anti-abuse, and anti-tax avoidance goals. The wording of the final provision does not explicitly specify who bears the burden of proof of the existence of the “substantive economic activity”. As pointed out in Preamble number 12 “It is important that tax administrations and taxpayers cooperate to gather the relevant facts and circumstances to determine whether the carve-out rule is to apply”. It can be inferred that the council aimed at promoting cooperation between the taxpayer and the tax authorities to collect evidence about facts and circumstances to determine whether the CFC carries a substantive economic activity. Only through the effective collaboration between the parties it can it be possible to gather grounded evidence about staff, equipment, assets, and premises of the CFC. So, the final developments of the rule espose the onus lying on both the taxpayer and the tax authorities. Notwithstanding the content of Preamble number 12, Austria and the Netherlands have designed the burden of proof to be on the taxpayer.

S. 2606. Final considerations

Exhaustive treatment of all of the implications related to the implementation of the categorical approach proposed by the ATAD directive is not easily reacheable due to the multifaceted way Member States have implemented it. Nevertheless, it can be inferred that Model A combined with the option of Model B constitutes an efficient and reasoned attempt to achieve the objectives of the directive, namely, the effect of reattributing the income of a low-taxed controlled subsidiary to its parent company. However, a list of passive income based on a characterization on legal grounds still may entail some administrative burdens and possible risks of double taxation for which implications have not yet been thoroughly addressed.