This is the first part of the input regarding the MLI’s LOB. It addresses the active conduct of a business, which constitutes the first subtest under the MLI’s LOB rule. The second part, in turn, will draw attention to the second and the third subtests, i.e. the concept of income emanating from or being incidental to the taxpayer’s active conduct of business in a State of residence and a State of source or income derived from a connected person, respectively.
The MLI’s LOB rule, with its complexity and specificity, has a great potential to properly reflect the nature of tax treaties and treaty abuse. This rule could therefore constitute an appropriate response to inappropriate use of the treaties by taxpayers. The MLI’s version of the LOB rule, however, seems to be a lost chance to appropriately and specifically address the treaty abuse insofar as, despite its extremely expansive and complex wording, it has serious defects, which could undermine its effectiveness and even cause it to miss its target. This input focuses on the seemingly most important test under the MLI’s LOB rule – the active business test.
Article 7(10) of the MLI consists of an active business test whereby taxpayers, regardless of their legal form, who are not qualified persons, may still gain access to treaty benefits if they operate an active business. In contrast to the other tests, however, this test does not confer all treaty benefits, only those applying to particular items of income related to the business activity. This “item of income by item of income” approach could become a very precise tool in the prevention of abusive treaty shopping, neither too broad nor too narrow. Releasing this potential, however, will depend on the overall wording and structure of the active business test.
The test draws attention to the functional relationship between the active business conducted by taxpayers in their residence State and the treaty-benefited income received by the taxpayers from such business operations. The existence of this relationship is taken as evidence that the taxpayer is not merely serving as a conduit for income which might “normally” have been routed elsewhere . In this sense, the active business test is based on the concepts of economic substance/business purpose and aims to exclude from the scope of abusive treaty shopping practices all arrangements or transactions with sufficient degree of economic substance/business purpose. Since conduit companies are typically used only for channelling income and lack or have only minimal economic substance and/or business purpose, apart from obtaining treaty benefits, the business purpose test seems to do a good job in precluding conduit companies from treaty benefits. Nevertheless, this does not seem to hold up after a close analysis of the active business test under the MLI.
First subtest: The active conduct of a business
The focal concept under the active business test is, of course, the active conduct of a business. As a fundamental concept, it should be defined clearly and exhaustively. But it is not.
Domestic vs autonomous definition
The OECD said that the term “business” is not defined and, under the equivalents of tax treaties’ Article 3(2) of the OECD Model, must therefore be given the meaning that it has under domestic law of the Contracting State applying the active business test. This statement raises several problems because large parts of the doctrine and jurisprudence currently recommend interpreting the term “business” autonomously, using domestic law only supplementary to confirm the autonomous meaning.
The main reasons for not construing a definition of “business” under domestic law, partially or wholly, are that the concept is absent in the domestic laws of many countries and, in countries where it exists, its definition may vary significantly such that in many instances it would be simply impossible to define “business” under domestic law, or may lead to wide discrepancies in its application, making it harder to apply the business active test.
Defining a term under domestic law via the equivalent of Article 3(2) of the OECD Model under a tax treaty is allowed only if the treaty’s context does not require otherwise. One may say, therefore, that when an interpretation of a tax treaty via Article 3(2) by referring to the domestic tax law of the Contracting State applying the business active test results in double taxation or double non-taxation, while an interpretation based on context (an autonomous interpretation) avoids such results, the latter should prevail.
Accordingly, it may not be such a good idea to encourage Contracting States to always define “business” under their domestic law. Indeed, the OECD’s approach seems largely to have followed the US view on interpretation under Article 3(2), to the relative exclusion of the views of other stakeholders.
Partly autonomous (negative) definition
Partly in contradiction with its previous statement (or, perhaps, to rectify it), the OECD decided to define the term “active conduct of a business” in the Commentary, where it states that “[a]n entity generally will be considered to be engaged in the active conduct of a business only if persons through whom the entity is acting (such as officers or employees of a company) conduct substantial  managerial and operational activities.” Moreover, Article 7(10)a) of the MLI excludes from the meaning of the term “active conduct of a business” several activities and combinations thereof: (i) operating as a holding company; (ii) providing overall supervision or administration of a group of companies; (iii) providing group financing (including cash pooling); or (iv) making or managing investments, unless these activities are carried on by a bank, insurance company or registered securities dealer in the ordinary course of their business as such.
A negative condition for “active conduct of a business” under Article 7(10)a) of the MLI and the Commentary to Article 7(10) of the MLI indicates what clearly does not constitute the “active conduct of a business”. In this sense, the activity of a purely management, headquarters or holding company would not fall within the scope of the term “active conduct of a business” even though such activity constitutes “business” pursuant to the domestic tax law of the State applying the active business test. Accordingly, the term “active conduct of a business” appears in certain respects to have an autonomous meaning. To put it more succinctly, the term “business” is not the same as “active conduct of a business”. The latter is more specific and thus narrower than the former. Therefore, the only sense in which to define “active conduct of a business” in the Commentary is when it is accepted that the definition of “business” partly overlaps with the definition of “active conduct of a business”. This, in turn, means that the term “business” must be understood at least partly autonomously so that it can contain the autonomous partial definition of the term “active conduct of a business”.
The above shows some of the inconsistency in the OECD’s approach to the term “active conduct of a business”. If the OECD followed this approach consistently, there would be no partial definition of “active conduct of a business”, as the a fortiori reasoning shows. Since “business” is not to be understood autonomously but under domestic law only, it should therefore be understood relative to the narrower term contained in the term “business”, i.e. the term “active conduct of a business”.
Furthermore, the partly autonomous definition of the term “active conduct of a business” actually corresponds to the autonomous definition of the term “business” in large part, as provided by scholars following an in depth analysis of the relevant provisions of OECD Model and relevant jurisprudence, i.e. every independent income-generating activity, apart from merely passive behaviour that relies on receiving income from holding a property (e.g. shares in companies) or the management of assets (e.g. the renting-out or leasing-out of property). Consequently, the term “active conduct of a business” seems effectively to merge with the treaty autonomous meaning of the term “business”.
This leaves a very small margin for defining the term “active conduct of a business” under the domestic law of a State applying the active business test, although the Commentary explicitly encourages Contracting States to interpret “active conduct of a business” under their own laws. And this is not the last of the inconsistencies. The attribution rule blurs much of the understanding of the term “active conduct of a business” as well.
The attribution rule
Pursuant to the attribution rule, activities conducted by the taxpayer’s connected persons  are attributed to the taxpayer. A company holding shares in its wholly owned subsidiaries (i.e. typical connected persons) is therefore considered to conduct an active business if one or all of its subsidiaries carry out such business activities (e.g. manufacturing using its own officers and employees). This means that, under the attribution rules, the reality – a pure holding company – is replaced with a legal presumption – a pure operating company. In that regard, one may ask: Why are some types of activities excluded from the scope of the term “active conduct of a business” if a company entirely engaged in such activities may still be considered to be conducting an active business? Why does the activity of one particular person matter for the purpose of granting treaty benefits to another person?
Under company law, a company may be established in one State while carrying out all of its business via a foreign branch or subsidiary. Freedom of establishment is particularly protected under EU law, including the CJEU case law, apart from wholly artificial arrangements. The absence of the attribution rule would therefore restrict a taxpayer’s right of establishment too much and be a serious blow especially to financing companies within the same international group (intra-group financing) as the taxpayers would in many cases be forced to establish a finance company in their residence State (“home” jurisdiction) in order to obtain treaty benefits under the MLI’s LOB rule (the active business test). Consequently, although the attribution rule seems to haze the understanding of the term “active conduct of a business”, its existence under the active business test is pretty much justified.
Conclusion and postulate de lege ferenda
All in all, for the sake of clarity and legal certainty, it would be advisable to include in Article 7(10) or (13) of the MLI a more precise definition of the term “active conduct of a business” along with an open-ended list of activities that constitute the active conduct of a business. The treaty autonomous definition of the term “business” could be recognized, as the jurisprudence and scholars imply, and used as a yardstick for defining “active conduct of a business”.
* * * * * * * * * *
 See American Law Institute, Federal Income Tax Project, International Aspects of United States Taxation II, Proposals on United States Income Tax Treaties, 1992, p. 160.
 The term “substantial managerial and operational activities” is not defined and therefore exacerbates the vagueness of the definition of “active conduct of a business”.
 Article 7(13) e) of the MLI says that “two persons shall be ‘connected persons’ if one owns, directly or indirectly, at least 50 per cent of the beneficial interest in the other (or, in the case of a company, at least 50 per cent of the aggregate vote and value of the company’s shares) or another person owns, directly or indirectly, at least 50 per cent of the beneficial interest (or, in the case of a company, at least 50 per cent of the aggregate vote and value of the company’s shares) in each person; in any case, a person shall be connected to another if, based on all the relevant facts and circumstances, one has control of the other or both are under the control of the same person or persons.” Neither the MLI nor the Commentary to Article 29 of the OECD Model (or BEPS action 6) suggest that the attribution rule is limited to connected persons in the same country or in different countries. Hence, the attribution rule applies to connected persons irrespective of their location/tax residence. Pragmatically speaking, the connected persons will be most often located in different countries because the MLI’s LOB rule targets treaty shopping, a practice which is inherently related to persons in different countries. Presumably, the connected persons will overlap with associated entities under transfer pricing rules and/or controlled and controlling companies under CFC rules, as applied in various countries.
To make sure you do not miss out on posts from the Kluwer International Tax Blog, please subscribe to the blog here.