This is a chapter from the book: "Taxes Crossing Borders (and Tax Professors Too) - Liber Amicorum Prof Dr R.G. Prokisch".
Prof. dr Hans van den Hurk
An icon leaves Maastricht University. I still remember my first acquaintance with Rainer Prokisch. In a nice pub in Maastricht, we had our first conversation about the dynamic world of international and European tax law. And, of course, about the international LLM that Rainer was setting up. He needed a partner in crime for that. He, therefore, asked me to apply. I did not have to think twice about that and 'the rest is history'.
What makes Rainer Prokisch so special? First of all, of course, that he is a fantastic tax expert with a great knowledge of international and European law. Secondly, he has lived in several cultures (also in terms of taxation) and that makes him able to sketch valuable broad views. Incidentally, this also means that you could have discussions with him on various points where you will remain in an ‘agree to disagree’ status. Thirdly, he has an incredibly large international network that has enabled him to quickly make Maastricht University one of the best-known universities in the world. And last but not least, he is a fantastically nice man, a dear friend and a colleague whom I will miss in the years to come.
In my contribution I will discuss offshore indirect transfers (OIT), the reasons for having this subject high on the UN-agenda, the UN proposal, the alternatives developed by the Platform1 and the possible negative consequences for corporations which are mostly caused by a lack of mandatory dispute resolution and difficult double tax relief due to the fact that more than two states can be allowed to tax capital gains. Tax payer rights seem to deteriorate if these proposals are going to be introduced. Due to the limits in the number of words I will focus on the core issues.
Offshore indirect transfers is an international tax topic which gets increasingly more attention. The reason for this is that in a few well known cases during the last couple of years developing countries experienced effective tax planning by multinationals whereby these states saw a change in the ownership of domestic companies but were not able to tax the results accordingly.2 Specifically, in situations where the domestic activities deal with natural resources, the issue is that, if the source state is not allowed to tax offshore indirect transfers, at a certain moment in time the natural resources will be exhausted and the source state has missed possibilities to enjoy these revenues which are so strongly connected with that country. During the last covid years developing countries have paid new attention to these issues since they need the resources to cope amongst others with this crisis.
Offshore indirect transfers have been regularly in the global news last couple of years. Indirect transfers became world news as the consequence of, amongst others, two Indian court cases under a bilateral investment treaty. Interestingly, these cases where not just discussed by tax experts but also outside the tax world.3 India changed the law retroactively by applying rules designed to tax offshore indirect transfers in years which were more than ten years behind the year of creation of these rules. India motivated this retroactive taxation by saying that it was not so much a retroactive taxation as well as a clarification of existing rules. In one of the two cases, Cairn won the dispute before an arbitration court but India refused to follow up on this and Cairn actually seized legal action on Indian property around the globe, including embassy buildings in Paris and India’s domestic aircraft carrier Air India.4 Tax has never ever been in the news this way. In the meanwhile Cairn and the Indian government reached a solution and Cairn has dropped the court case against India.5
For this reason I will also discuss the influence of the bilateral investment treaties and specifically the so-called ‘fair and equitable’ treaty criterion from art. 4 which is part of most of these investment treaties.
Before in the previously mentioned recent Cairn case the retroactive application of Indian law against the ‘fair and equitable’ treatment condition under the bilateral investment treaty between India and the United Kingdom was tested, the world saw the Vodafone case being decided under comparable rules. A short summary of this first landmark case:
In 2006, Vodafone purchased Hutchison’s participation in a joint venture to operate a mobile phone company in India (the owner of an operating license), for nearly US$11 billion. This transfer was accomplished by Hutchison, a Hong Kong-based multinational, by selling a wholly owned Cayman Islands subsidiary holding its interest in the Indian operation to a wholly owned subsidiary of Vodafone incorporated in the Netherlands where it also resides. The transaction thus took place entirely outside India, between two non-resident companies.
The Indian Tax Authority (ITA) sought to collect over US$2 billion of tax on the capital gain realized by Hutchison on the sale of the Cayman holding company. As per the Indian Law, the purchaser is required to deduct tax at source while making payment to the non-resident seller. Accordingly, the Indian Tax Authority (ITA) held the purchaser, Vodafone’s Dutch subsidiary, liable for failure to comply with its obligation to withhold tax from the price paid by it to Hutchison on the ground that the capital gains realized by the seller were taxable in India. This sparked a protracted court case, with the Supreme Court of India ruling in 2012 in favour of the taxpayer. The Supreme Court denied the ITA’s broad reading of the law to extend its taxing jurisdiction to include indirect sales abroad, though it took the view that the transaction was in fact the acquisition of property rights located in India.
The government of India subsequently brought in a clarificatory amendment with retroactive effect to overcome the technical difficulty arising out of the Supreme Court ruling so as to allow taxation of offshore indirect sales and to validate the tax demand raised against the Vodafone's Dutch subsidiary. The legality of a retroactive effect of the law was not challenged by Vodafone in the Indian courts and instead it has submitted the action of the government of India to arbitration under the India-Netherlands Bilateral Investment Treaty.
From the above it is clear that countries challenge any OIT’s because they believe they are entitled to part of the revenues. But sometimes their fights exceed reasonable limits. In the recent Cairn case, the retroactive taxation from OIT led to an appeal to the Permanent Court of Arbitration under the bilateral investment treaty between India and the United Kingdom. But can a bilateral investment treaty have any effect on an international tax case? In the following I only address a few elements of this discussion. I refer to the Cairn case if the reader wants to know more.
In most BIT’s two paragraphs are of the utmost important. First of all, art. 4, par.3 which reads:
“The provisions of this Agreement relative to the grant of treatment not less favourable than that accorded to the investors of either Contracting Party or of any third state shall not be construed so as to oblige one Contracting Party to extend to the investors of the other the benefit of any treatment , preference or privilege resulting from:
a. any existing or future customs union or similar international agreement to which either of the Contracting Parties is or may become a party, or
b. any international agreement of arrangement relating wholly or mainly to taxation or any domestic legislation relating wholly or mainly to taxation.”
Article 4 deals with investments. With respect to these investments, many BIT’s contain a most favourite nations clause in article 4, 1 and 4, 2. But the above par. 4,3 clearly states that an exception has been foreseen with respect to e.g. a bilateral tax convention and any domestic tax rules. These exceptions are the reasons why states have often considered that an appeal on a BIT could only be of limited importance. The issue however in many tax cases, including amongst others the Cairn case, is that the country in stake applied new rules retroactively and by doing so the taxed corporation has in principle no feet to stand on. In the Cairn case the Permanent Court of Arbitration therefore applied the so-called ‘Fair and Equitable’ treatment provision of article 3. This reads:
“Investments of investors of each Contracting Party shall at all times be accorded fair and equitable treatment and shall enjoy full protection and security in the territory of the other Contracting Party.”
Without going in all the details, it is relevant to understand that India, according to the permanent court of arbitration, unlawfully expropriated Cairn’s investment without providing fair and equitable compensation. In such a situation in the case at hand ‘tax as such’ is not really the issue but the expropriation is. And therefore art. 3, 2 can be used. The Tribunal decided the following:
“1. Declares that it has jurisdiction over the Claimants’ claims and that the Claimants’ claims are admissible;
2. Declares that the Respondent has failed to uphold its obligations under the UK- India BIT and international law, and in particular, that it has failed to accord the Claimants’ investments fair and equitable treatment in violation of Article 3(2) of the Treaty;”
India lost the case and appealed in the Netherlands since the primary case was discussed before a PCA which has its seat in the Netherlands. However, before in this procedure even the hearing has taken place, India gave in and promised to pay back to Cairn part of the money which was expropriated. It was suggested that one of the reasons for this change in attitude was the belief in India that the country could benefit from a more investment friendly tax climate.
Still, from a state point of view, the frustration was immense that India had to pay back a major part of its necessary tax revenues.7 And for the future India now has to renegotiate all its tax treaties to make sure that OIT’s can be taxable by India. So for India this was quite a difficult decision after a disappointing case with the PCA.
But also Cairn is not very happy. The total tax claim was 1,7b USD and, if the press reports are correct, India promised in a compromise proposal which was accepted by Cairn to pay back 1b USD. Cairn therefore still lost 700m USD in a case in which it was decided that Cairn was totally right.
So, in situations where governments take actions as in the above mentioned India cases, a BIT can certainly play a role.
When does a state have the right to tax? Nexus is here the word that literally connects an activity or a source with a country which has as the consequence that this country has the right to tax. In the words of OECD’s BEPS Action 18, Nexus is one of the fundamental elements of the global tax system and as such determines where taxes should be paid. Under existing rules this is mainly the case where there is physical presence. The question what to allocate to that physical presence is answered by applying the Arm’s Length Principle.9 In the current tax world nexus has more than one meaning. After BEPS we have amongst others seen a ‘modified nexus approach’ which dealt with patent box regimes and the question how to determine the basis for profit allocation.
What is getting also more important is a further selection within the category of nexus situations which can be called situations in which there is ‘economic nexus’ or ‘real economic activity’. In the post-BEPS era this is getting much more importance since the traditional legal approach has been replaced with an economic one.10 Also in my country this new approach is getting much more important. Tax Authorities will only issue a ruling for activities of internationally operating companies where: ‘there is an 'economic nexus' with the Netherlands. It must concern commercial operational activities that are actually performed in the Netherlands for the account and risk of the company. These activities must match the position of the company within the group. For these activities, sufficient relevant personnel must be available in the Netherlands at group level and the number of personnel must be in proportion to the total personnel of the group. The level of the operating costs must also match the activities performed.’11 It is clear that the times are over where companies could acquire a ruling in the Netherlands on situations which were created just for tax avoidance based on a very simple interpretation of nexus.
Also in the situations of offshore indirect transfers the discussion seems at a first glance to be clear. However it is not. An ‘offshore indirect transfer’ (OIT) is in essence, the sale of an entity owning an asset located in one country by a resident of another country. Taking into account all other changes as we know them from OECD’s BEPS Actions, one could expect that offshore indirect transfers would find a place in the list of topics covered by these BEPS Actions. However disappointingly this is not the case. One reason why not, is possibly that in some situations the value creation as such is (partially) attributable to the owners and managers of such assets. If for example McDonalds would sell all the restaurants in one country via an offshore indirect transfer than the value of that transaction is not just the value calculated via e.g. a discounted cashflow since the quality of the cashflow is to a large extent based on its US brand. Under a different name these restaurants will have less value. And this is an example why some countries may choose not to accept a change in the treaties whereby the source state is allowed to tax OITs. The same goes if we look from a transfer pricing perspective to OIT’s, since then the question is mainly what the value drivers are. And in many situations part of the value drivers can be found in the central management from the company. But whether this is enough argumentation against allowing the source state to tax offshore indirect transfers? I doubt that.
This becomes most clear in situations where natural resources is the leading value driver in a country. Fine management can help increase the value of the activity, but without the resources there is no value. This is probably different when the discussion deals with communication networks since here users play a big role. But marketing and brand name are the first factors which, when dealt with adequately, lead to an increasing number of users. And since the activities relating to marketing and intellectual property can also be performed outside the state where the communication network is built, it does not go without saying that the state within which the communication network is running, has a taxing right on the sale of the shares by an entity outside that country holding the corporation which exploits the investment in that country. Actually possibly both states could be entitled to tax part of that capital gain but the determination which country may tax what, is quite difficult. Only an economic approach on value chain analysis (a transfer pricing approach) can possibly solve this but it is more likely than not that in these situations states often have a different view on how to perceive these value chain analyses and therefore double taxation may be the result. Therefore I can understand why OECD did not want to include IOT’s in the BEPS plans. In order to conquer these problems, it would be highly recommendable that in these situations profit allocation would play a role. If activities which create value are divided amongst several states, the only reason for not allocating profit to these states is that this seems to be difficult to execute. But that goes for many new rules and therefore this subject should without doubt be back on OECD’s agenda also.
Despite the above, there is one main exception where the interest for the source state should always be undoubted. Independently how good the management will be, how optimal the marketing development is, natural resources will be exhausted at a certain moment in time. And for me this is the main reason why I believe that these interests should create a taxing right for the source states, provided that it is based on rules which can be tested independently.12 Suppose a British corporation holds a Vietnam entity exploring the possibilities for lithium. At a certain moment Lithium is found and during a period of 20 years lithium has been exploited in that mine. After twenty years of exploitation the mine is no longer economically relevant anymore and therefore it is closed. Suppose also that in those twenty years the entity which holds the mine has been sold five times. Then only when the specific tax treaty contains an article like article 13,4 UN13, taxation is possible.
But suppose that this Vietnam entity is being held by an Irish entity which acts as the intermediary holding under the British headquarter and the shares in this Irish entity are being sold? In that situation this article 13, 4 UN is not applicable and the source state has no possibility to tax the five transactions in twenty years’ time in which the ‘capital gain as such’ was mainly the result from the value of the natural resources. For me this justifies special attention for the issue of natural resources in the global discussion on whether or not OIT’s should be taxable. 14 But more water has to flow through the rivers of OECD and the UN before this is possible.
Under current tax rules there are certainly positive developments to come up with legislation/treaties which enable the taxation of OIT which would work. Offshore indirect transfers (OITs) are being defined by the Platform for Collaboration on Tax:
“Offshore Indirect Transfer: an indirect transfer in which the transferor of the indirect interest is resident in a different country from that in which the asset in question is located.”15
This covers situations where for example a US company sells the shares in its Dutch intermediary which again holds a Dutch subsidiary which again owns a subsidiary in the source state.
The state which holds the assets within its border cannot tax the sale of the shares since that capital gain is allocated to the resident state from the corporation which holds that local subsidiary which owns the natural resources. And in most situations these revenues are even tax exempt based on a participation exemption or a comparable system.
The taxation of indirect transfers of assets such as mineral rights, and other assets generating location-specific rents such as licensing rights for telecommunications is a concern in many developing countries, magnified by the revenue challenges that governments around the world face as a consequence of coronavirus. But where location-specific rents like licensing rights for telecommunication will possibly only increase in value, with natural resources it is clearly different and therefore the urge for the source states is given. In the below I will first describe existing treaty rules. What is the role of article 13 OECD? Clearly, if the source state would acquire a right to tax the sale of the shares realised in the other country a taxing right must be found in article 13.
Article 13, 4 reads:
“4. Gains derived by a resident of a Contracting State from the alienation of shares or comparable interests, such as interests in a partnership or trust, may be taxed in the other Contracting State if, at any time during the 365 days preceding the alienation, these shares or comparable interests derived more than 50 per cent of their value directly or indirectly from immovable property, as defined in Article 616, situated in that other State.”
How would this work under UN rules? Under UN rules par. 4 reads exactly the same:
“4. Gains derived by a resident of a Contracting State from the alienation of shares or comparable interests, such as interests in a partnership or trust, may be taxed in the other Contracting State if, at any time during the 365 days preceding the alienation, these shares or comparable interests derived more than 50 per cent of their value directly or indirectly from immovable property, as defined in Article 6, situated in that other State.”
Both identical paragraphs allocate a taxing right to the source state for taxing the capital gain where the capital gain, as a consequence of the alienation of the shares, is based directly or indirectly from immovable property. Well, if the shares in the Dutch entity holding the Dutch intermediary which holds the Vietnam subsidiary are being sold, then the source state could, if all other conditions have been fulfilled, tax the revenues since it deals with a resident of a contracting state realising a profit which is based on the value of immovable property in the other state.
But this can be different if a UK holding holds a Luxembourg intermediary which holds a Dutch intermediary owning the Vietnamese subsidiary and the UK parent sells the shares in the Luxembourg entity. The tax treaty applicable in relation to article 13, 4 and Vietnam’s natural resources, is the Vietnam treaty with the Netherlands. The question remains to be answered whether Vietnam could possibly apply the Vietnam treaty with the United Kingdom also. From the current text this seem not to be possible although the United Kingdom is the resident state of the holding company realising a capital gain. The reason for this is mainly that the transaction in the UK, with respect to the sale of the shares in the Luxembourg entity, cannot be taxed in Vietnam since that share transaction takes place in different countries. And last but not least, the words ‘directly or indirectly’ which in a first reading possibly shines another light on this discussion, do not help Vietnam since these words do not refer to the participation but to the ‘immovable property’. And therefore Vietnam remains with empty hands.
But even if Vietnam would have the taxing right, it will be quite uncertain whether Vietnam knows that a sale takes place. That alone is another reason for new legislation. Later I will discuss the alternatives developed by the Platform which try to mitigate this issue.
The UN Model Tax Convention includes nowadays an additional paragraph in article 13, 5 which reads:
“5. Gains, other than those to which paragraph 4 applies, derived by a resident of a Contracting State from the alienation of shares of a company, or comparable interests, such as interests in a partnership or trust, which is a resident of the other Contracting State, may be taxed in that other State if the alienator, at any time during the 365 days preceding such alienation, held directly or indirectly at least ___ per cent (the percentage is to be established through bilateral negotiations) of the capital of that company or entity.”
Here ‘directly or indirectly’ seems to refer to the participation as such and this might help the source state to a certain extent. Depending on the negotiated percentage of the shareholding, the source state is able to tax the capital gain, provided that the respective treaty is one which is modelled after the UN model and the treaty contains this article 13, 5. However there is a main reason why this provision still does not work. The first line has clearly been drafted that, in order to apply this paragraph, it has to deal with the alienation of shares in an entity in the other state. Let us look at an example to see the limitations of article 13, 5. What would the above article 13, 5 UN mean for the case of the UK parent with a Luxembourg intermediary which holds again a Dutch entity which holds the Vietnamese subsidiary and the UK would sell the shares in the Luxembourg company?
The first part of article 13, 5 (Gains, other than those to which paragraph 4 applies, derived by a resident of a Contracting State from the alienation of shares) refers in this example to the United Kingdom entity which sells its Luxembourg intermediary holding.
The second part of article 13, 5 (shares of a company,…, which is a resident of the other Contracting State, may be taxed in that other State) defines the corporation on whose shares a capital gain has been realised. Can Vietnam use this provision to tax the capital gain in this situation? Unfortunately Vietnam cannot. If I fill in the respective corporations, article 13, 5 reads as follows:
“Gains derived by the UK holding from the alienation of shares of a Luxembourg corporation, may be taxed in Luxembourg if the UK holding, at any time during the 365 days preceding such alienation, held indirectly at least 10017 per cent of the capital of that Luxembourg entity.”
This would be the literal reading. The problems for Vietnam is simply that art. 13, 5 is not applicable since no shares in the Vietnam entity have been alienated. And although the words ‘direct and indirect’ improve the possibilities for applying article 13, 5, it does not help India here. In the words of the Platform:
“This allocates to country (source) taxing rights over the gain derived by a non-resident of country (source) from the disposal of shares (or comparable interests) of a company, partnership or trust that is itself resident of country (source). However, this only applies to offshore direct ownership of such entities. For that reason, while Article 13(5) may help address certain tax avoidance arrangements (e.g. certain dividend-stripping or change of residence strategies), it is not suitable as a provision to ensure the source taxation of gains on indirect transfers. According to the UN Manual for the Negotiation of Bilateral Tax Treaties between Developed and Developing Countries, treaty practice varies with regard to the use of Article 13(5): some countries explicitly exclude gains on listed shares; others restrict the scope to gains realized by individuals who were previously residents of the source State; and many countries do not include this provision at all in their treaties.”18
Here we see the lacking possibilities from bilateral tax treaties in a reverse way. In January 2022 I published an article on triple taxation and the impossibilities for companies to get triple taxation being relieved, here we see countries suffering from the same.19 If there would be room for third state intervention, quod non, 13,5 could, in my last example, be read as follows:
“Gains derived by the UK holding from the alienation of shares of a Luxembourg corporation, may be taxed in Vietnam if the UK holding, at any time during the 365 days preceding such alienation, held indirectly at least 10020 per cent of the capital of that Vietnam entity.”
This demands that tax treaties would no longer have a bilateral character. But unfortunately, this is possibly one bridge too far, yet. The latter approach would only be possible if there would be a multilateral instrument which creates taxing rights for the source state on capital gains related to the source state resources which have been realised between a second and a third state.
The previously discussed disputes between India and two British multinationals can be looked at from several angles but for me it is evident that, since real value goes from one hand to the other without the source state (even more when it is a developing state) being able to tax this, this is, to say the least, quite strange. The source states should have a legitimate right to tax the sale of natural resources but changing the system is unfortunately not easy. And it is not just India suffering from this. In the meanwhile, several case around the globe are pending and a call for a change in the allocation of taxing rights is heard around the globe.21
One of the reasons why developing states are desperate in their wish to tax OIT is that, by using indirect transfers in order to alienate shares, a legitimate but unreasonable way of tax avoidance is used which is difficult to challenge by these states. Again, suppose OIT should remain untaxed as it actually is at the moment, then source states with natural resources will see all kind of changes in the ownership of these resources without being able to tax them and possibly in the end the natural resources are exhausted.
These are the reasons why several organisations in the world work on proposals for taxing OIT’s, which initiatives I fully support.22 In the below I will therefore focus on part of the UN-plans to tax OIT’s but also on the plans of the Platform in which UN plays also a relevant role.
The United Nations developed a reasoning whereby OIT should also fall under Article 13(4). This is clarified in the report of the virtual meeting from October 20-29, 2020 which states the following:
“At its nineteenth session (Geneva, 15-18 October 2019), the Committee discussed note E/C.18/2019/CRP.22 which dealt primarily with the issue of so-called offshore indirect transfers (OITs) and how the gains from such transfers should be dealt with in the UN Model.
After discussion, the Committee welcomed the Subcommittee’s decision to work on a treaty provision that would allow source taxation of capital gains on OITs not already covered by Art. 13(4) without prejudging the question of whether that provision would be included in Article 13 of the UN Model or presented as an optional provision in the Commentary.”23
It has to be seen whether a solution in the commentary as suggested in the above would actually work. Specifically when the other state is a OECD state, it could be difficult to actually negotiate that taxing right. And a change in the text is neither going to bring a solution since the source state wants to tax a capital gain which a company in a third state will realise on the shares of the second state. Suppose that a Dutch holding company sells its shares in a UK intermediary holding which holds a Vietnam operating company, then Vietnam would like to use article 13, 4 in order to tax the Dutch entity. But the Netherlands (or any other country) could simply reject the claim of Vietnam since its treaty with Vietnam is not applicable on this transfer.24
Are there any other ways taxing an OIT without changing the text of a treaty? Possibly the Principal Purpose Test could be applied in situations where an entity is inserted in a structure in order to enable an offshore indirect transfer to be created instead of a direct transfer.25 Article 29, 9 reads as follows:
“Notwithstanding the other provisions of this Convention, a benefit under this Convention shall not be granted in respect of an item of income or capital if it is reasonable to conclude, having regard to all relevant facts and circumstances, that obtaining that benefit was one of the principal purposes of any arrangement or transaction that resulted directly or indirectly in that benefit, unless it is established that granting that benefit in these circumstances would be in accordance with the object and purpose of the relevant provisions of this Convention.”26
Based on this provision the treaty benefits can be ignored by Vietnam in relation to the (in my latest example) UK. But the UK entity is being sold and why would Vietnam be able to tax the capital gain realised by an entity of a third state if no direct taxable transaction between Vietnam and the UK takes place. And even if Vietnam would act as if the treaty between Vietnam and the UK should not be applicable and therefore apply domestic legislation, the tax money should come from the Netherlands. And although the latter is not always a problem (the claim will be put on the source state entity like India did in the Vodafone case) it is not my preferred alternative.27 28
In the meanwhile, and this is very important, the International Chambers of Commerce does seem to accept taxation of OIT though with one sidenote. International Chambers of Commerce has responded to the UN plans and agreed to a certain extent with one, to my opinion reasonable exception. ICC strongly recommends to exclude intra-group reorganisations since no true alienation took place. See:
“Similarly, exclusions for intra-group reorganisations should be included – there should be no need to tax where the asset has not been truly alienated.”29
And since to my opinion there should be arrangements available which allow the source state to tax OIT provided that mandatory dispute resolution is provided, their position stems hopeful.
Despite this last positive note, I have to conclude that the UN specific developments in order to improve the possibilities to tax offshore indirect transfers do not really make sense. This is probably the reason that the Platform for Collaboration on Tax, in which UN also participates is considering alternative approaches.
A lot of work in the field of OIT has been performed on instigation of the Development Working Group (DWG) of the G20 to the International Monetary Fund (IMF), Organization for Economic Cooperation and Development (OECD), World Bank Group (WBG) and the United Nations (UN)—the partner members of the Platform for Collaboration on Tax—to produce “toolkits” for developing countries for appropriate implementation of responses to international tax issues under the G20/OECD Base Erosion and Profit Shifting (BEPS) project, as well as additional issues of particular relevance to developing countries that the BEPS project does not address. The platform aims also at helping developing countries to find a way to effectively tax offshore indirect transfers.30
The report outlines two main approaches to the taxation of OITs by the country in which the underlying asset is located. One of these methods (‘Model 1’) treats an OIT as a deemed disposal of the underlying asset. The other (‘Model 2’) intends to treat the transfer as being made by the actual seller, offshore, but sources the gain on that transfer within the location country and so enables that country to tax it. The Platform wants to share both alternatives without expressing a general preference between the two models. They leave that to the countries’ circumstances and preferences.
Model 1 has been defined by the Platform as follows31:
Model 1 (taxation of a deemed direct sale by a resident): This model seeks to tax the local entity that directly owns the asset in question, by treating that entity as disposing of, and reacquiring, its assets for their market value where a change of control occurs (e.g. because of an offshore sale of shares or comparable interests). This model takes into account the fact that a capital gain has been realized through the indirect transfer, triggered by a change of control. The relevant taxpayer under this model is not the seller entity, which effectively disposes of the shares, but rather the entity which actually owns the assets from which the relevant shares derive their value. Model 1 needs to be supported by a deemed disposal and reacquisition rule of the assets from which the shares actually disposed of derive their value.32
The thought behind model 1 is a simple one. The Platform just deems that everything happened in the source state.33 But if an OIT takes place, the entity in the source state which owns the assets is liable for the tax burden and has to pay the taxes due. Possibly that entity does not own the resources to pay this tax claim since this entity does not possess the revenues from the OIT. The entity which sold the shares is not going to support the source state entity since it no longer owns that entity. So in an OIT process the buyer will have to be aware that an additional tax burden can be triggered by the transaction.
Still it is clear that, since the ‘source state entity’ will be taxed, taxation will take place by a country which has the right to tax an entity both a resident and a source state basis. This specific model seeks to tax the accrued gain on the entire asset when a change of control occurs from the sale of interests whose value is principally (e.g. more than 50 percent) derived from the asset. The Platform is quite consistent in some parts of the explanation by confirming that losses should also be recognized where there are no accrued gains, and should be subject to appropriate loss utilisation rules (as applicable). What I find quite difficult to understand in the Platform’s reasoning is the link to the European Union and the Anti-tax avoidance directive, which link is to say the least deviate. In the words of the Platform:
“The taxation of unrealized capital gains derived by tax residents is not uncommon. The most recent example of taxing the unrealized capital gains of corporate taxpayers is the EU’s exit taxation rule, which is being implemented by all 28 EU Member States”
It is true that the European Union did introduce capital gain legislation but only in the situations below:
“(a) a taxpayer transfers assets from its head office to its permanent establishment in another Member State or in a third country in so far as the Member State of the head office no longer has the right to tax the transferred assets due to the transfer;
(b) a taxpayer transfers assets from its permanent establishment in a Member State to its head office or another permanent establishment in another Member State or in a third country in so far as the Member State of the permanent establishment no longer has the right to tax the transferred assets due to the transfer;
(c) a taxpayer transfers its tax residence to another Member State or to a third country, except for those assets which remain effectively connected with a permanent establishment in the first Member State;
(d) a taxpayer transfers the business carried on by its permanent establishment from a Member State to another Member State or to a third country in so far as the Member State of the permanent establishment no longer has the right to tax the transferred assets due to the transfer.”
In all of these situations the gain has been realised and been taxed in the same country. This is principally different in the Platform’s Model 1. And this seems to me also the main shortcoming for this system to be accepted. Model 1, as described by the Platform does not really address possible difficulties and how these should be challenged. Yes, I do read an exception for internal reorganisations and the Platform clearly has a view on tax avoidance schemes like staggered sell-downs since there is no minimum threshold. But the only element where the Platform does see some complications is the valuation element. Where a change of control occurs, the model treats the local asset owning entity as the entity disposing its assets for their market value. The market value of the local assets which are deemed to be sold, could be determined administratively using assumptions and adjustments based on the price at which the actual shares are sold since this price is, to a high extent, based on the value from the local assets. It is convincing argument that this approach is quite complex and will possibly lead to disputes about the real value which is deemed to be realised. That leaves alone that the source state should know about this transaction. This last problem can be solved according to the Platform by imposing a reporting obligation of the price of the shares to the resident entity.34
Then there is also the discussion on possible double taxation. Since the alienation is being taxed in the country where the assets are located, double taxation could arise if at a later moment this entity is being disposed again. For this situation the model treats the local asset owning entity as reacquiring the assets for their market value. This means that its tax cost in those assets is stepped up to market value. In the words of the Platform:
“This means that its tax cost in those assets is stepped up to market value— which is important to ensure that double taxation does not arise in the location country in the event that another subsequent change of control occurs. The effect of the step up is thus to neutralize the double taxation in the location country. It also effectively leads to the outcome that, over time, potential international economic double taxation is mitigated—not by immediate relief for the same taxpayer, but by mitigating the economic effects of this potential international double taxation on the side of the entity owning the relevant assets, when the stepped-up values of assets are taken into account for the determination of their tax base through depreciation, amortization or other deductions related to the values of those relevant assets.”
This does mean that whenever a holding company performs an OIT, there will be cash-out at the source entity which will only be compensated in time if that entity is producing enough profit to set these additional depreciation rights off.
But what in an imaginary situation where in a couple of years several OIT’s take place? If in three years’ time three times an OIT takes place, then it is obvious that the compensation for tax cost (depreciation of the tax cash-out) will not be experienced as a reasonable compensation. The possibility to depreciate these tax costs does not really solve the problem since it remains cash out and, depending on the depreciation period, there remains a cash-flow issue. Therefore some finetuning would certainly be desirable. Still the Platform seems to be convinced that the application of Model 1 mitigates this potential economical taxation fully. The Platform is of the opinion that by mitigating the economic effects of potential international double taxation on the side of the entity owning the relevant assets, when the stepped-up values of assets are taken into account for the determination of their tax base through depreciation, amortization or other deductions related to the values of those relevant assets, this is taken care of.
Another possible source for double taxation could be found in situations where the country in which the sale takes place, has no territorial system or participation exemption. The Platform suggests that more and more countries have one of these systems. And therefore, any taxation in the source state would not lead to adverse consequences from a sellers perspective. The Platform also adds to this that if countries would not have such a system, then still the Mutual Agreement Procedure could possibly be applied. But since the MAP procedure does not lead to a mandatory decision eliminating double taxation, that alternative does not make me very enthusiastic.
Model 2 has been defined by the platform as follows:
“Model 2 (taxation of the non-resident seller): This model seeks to tax the non-resident seller of the relevant shares or comparable interests via a non-resident assessing rule. Model 2 must be supported directly or implicitly by a source of income rule, which provides that a gain is sourced in the location country when the value of the interest disposed of is derived, directly or indirectly, principally from immovable property located in that country. A source rule relating to gains from the disposal of other assets may also be considered, as discussed above in Section II—including substantial shareholdings in resident companies. A source of income rule may be further supported by a taxable asset rule dealing with such matters as whether taxation only applies to disposals of substantial interests (such as a 10 percent shareholding rule) to exclude from the scope of tax changes in ownership of portfolio investments. The rule can also prescribe whether the entire gain will be subject to tax when the value of the indirect interest is less than wholly derived from local immovable property or, alternatively, whether the gain will be subject to tax on a pro rata basis. Each legal design option under this rule is discussed and explained further below.”
Model 2 seems to be a little bit more complex. The source of income rule does not do more than defining the sort of income and determining whether that income fulfils the conditions to make it taxable. In other words: derived from sources is a gain arising from the alienation of immovable property in the source state if, at any time during the 365 days preceding the alienation, holds more than 50% of the value of the assets. The additional taxable assets rule defines what to do when the alienator of the participation holds less than 100%. Apparently Kenia has legislation whereby when the participation is at least 50%, the whole capital gain is taxable and if it is less than 50% a reciprocal part of it.35
Here we see the taxation not taking place in the source state as such but in the state of the country that alienates the shares. One of the main problems is how to actually get this done. The Platform suggests a little logically to obey existing treaty obligations. The Platform suggests:
“The development and application of any domestic legislative provisions will need to take into account any existing tax treaty obligations. However, as discussed above, the taxing right over gains realized on offshore indirect transfers which are principally (e.g. more than 50 percent) derived from local immovable property is generally preserved in Article 13(4) of the OECD and UN MTCs. Both MTCs permit the location country to capture gains from the sale of relevant interposed holdings at different tier levels. It is important that the domestic legislative provisions of the location country be designed and drafted to preserve this taxing right over relevant interests which derive more than 50 percent of their value, directly or indirectly, from immovable property in the location country as permitted by the MTCs.”36
One of the ways to enable the source state to tax the capital gain is a the introduction of a withholding tax system. Several countries already use a withholding mechanism to collect tax with respect to a non-resident seller’s gain. The Platform sees this as a practical alternative since this alternative is not new. Of course there is then a need for a new specific withholding tax regime which should be designed and drafted to apply on payments to a non-resident seller. Withholding taxes can represent all or a portion of the tax liability (or possibly an estimate) of the recipient of the payment. The tax must be withheld from the payment by the payer and paid to the tax authority in the location country. It is a stronger alternative if embedded in tax treaties than just taxing the source state entity as done in the first alternative. But for both goes that there should be a remedy in place for situations where discussions can/will arise like for example with respect to the determination of the value.
The Platform also suggests to design a regime to impose a withholding tax by the payer, as either a final or non-final charge on the payee. A final withholding tax represents in that alternative the final tax liability for the person receiving the payment on which part is withheld and paid to the source state. A non-final withholding tax is collected as an estimate of the recipient’s final income tax liability. In that alternative the recipient is ordinarily still required to file a return and pay any outstanding balancing amount after claiming a credit for the amount of tax withheld (or receive a refund, if the withheld amount exceeds the tax due).
The Platform suggests that if a withholding tax regime would be applicable to OITs then it should be designed as a non-final regime. The issue I see in this alternative that if the states are in agreement on the value, the withholding tax can be paid in one payment, and if not, the 2nd part of the payment will certainly lead to disputes.
Still there is a lot of experience with some kind of withholding taxes. A withholding tax regime applies according to the Platform to OITs in a number of jurisdictions, including the U.S., Canada, India, China and Australia. If I read the Platform’s report right, in several situations taxation at source takes place by simply accepting treaty override. Or by applying unclear anti-avoidance rules. And although there is a legitimate interest for the source state to tax, this solution would simply deteriorate company’s tax payers rights and therefore we should consider alternatives.
First of all, there is the possibility that these two Models are not necessarily mutually exclusive which leads to these Models being applied in the same time. Therefore, as suggested by the Platform, if both were to be adopted an ordering rule would need to be clearly established to ensure they do not both apply at the same time. Where it is obvious that both Models have several issues which are not yet clear and which have adverse consequences for companies, it is difficult to understand why no ordering rule has yet been designed. If the world would proceed in one or both alternatives let then tax treaties at least have a rule to deal with this unexpected complexity.
The good thing is however that at least the Platform came up with two alternatives which have possibilities to be effective. Both models are based on the idea that source states should have more taxing rights with respect to OIT than they currently have. I support that thought. However, as suggested by the Platform, there are many issues which still should be considered before these Models can be introduced. Complexity has stopped the Platform to further dive into this yet. And that might be an indication that these Models are, in the current status of development, not the most appropriate solution.
But what I also find missing, and I am repeating myself, is the balance between taxing rights for countries versus tax payer rights. Often this is due to the fact that not in all states of this world the tax courts have the independent view e.g. the Dutch or German tax courts normally have. And not in all states of the world the legislator acts in such a way that no retroactive taxation will take place like in the Indian case. And where in the latter case an appeal on a bilateral investment treaty helped the company to overcome the issue, if comparable problems arise in the context of one of the alternatives as developed by the Platform, then neither tax treaties nor bilateral investment treaties will probably be able to take care of the job totally.
One of my other worries is the application of domestic anti-abuse rules. From the report of the Platform I learned that apparently China applies such a system. As described by the Platform:
“if an indirect transaction of assets located in China fails the comprehensive test of its reasonable business purpose, i.e., if meets all of the following conditions: (1) the overseas holding company is situated in a jurisdiction where the effective tax burden is significantly low, or where offshore income is not taxed; (2) the value of the asset directly transferred derives at least 75 percent (directly or indirectly) from Chinese taxable property; (3) 90 percent or more of the total assets or income (directly or indirectly) of the overseas holding company is based on investment or income from China; (4) the overseas enterprise does not undertake substantive functions and risks; (5) the tax consequence of the indirect transfer in the foreign country is less than the Chinese tax payable if the sale had been made directly, then the Chinese tax authority may determine that there is no reasonable commercial purpose to the offshore transaction other than avoiding the Chinese tax, and re-characterize the indirect transfer as a direct one.”
Specifically the 4th criterion refers to something comparable with the Principal Purpose Test. Whatever substantive functions and risks are, is even from a transfer pricing perspective black or white. So, if a Belgium company would dispose its Swedish intermediary which holds a China entity, then condition 1 has been fulfilled (Belgium has a participation exemption), let us assume that condition 2 has also been fulfilled, then possibly condition 3 and 4 would create a discussion with the tax authorities if we also accept that the 5th criterion has been met. It is imaginable that this discussion won’t be easily won by the company and that therefore China taxes the revenues from the indirect transaction as if it was a direct one. But this situation should only applicable under a tax treaty if a person of one state sells real estate in the other (read: China). And possibly the only way how China will be able to effectuate its tax claim, is applying powerplay by taxing the Chinese entity. It might be a solution but clearly not a preferred one. We need rules which balance the source state taxing rights with the tax payer rights since the company won’t have no real means to challenge the position of the Chinese tax administration.
When an OIT takes place, not being an internal reorganisation, the buyer has to take into account that the ‘source state entity’ has a tax obligation. This tax cost is based on the capital gain realised on the indirect sale from the source state entity. Determining what that capital gain is, is not easy. But due diligence processes will be able to help to determine this. Still, it is not black or white and therefore issues will certainly arise.
It would be a first step that the buyer pays part of the price (capital gain (in a 100% sale of the shares) multiplied with the specific capital gain tax rate) to the source state. Model 2 calls this situation the withholding tax variant. But the characteristic of most tax treaties, is that they have a bilateral character.37 And in a situation where a UK corporations sells the shares in a Dutch entity which holds a US entity which again holds an Indonesian entity, the Dutch entity should have to pay the withholding tax but this seems quite difficult to effectuate. Still some states tax part of OIT by allowing treaty override and tax where it is possible to do so. An example is the US Foreign Investment Real Property Act.38 Still I would recommend that the world develops a clear international approach which is effective for states to tax their legitimate part of a capital gain and for companies being able to rely on rules and treaties. Tax treaty override is to me not a reasonable alternative.
Another alternative would be to adapt article 13, 5. However, as mentioned previously, this is only possible if it would have a legal basis where more than two states will be able to allocate taxing rights. If this would exist, art.13, 5 UN could be rewritten in:
“5. Gains, other than those to which paragraph 4 applies, derived by a resident of a Contracting State from the alienation of shares of a company, or comparable interests, such as interests in a partnership or trust, which is a resident of a second Contracting State, may be taxed in the third contracting State if the alienator, at any time during the 365 days preceding such alienation, held directly or indirectly at least ___ per cent (the percentage is to be established through bilateral negotiations) of the capital of the company or entity in the third contracting State.”
Possibly the best way to get the job done, is to enable this system via a new Multilateral Instrument. I would prefer this, since these kind of instruments enable the countries also to include ways to exchange necessary information and to prevent double, triple or even worse taxation. And the world has acquired some relevant experience with these kind of instruments.
Resuming, for both suggestions it goes without saying that a bilateral approach is not going to do the job. It is time that the world starts considering setting a next step in international taxation. In fact, considering a new approach where more states than two will be involved in a taxable transaction.39 That would be a much better approach than installing national measures which override tax treaties.40 It supports states in their wish to tax OIT’s and it helps companies in their drive to be in control of their tax liabilities.
Offshore indirect transfers is a complex subject. The current tax treaties make taxation of capital gains realised on shares which envisage the value of natural resources hardly possible. To my opinion, these source states have a sufficient qualifying connection between the state and the natural resources to be allowed to tax these, provided that the capital gain which is realised outside the state of the natural resources is directly linked to those natural resources. In other words, there is sufficient territorial nexus to tax this. Current international tax rules make this taxation in most situations impossible while at the other end of the transaction (seller state) most states will exempt the capital gain on the sale of the shares. Unilateral approaches do the job for the states but lead to a disbalance between taxing rights and tax payer rights.
Still it is clear that the source states should acquire a clear taxing right in observance of tax payer rights. But I see too many issues with the existing and the proposed systems. Therefore new treaty perspectives have to be developed. And these new initiatives should be based on a multilateral tax treaty which help states not just with providing a taxing right but also, that in situations where the value is not just based on the assets in the source state, that a 2nd or 3rd state acquires a taxing right. Transfer pricing principles could help here by using a value chain analysis and accordingly a profit allocation based on this. And, if we would get a better dispute resolution system than the current BEPS Action 14 that there is a balance between taxing rights and tax payer rights. And it goes without saying that this new multilateral instrument should focus also on other issues which demand a multilateral solution.41 I do have to emphasize that to my opinion BEPS Action 15’s MLI is only to a certain level a multilateral treaty. I envisage a multilateral treaty to be developed which is also supportive to existing treaties but which directly empower taxing rights when countries have ratified this. After the BEPS plan many new open norms have been created which the world clearly needed in order to create a more reasonable tax system and which helped the world to challenge unacceptable tax avoidance.42 But where the BEPS plan have been mainly focusing on taxing rights, this new plan should focus on a balance. I suppose I could call this new instrument BEPS Action 16 but better is assumingly Pilar 3. This is not to come up with analogies but mainly to emphasize that we should make progress here. In the interest of countries and companies.
Rainer, with respect to my contribution, I know that you care very much about making sure that taxation is based on principles like the old ‘no taxation without representation’ but also that you strive to create a balance between source state taxing rights and taxpayer rights. In this contribution I gave you my views on this relevant subject which is specifically quite important for developing countries. Countries like China do not need our help in order to enable taxation. But many developing countries are not that strong. And by approaching this in a multilateral way we help developing countries in more easily applying their taxing rights and also making sure that China (and several other states) apply them in a fair way from the perspective of companies.
The report of the Platform was an attempt to ignite new discussions on this topic and I still believe that it helps, but only a little bit. For me this means that this topic will be one of my mean research topics in the years to come. And I am fully convinced that you and I will have very interesting discussions on this topic also after you left university. International tax law is complex but also fun. So nobody in this field will ever really retire. As I wrote in the introduction, you will be missed as my direct university colleague, and as a friend. But not just by me. We will stay in touch. In the words of the title of a famous Carole King song: You’ve got a friend! Het ga je goed!