“A statesman in these days has a difficult task. He has to pursue the policy he deems advantageous to his country, but he has at the same time to recognize the force of popular feeling. Popular feeling is very often sentimental, muddleheaded, and eminently unsound, but it cannot be disregarded for all that.” Hercule Poirot – Agatha Christie

Two years after the publication of the report titled Addressing Base Erosion and Profit Shifting, OECD countries together with the G20 and partner jurisdictions launched the final OECD/G20 Base Erosion and Profit Shifting Package in October 2015. In totality 15 Actions points were recommended as a part of the Action Plan. The aim of the BEPS plan was to address tax challenges faced by various taxing jurisdictions both at the national and International level, with particular reference to addressing taxation concerns where economic activities were undertaken and value was created, while at the same time giving business greater certainty with regards to tax policies, and effective and efficient methods of dispute resolution. Emphatically, most of the recommendations in the project were incorporated in the 2017 OECD Model Tax Convention and commentaries, specifically aimed at addressing:

• Neutralizing the effects of hybrid mismatches.

• Prevention of Artificial PE avoidance.

• Granting of inappropriate treaty benefits or state aids.

• Improving direct tax dispute resolutions between jurisdictions.

Some of which have been already highlighted in an alternate context by your author in the August 2020 issue.

Undoubtedly the Multilateral Convention to Implement tax Treaty Measures being a historical turning point in international taxation, that resulted in the introduction of a third paradigm in the arena, in addition to the existing private international tax law and bilateral treaties, revolutionises the international tax treaty network by way of a multilateral as opposed to a bilateral modification of Covered Tax Agreements (Under the MLI Convention) by a single stroke.

According to the Executive summary on Action point 15 of the BEPS report “Tax treaties are based on a set of common principles designed to eliminate double taxation that may occur in the case of cross-border trade and investments. The current network of bilateral tax treaties dates back to the 1920’s and the first soft law Model Tax Convention developed by the League of Nations…….. Globalisation has exacerbated the impact of gaps and frictions among different countries’ tax systems. As a result, some features of the current bilateral tax treaty system facilitate base erosion and profit shifting and need to be addressed. Beyond the challenges faced by the current tax treaty system on substance, the sheer number of bilateral treaties makes updating the current tax treaty network highly burdensome.”

The front note Multilateral Instrument (MLI) in accordance with action point 15 Base Erosion and Profit Shifting Report reads the purpose as to “Analyse the tax and public international law issues related to the development of a multilateral instrument to enable jurisdictions that wish to do so to implement measures developed in the course of the work on BEPS and amend bilateral tax treaties.” (https://www.oecd.org/ctp/BEPSActionPlan.pdf)

It goes without saying that bilateral tax treaties are a result of multiple years of hard labour and diplomatic negotiation and any amendments to any of them would also result in similar sort of labour. A mechanism thus which could lead to modification of multiple of them by a single stroke would be highly preferable, particularly which would entail lesser negotiation and would drive home the relevant point.

A brief introduction into the structure of the Multi Lateral Instrument would be in order at this juncture.

The MLI text introduces “Recognising that governments lose substantial corporate tax revenue because of aggressive international tax planning that has the effect of artificially shifting profits to locations where they are subject to non-taxation or reduced taxation………. “(https://www.oecd.org/tax/treaties/multilateral-convention-to-implement-tax-treaty-related-measures-to-prevent-BEPS.pdf)

The MLI document consists of 39 Articles with the following headings

Part I : Scope and Interpretation of Terms (Articles 1-2) : Sets out the scope of the Instrument and definitions thereunder

Part II : Hybrid Mismatches (Optional) (Articles 3-5) : BEPS Measures

Part III : Treaty Abuse (Articles 6-11) : BEPS Measures

Part IV : Permanent Establishment (Optional) (Articles 12-15) : BEPS Measures

Part V : Dispute Resolution (Articles 16-17) : BEPS Measures

Part VI : Arbitration (Optional) (Articles 18-26) : Provisions related to Mandatory Binding Arbitration

Part VI : Final Provisions (Articles 27-39) : Procedural provisions related to acceptance, reservations amendments, entry and exit provisions etc……

For a detailed overview, the reader might go online at https://www.oecd.org/tax/treaties/multilateral-convention-to-implement-tax-treaty-related-measures-to-prevent-beps.htm

It might be noted that as of December 2020 95 countries have signed the MLI agreement. For the Indian timeline, MLI document was published in November 2016, by June 2017 India was amongst the 68 jurisdictions that signed the MLI, June 2019 India Ratified and deposited the instrument of the MLI, MLI entered into force in India on the 1st October 2019, and on 1st April 2020 Indian MLI provisions became effective with 23 countries. While a detailed discussion on the Indian scenario would be out of scope, it can be said that India has been an active participant towards the formulation and continuity of the MLI.

Without going into the detailed academics of what the MLI Instrument actually is and what it strives to achieve, and by what mechanisms, this article tries to delve on what the direct and indirect issues of the MLI have been at large to the tax administration and the comity of nations thereof.

Its pertinent to note that the MLI applies only where all concerned parties have designated CTA’s, and application is limited only those provisions that all the parties to the relevant CTA’s have accepted. Consequently, the impact is limited by not only the jurisdictions ratifying the MLI but also the choices within the MLI which the contracting jurisdictions have chosen to adopt.

1. Covered Tax Agreements:95 jurisdictions signed the MLI as of going into press (https://www.oecd.org/tax/treaties/mli-matching-database.htm), with over 2800 agreements notified. Of the notified agreements 59 jurisdictions have ratified and deposited their instruments, 36 although have signed but are still to notify their CTA’s. For 10 jurisdictions the MLI comes to force in the year 2021, mostly in the about the first quarter of 2021 as per rules, a notable name is Germany which although signed the instrument in June 2019, ratified only on 18/12/2020. Some notable exceptions from signing the MLI are Brazil and the United States. Among the tax havens the Isle of Man is a notable signatory. The United states has cited its tax agreements already being consistent with most of the proposed solutions under the MLI (https://www.orbitax.com/news/archive.php/Treasury-Official-on-Why-U.S.–25360 ) for not being a party to the MLI. Further the United States deems the provisions of the PPT/LOB(Principal Purpose Test/Limitation on Benefits), although desirable but not effective.

2. Minimum Standard and Treaty Abuse : All jurisdictions who are signatories are required to include measures to prevent treaty abuse as a minimum standard under article 6. The text of the article 6(1) is self explanatory “ A Covered Tax Agreement shall be modified to include the following preamble text:

   Intending to eliminate double taxation with respect to the taxes covered by this agreement without creating opportunities for non-taxation or reduced taxation through tax evasion or avoidance (including through treaty-shopping arrangements aimed at obtaining reliefs provided in this agreement for the indirect benefit of residents of third jurisdictions)”.

   Although there is no indicator from the OECD MLI database as to the specific implementation of this article in all about 61 jurisdictions have agreed to implement this as a minimum standard. India’s position on this article is interesting post the landmark Azadi Bachaao Andolan case in which the Supreme Court categorically held “ 125. There are many principles in fiscal economy which, though at first blush might appear to be evil, are tolerated in a developing economy, in the interest of long term development. Deficit financing, for example, is one; treaty shopping, in our view, is another. Despite the sound and fury of the respondents over the so called ‘abuse’ of ‘treaty shopping’, perhaps, it may have been intended at the time when Indo-Mauritius DTAC was entered into. Whether it should continue, and, if so, for how long, is a matter which is best left to the discretion of the executive as it is dependent upon several economic and political considerations. This Court cannot judge the legality of treaty shopping merely because one section of thought considers it improper. A holistic view has to be taken to adjudge what is perhaps regarded in contemporary thinking as a necessary evil in a developing economy.” Consequently India has not notified article 6(3) of the MLI.

   In addition to this a number of jurisdictions have also implemented article 6(3) of the MLI which reads “Desiring to further develop their economic relationship and to enhance their co-operation in tax matters”.

   This is possibly aimed at jurisdictions who are not OECD members nor have signed the Convention on Mutual Administrative Assistance in Tax Matters, however it might be noted that India has not notified the additional provision of article 6(3) over and above the minimum standard. Under article 7 most jurisdictions have opted for the simpler PPT Test as opposed to the more mechanical LOB model. However, it might be noted that in most US based treaties there are detailed LOB provisions and also they are deemed to be contradictory to the EU fundamental freedoms.

3. Hybrid Mismatches :With the notable exception of Netherlands and Australia, who have chosen unrestricted rights as to determination of hybrid mismatches, 66 jurisdictions have notified clause 3(1) of the MLI instrument, which disallows hybrid mismatches. It is interesting to note India’s position of approval thereon where private international law has held that a partnership is a taxable unit under the Income Tax Act 1961 although it’s a transparent entity under English Law (P&O Nedlloyd v. ADIT, Calcutta High Court WP 457 & 458 of 2005). 35 Jurisdictions in all have adopted the rules under dual resident entities, and 38 jurisdictions have adopted the rules under elimination of double taxation.

4. Dispute resolution :All jurisdictions are required under the MLI to adopt article 16 as a minimum standard, despite the fact that almost all treaties incorporate some form or the other with regards to dispute resolution. As a result almost 1700 non duplicative CTA’s are supposed to be impacted by the measures. For corresponding adjustments under article 17 most treaties already included a provision for the same even pre BEPS. However, there are only a handful of jurisdictions that have chosen to apply mandatory arbitration. It appears from the database that only about 30 jurisdictions have opted to apply for arbitration provisions. Interestingly only eight jurisdictions have chosen the “independent opinion” approach. On the face of it, it appears that most of the jurisdictions have a reluctance to get into arbitration possibly because of independent sovereignty, and also due to uncertainty over resources, both human and financial to handle arbitration procedures. Norway for example, takes the position that its own treaties, which it has renegotiated, have a better scope.

5. Issues of adoption : Majority of the states who are signatories to the MLI convention have already or are in the process of adoption of the same with respect to their legislative features. Depending on the legal structure whether monistic or dualistic with reference to adoption into domestic law, the ratification process might entail further steps. In some cases it has been much simpler, bought to the executive to vote(New Zealand) or the authority been delegated to government (Singapore), however, there are some cases where legal structures are such that it might need cascading approvals it has gone through a number of approvals (Belgium).

   There has been a varying degree of success with regards to ratification. For example some states have cited lack of adequate budgetary support with regards to assessment of impact (Mexico, New Zealand). In many jurisdictions the opinion is that parliament has just given approvals out of a sheer need for tackling base erosion and profit shifting without any actual impact analysis, which explains the perspective of there being no substantial difference between the clauses chosen by the relevant department of Finance and the executive approval thereof. An interesting note can be made of the French position of its Court of Accounts wherein they have observed that the discrepancy between the scarcity of the quantitative analysis and of the human resources invested by the tax authorities during the ratification process are in sharp contrast with the financial amounts at stake, in particular reference to a country such as France which is a resident state for many MNE’s. (Duff and Gutmann-IFA Cahiers)

   Publishing of synthesized texts has been a follow-through, in accordance with the guidelines of the OECD on the same. However, its pertinent to note that synthesised texts in many jurisdictions are of a guiding nature without any legal standing whatsoever. The underlying principle of mistrust of the taxpayer by authorities holding in good vogue. There have been examples of countries preparing synthesized texts without consulting a treaty partner and India stands exemplary in that its version of synthesized texts actually contradict the other sides of UK, Japan and Singapore amongst others. An interesting observation with regards to synthesized texts has been in Liechtenstein with regards to consolidation of synthesised texts wherein the legal provisions published in the official gazette must be shown in a consolidated fashion. This obligation has not found proper ground, since, after internal debates thereof, it was arrived that the tax authorities took the view that the provisions of the MLI do not have a similar effect on a CTA as a revision protocol, which implies that the MLI and the relevant treaty would have to be read hand in hand. Another interesting departure from the OECD practice can be noted from France wherein the authorities published afresh novel CTA’s, to project the relevant treaties as modified by the MLI.

6. Issues of interpretation: One of the very basic issues that arise out of the MLI is the language. The MLI and its official statement are to be in two official languages English and French, and any translations by the AD HOC committee ( https://www.oecd.org/tax/treaties/multilateral-convention-to-implement-tax-treaty-related-measures-to-prevent-beps.htm). This by itself raises questions of interpretation. It might be that translated versions of the MLI are used, or it might be that the CTA itself is in a different language. Explanatory statement no 317 of the MLI provides an answer to this as thus “Accordingly, where questions of interpretation arise in relation to Covered Tax Agreements concluded in other languages on in relation to translations of the Convention into other languages, it may be necessary to refer back to the English or French authentic texts of the Convention”.At the end of the day, it is not necessary that Competent Authorities or Courts be native English users and their interpretation regarding the same is going to matter a lot. Human behaviour cannot be ruled out.

   One common ground that arises with regards to interpretation is how will private laws distinguish between the operative interaction between the MLI and substantiative interaction of the CTA. This follows from the very nature of the MLI which in certain terms is a hybrid instrument. An interesting observation can be found in the explanatory statement in para 12 “It should be noted that while in some cases, as noted below, the provisions of the Convention differ in form from the model provisions that were produced through the BEPS Project, unless noted otherwise, these modifications are not intended to make substantive changes to those provisions. Instead, they are intended to implement the agreed BEPS measures in the context of a multilateral instrument that applies to a widely varied network of existing treaties.” Furthermore, in para 13 of the explanatory statement “It will not function in the same way as an amending protocol to a single existing treaty, which would directly amend text of a Covered Tax Agreement; instead, it will be applied alongside existing tax treaties, modifying their application in order to implement BEPS measures.” Also, para 12 substantiates that article 3-17 related to BEPS measures “should be interpreted in accordance with the ordinary principle of treaty interpretation, which is that a treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in light of its object and purpose.”, which is more or less in line with the article 31(1) of Vienna Convention on the Law of Treaties and the observations of the House of Lords in the case of n Fothergill v. Monarch Airlines Ltd [1981] AC 251 and the rulings in the landmark Commerzbank AG case (CIR v. Commerzbank AG [1990] STC 285, 63 TC 218 (1990) STC 285).

   However, again in an ironical twist on what parameters can the Explanatory statement of the MLI be regarded on legal fortitude? Very interestingly while the official texts might validly draw from article 31 (1) of the Vienna Convention, the fact of questioning the validity of the Explanatory notes on the same grounds is very dicey. Again, the question arises whether it can constitute valid preamble and annexes under Art 31(2), it might do justice to recall that the same are prepared by an Ad-Hoc group, and thus it might not meet the standards of “agreement” of Article 31(2)(a), or in consonance with Article 32(2)(b), as universal consensus on the Explanatory Statements do not exist.

   So, it could actually satisfy the supplementary criteria under article 32(a) as supplementary material. Very interestingly although Russia has taken the stance of Explanatory statements being in consonance with Article 32(a), the Swiss have taken the position that the Explanatory statements fall with the definitions under Article 31(2).

   Belgium, one of the signatories has gone to the extent of making the explanatory statement codified law in its jurisdiction by submitting the Explanatory statement along with the instrument before its parliament for approval.

   It follows that there will be ambiguities in interpretation relating to the operation of MLI, alongside the operative mechanism of the MLI would result in issues regarding interpretation of the underlying CTA. Interestingly, in an observation by Duff and Gutmann in the IFA Cahiers Norway has reserved the right for the entirely of article 17 not to apply as it believes that its covered tax agreement with India already contains the provision for the corresponding adjustment described in the article. However, the on the Indian side there is a note that the view of a closer look at article 9(2) of the covered tax agreement reveals that the other sate “may make an appropriate adjustment” while article 17(1) requires such an adjustment to be mandatorily made. Such a provision which gives discretionary powers to a contracting state to make cannot be reserved, which again is an issue of interpretation by Norway. It is likely that such issues would continue to arise further in case of interpretation.

   Another important aspect is the “ex post” approach to the modification proposed by the MLI. Can such a modification result in a different interpretation of a pre-MLI status? In this case a reference might be made to Article 28 of the Vienna Convention which expressedly forbids such actions. However, as with all, even this aspect is riddled with complexities.

   An interesting question would be when a jurisdiction considers the wordings of an article to be merely clarificatory in nature and decides not to implement them, like for instance the position of India on Article 5(6). Does it lead to a conclusion that non-notified portions of a CTA be used to interpret ?

   At the end of it all, it should be borne in mind that private international law considerations of various jurisdictions and their interpretations regarding the same by courts will have an overbearing impact on the entire scope of the MLI. Be it the German Treaty Override case, or Black v. R 2014 (TCC 12), it should be borne in mind that jurisdictions might be prone to apply the Revenue Rule (Govt of India v. Taylor) in order to protect their own.