Tax Treaties, Transfer Pricing and Financial Transactions Division Centre for Tax Policy and Administration Organisation for Economic Co-operation and Development By email: taxtreaties@oecd.org 9 April 2014 Dear Sirs Discussion draft on Action 6 (Prevent Treaty Abuse) of the BEPS Action Plan We wish to make a submission with respect to the public discussion draft on BEPS Action Plan 6 (the Discussion Draft ). Matheson is an Irish law firm and our primary focus is on serving the Irish legal and tax needs of Irish and international companies and financial institutions doing business in Ireland. Our clients include over half of the Fortune 100 companies. We also advise 7 of the top 10 global technology companies and over half of the world s 50 largest banks. We are headquartered in Dublin and also have offices in London, New York and Palo Alto. More than 600 people work across our four offices, including 75 partners and tax principals and over 350 legal and tax professionals. 1 Executive summary The main points of our submission are as follows: In our view, the proposed limitation of benefits clause should not be adopted, as it creates bias in favour of larger jurisdictions. Jurisdictions with smaller economies which are reliant on a broad-based investor profile would be disadvantaged. The proposal would also create complexity and uncertainty and is in conflict with the fundamental freedoms of the European Union. If a limitation of benefits clause is retained, it must be refined fundamentally. Such a clause must contain a derivative benefits clause and the reference to recognised stock exchanges must be expanded to include all recognised OECD stock exchanges.
The tax residency tie-breaker test for dual tax residency should retain an objective measure by which taxpayers can determine their tax residency position and which would underpin a mutual agreement procedure. The proposed limitation of benefits provision should not apply to collective investment vehicles. Instead, the approach adopted by the OECD in its 2010 report on collective investment vehicles should be applied and retained. The proposed limitation of benefits provision should not apply to securitisation companies. Cross-border securitisations are an important part of the financial system, which would be materially impeded by an application of the proposed limitations of benefits provision. 2 General comments on the limitation of benefits provision We understand and agree with the intent of the OECD to target the abuse of double taxation agreements. Proposals to counter such abuse should be clear, consistently applied and should not inadvertently impede genuine substantive commercial arrangements. With this objective in mind, we would make the following points: 2.1 We agree with the proposal to introduce an anti-abuse provision into the Model Convention as this mechanism should provide the flexibility to counter treaty abuse in a targeted manner. 2.2 In our view, the proposed limitation of benefits provision should not be adopted. There are at least three reasons for this conclusion: 2.2.1 The limitation of benefits provision is a mechanism which is inherently biased in favour of larger jurisdictions and furthermore is biased against jurisdictions with smaller economies which are more reliant on a broad-based investor profile. This is because limitation of benefits clauses automatically grant treaty access where the majority of investors are same-country investors, but imposes much more restrictive conditions for smaller economies, which inevitably have a smaller domestic capital base with local businesses looking to raise capital from non-local investors. 2.2.2 The proposed limitations of benefits provision would also introduce significant complexity and uncertainty into cross-border dealings. This is because many of the concepts in the limitations of benefits provision are subjective in nature. For example, it will require a subjective determination to conclude whether a company has an active trade or business. It will also require a subjective determination to conclude where a company s primary place of management and control resides, in cases where management functions are split between different jurisdictions, which is commonly the case. These uncertainties will likely lead to an increase in competent authority cases, tying up resources of both taxpayers and tax authorities. 2
2.2.3 Finally, in our opinion, the proposed limitation of benefits provision is in conflict with the principles on which the EU is founded. In particular, it cuts across the freedom of establishment, the freedom of capital, and the freedom of services. Therefore, in our view, the limitations of benefits provision should not be adopted. Instead, the OECD should focus on the introduction of a general antiavoidance rule, together with certain targeted anti-avoidance rules. 2.3 Should a decision be made to progress with the inclusion of a limitation on benefits provision in the Model Convention, we would have the following comments: 2.3.1 A derivative benefits provision must be included. The policy reasons in favour of introducing such a provision strongly outweigh any perceived potential for abuse. In particular, if all the investors in a company are entitled to the same degree of treaty benefits, the company itself should not be denied such access. Furthermore, there seems to be no policy or logical basis to restricting the derivative benefits provision to seven or fewer persons. For example, if a company is owned by 20 investors, all of whom are entitled to the same degree of treaty benefits, why should the company itself be denied access? In our view, there should be no numerical limitation on such a derivatives benefit provision. It is important to note, in this regard, that a derivative benefits provision is particularly important to small open economies, to ensure that intermediate group companies should not be prevented from claiming treaty benefits, where the parent company is itself entitled to treaty benefits. 2.3.2 Active trading of shares on any recognised stock exchange in the OECD/G20 (and not just exchanges in the home country of the relevant company) should always satisfy paragraph 2(c)(A) of the limitations of benefits provision. Any restriction to stock exchanges located in the home jurisdiction will create a strong and unjustified bias in favour of larger jurisdictions with well-developed stock exchanges. There is no policy reason whatsoever to discriminate against companies resident in smaller economies who chose (for entirely reasonable commercial reasons) to raise capital by listing their shares on the larger and more liquid stock exchange of another jurisdiction. Any such listing (irrespective of which stock exchange is used) indicates clearly that treaty-shopping is not a concern for the relevant company. 2.3.3 The active trade or business test at paragraph 3(b) of the limitation of benefits provision will be difficult, in most cases, to apply. In particular, it will prove challenging to determine whether the business conducted by one company is substantial in comparison to the business conducted by another company, where the functions and businesses of those companies are different, though complementary. Consequently, strong, clear and comprehensive commentary will be required in respect of this provision to ensure consistency in 3
interpretation. This commentary should provide detailed examples that deal with situations that are most likely to cause differences in interpretation between Contracting States. 2.3.4 Furthermore, the active trade or business test should be amended to include that the active management of investments and subsidiaries would qualify as an active trade or business. Multinational groups frequently have intermediate companies in their corporate structure, for various commercial reasons including centralising risk management, granting security over certain assets, regulatory requirements, centralising regional operations, and as a result of the acquisition of new groups / companies. The management of such investments can be an active business in itself in many cases. The proposed limitation of benefits clause should recognise this, and permit treaty access in such situations where sufficiently active business functions are being carried on with respect to such investment management operations. 2.3.5 Reference also needs to be made in the limitation of benefits provision to non-corporate entities which are persons for treaty purposes. For example, unit trusts are generally treated as persons for treaty purposes, and can have their units traded on stock exchanges, and have their units owned by multiple investors. Consequently, paragraphs 2(d), 2(e) and 5 need to make reference to such noncorporate entities, and their units. 2.3.6 The reference in paragraph 2(e)(i) to persons who are residents of that Contracting State should be expanded to refer to persons who are resident of either Contracting State. This appears to be a logical and fair amendment, and reflects the approach adopted in the subsequent sub-paragraph 2(e)(ii). 2.3.7 The limitation of benefits provision should be limited to arrangements between associated persons, which is the focus of the BEPS project, and should not apply to dealings between independent persons. A general anti-avoidance rule, together with certain targeted antiavoidance rules, can deal with any perceived tax avoidance arising with respect to dealings between independent parties. 2.4 The proposal to have competent authorities agree on where a dual-resident taxpayer is resident for the purposes of the relevant double taxation agreement is positive. However, as the resources of competent authorities are increasingly being stretched, this may potentially result in lengthy periods during which taxpayers will not have certainty of tax residency pending completion of the mutual agreement procedure process. An objective tie-breaker test on a company s residence should be retained (for example, the effective place of management test), together with detailed commentary on the factors and criteria that will be considered by the tax authorities in determining tax residence. Competent authorities should commit to agreeing on dual-residence requests within a specified short period of time, with appropriate arbitration provisions. 4
2.5 The proposed burden of proof for the general anti-avoidance rule, namely one of reasonable to conclude, is much too low. It would result in too much uncertainty in the application of this paragraph 6 and would leave taxpayers in an unfairly weak position. We would note, in this regard, that tax considerations will always feature in international dealings and investments. However, for this general anti-avoidance rule to apply, it should be clear that, on the balance of probabilities, one of the main purposes of a particular transaction was to achieve a treaty benefit. We would therefore submit that such term be removed from paragraph 6, so that it would be applicable where having regard to all the relevant facts and circumstances, one of the main purposes of any arrangement or transaction was obtaining that benefit. 3 Collective investment vehicles and the limitations of benefits provision The proposal to introduce a limitation of benefits provision into tax treaties, as currently drafted, would have a significant (and we believe unintended) detrimental effect on collective investment vehicles ( CIVs ). For these purposes, CIVs mean collective investment vehicles which are widely-held, hold a diversified portfolio of securities and are subject to investor-protection regulation in the country in which they are established. In this regard, we would note the following points: 3.1 The question of how treaty benefits should apply to CIVs has recently been the subject of thorough OECD studies and in-depth reports, including reports of the Informal Consultative Group on the Taxation of Collective Investment Vehicles and Procedures for Tax Relief for Cross-Border Investors published in 2009, and the report adopted by the OECD Committee on Fiscal Affairs in 2010 entitled The granting of treaty benefits with respect to the income of collective investment vehicles (the 2010 Report ). This 2010 Report came to the general conclusion that CIVs should be entitled to the benefits of treaties, on their own behalf, and suggested specific provisions which member states may wish to include in their tax treaties to deal with any avoidance concerns. 3.2 We agree with the conclusions of the 2010 Report. It is entirely appropriate that CIVs should be granted treaty benefits. By their nature as regulated investment vehicles designed to promote collective investment, CIVs are not aimed at treaty shopping. 3.3 Consequently, the approach recommended by the OECD in the 2010 Report should continue to be the approach for CIVs, instead of the application of the proposed limitations of benefits provision. 3.4 Our concerns are focused, in particular, on the impact of the proposed limitation of benefits provision on cross-border CIVs, that is CIVs whose units interests, or shares are distributed on a cross-border basis. Unlike domestically-distributed CIVs, cross-border CIVs (by their nature) will not have a majority of investors resident in the jurisdiction of residence of the CIV itself. This was acknowledged in the 2010 Report which stated: The global CIV market is one in which the CIV and a significant portion of its investors are located in different countries. The global 5
CIV can be much more efficient it can benefit from the economies of scale described above to a greater extent than smaller CIVs. The 2010 Report highlighted the importance of cross-border CIVs, it explained: International diversification of investment portfolios is becoming more significant. For example, over 25% of all equity assets held by US CIVs are issued by non-us companies. About 30% of the assets of UK CIVs are invested outside the United Kingdom. More than onethird of the assets of Japanese CIVs are foreign securities. Assets of Luxembourg, Swiss and Irish funds are predominantly invested outside of their home market. As more investments are made crossborder, the issue of CIVs qualification for treaty benefits is becoming increasingly important. As cross-border CIVs (by their nature) will generally not have a majority of investors resident in the jurisdiction of residence of the CIV itself, paragraph 2(e)(i) of the proposed limitation of benefits provision will almost never be capable of being satisfied by a CIV. Nor would the derivatives benefits provision (as currently drafted) offer any benefit, as CIVs (by their nature) will generally have far more than seven investors. 3.5 Cross-border CIVs are good for the financial market and the stability and efficiency of the financial sector. Cross-border CIVs have been promoted both by regulators and law-makers over the past 30 years. As explained in the 2010 Report: regulators see the benefits of a smaller number of larger CIVs, and regulatory changes, such as the UCITS Directive within the European Union, are designed to encourage global business. The EU has sought since 1985, through legislation on undertakings for collective investment in transferable securities ( UCITS ), to incentivise the establishment of cross-border funds. One of the key aims of the recent UCITS IV directive was to enable UCITS located in one EU member state to be managed, distributed and administered by a management company located in another EU member state. The UCITS IV directive contained various provisions to enhance efficiencies in cross-border investment, including the introduction of a management company passport, a master-feeder structure and cross-border merger provisions. Similarly, the Alternative Investment Fund Managers Directive ( AIFMD ) promotes cross-border investments. Although AIFMD is focused on regulating alternative fund managers rather than the funds themselves, it encourages cross-border fund distribution by facilitating the marketing of alternative investment funds to professional advisors across the EU. 3.6 Cross-border CIVs, in general, and both the UCITS and AIFMD regimes, in particular, will be materially and unjustifiably undermined, should a limitations of benefits provision be applied to CIVs. As cross-border CIVs would generally be unable to satisfy a limitations of benefits provision and would thus suffer 6
additional costs (eg, withholding taxes) on their investments, investors will be incentivised to move their investments to CIVs which can (potentially) satisfy the limitation of benefits provision, namely domestically-distributed CIVs which are resident in their own jurisdiction. Consequently, there would be a real financial incentive tending towards the segmentation of the CIV market, with domestically-distributed CIVs having a real commercial advantage of crossborder CIVs. We presume that it cannot be the OECD s intention to segment the CIV marketplace in this way. 3.7 In addition to these concerns in respect of cross-border CIVs, an additional general concern arises in respect of all CIVs. This relates to the intermediatednature of the CIVs. The vast majority of investments in CIVs are made through intermediaries such as securities firms, banks, insurance companies and independent financial advisers. This system allows for greater efficiencies as intermediaries aggregate their customers daily transactions and effect a single net purchase or a net sale each day in a nominee account held by the intermediary on behalf of all of its customers. As a result, it is common for CIVs to have layers of intermediaries between itself and its customers. In many cases, these intermediaries are not located in the country in which the investor is located. The result is that many CIVs (both domestically-distributed and internationally-distributed) will not know who their investors are. Consequently, the CIVs will generally not (in practice) be in a position to confirm whether or not their investors satisfy the requirements of paragraph 2(e) of the limitation of benefits provision. It should be noted, in this regard, that CIVs are subject to the FATCA regime, and that the FATCA regime has accepted that it is entirely appropriate for CIVs to rely on the intermediated distribution process (so that the CIVs are not obliged to identify their ultimate investors for FATCA purposes). We would therefore submit that CIVs should not be required to satisfy a limitation of benefit provision in order to obtain treaty access. Instead, the approach described in the 2010 Report should be applied with respect to CIVs. 4 Securitisation companies and the limitations of benefits provision The proposal to introduce a limitations of benefits clause into tax treaties, as currently drafted, would also have a significant (and we believe unintended) detrimental effect on securitisation companies involved in cross-border securitisation transactions. In this regard, we would note the following points: 4.1 Securitisation is a key element of the international financial markets, and offers a channel for borrowers to directly access capital markets. Residential and commercial mortgages, auto loans, trade receivables and bank lending are all funded, to a material degree, by securitisation. 4.2 The importance of securitisation has been recognised by global regulators: [the International Organisation for Securities Commissions] believes that securitisation markets can play a role in supporting economic growth... Securitisation markets potentially [make] bank lending less 7
sensitive to abrupt changes to the cost of funds, ultimately affecting the availability of finance to economic growth. For that reason, access to these funding sources may be important to those economies experiencing slow growth 1 The ECB welcomes the [PCS] initiative, which aims at increasing the attractiveness of asset-backed securities among investors and originating banks. A well-functioning ABS market in the EU would allow investors to diversify their investments and... thereby contribute to a smooth financing of the real economy 2 I would like to reaffirm that for the [EU] Commission securitisation is considered as an efficient mechanism to increase the credit availability and lower the cost of credit in line with the G20 s November 2010 report that noted that re-establishing securitisation on a sound basis remains a priority in order to support provision of credit to the real economy and improve banks access to funding in many jurisdictions. Furthermore, there is no question that it is in the private and public sector interest to reactivate securitisation markets 3 4.3 In the European market, it is frequently necessary to carry out securitisations on a cross-border basis. For example, a bank may wish to securitise loans which it has made in a number of jurisdictions or a multi-national company may wish to raise finance by securitising its trade receivables owing from customers in a number of jurisdictions. In these instances, a single securitisation company is established in one (typically EU) jurisdiction, to acquire the relevant financial assets (eg, the loans or receivables) owing from obligors resident in a number of different jurisdictions. 4.4 In such securitisations, the securitisation company will typically look to make sure that it has access to double tax treaties. This is to ensure that no withholding taxes will apply to income streams earned by the securitisation company. 4.5 Cross-border securitisation companies share two similarities to CIVs. First, investors in cross-border securitisation companies are international in nature; they are not normally limited to investors resident in the jurisdiction of the securitisation company. Second, the securitisation company will not, in most cases, know the identity of its investors; in the main, this is due to the fact that investors hold their investments in securitisation companies through clearing systems such as Euroclear, Clearstream and the Depository Trust Company of New York. For example, the investors in a securitisation of trade receivables originated by a European multinational company may invest by way of commercial paper or medium term notes, issued by the securitisation company, which are held in Euroclear. 1 IOSCO, Final Report on Global Developments in Securitisation Regulation, November 2012 2 Mario Draghi, President of the European Central Bank in a letter to the European Financial Services Round Table supporting the PCS Initative, June 2012 3 Emil Paulis, European Commission, speaking at AFME s Funding Conference in Madrid, November 2011 8
4.6 As a result, cross-border securitisation companies will not be in a position to satisfy the proposed limitation of benefits provision. They will generally not know who their investors are, and (even if they do) those investors will in the main not be resident in the jurisdiction of residence of the securitisation company. 4.7 Like CIVs, cross-border securitisation companies are good for the financial market and the stability and efficiency of the financial sector. Cross-border securitisation companies are being promoted by both regulators and central banks, especially within the European Union. Mario Dragi recently said We think that a revitalisation of a certain type of [asset-backed security], a so-called plain vanilla [asset-backed security], capable of packaging together loans, bank loans, capable of being rated, priced and traded, would be a very important instrument for revitalising credit flows and for our own monetary policy (our emphasis). This revitalisation can only happen on a cross-border basis. 4.8 It is therefore critical that securitisation companies are not prevented from accessing the benefits of tax treaties by reason of the imposition of a limitation of benefits provision. Securitisation companies should, instead, be subject to the general requirements of being a resident of a contracting state, being liable (or subject) to tax in that contracting state, and of being the beneficial owner of the income or gains earned. If a securitisation company satisfies these requirements, then it should generally be entitled to treaty benefits. It should not also be required to apply a limitations of benefits provision, which it would inevitably fail to satisfy. If a particular state has a concern about potential avoidance situations, the solutions outlined in the 2010 Report for CIVs could equally be applied to securitisation companies. Therefore, we would therefore submit that securitisation companies should not be required to satisfy a limitation of benefit provision in order to obtain treaty access. Instead, an approach similar to that described in the 2010 Report for CIVs should be applied with respect to securitisation companies. 5 European Union law and the limitations of benefits provision We understand that the rationale for introducing a limitation of benefits clause is to provide objective rules to identify clear situations that should not qualify for treaty benefits. We have concerns that a limitation of benefits provision goes beyond what is required to tackle treaty abuse and can capture genuine commercial structures. In particular, the limitation of benefits provision potentially impacts on certain of the EU fundamental freedoms, including a company s freedom of establishment, the freedom of stock exchanges to provide services to companies located in other EU Member States and the free movement of capital. For example: (a) a company resident in one member state of the European Union could be denied access to tax treaties (even assuming the derivative benefits provision is included) if that company was owned by ten investors resident in other EU member states; and 9
(b) a listed company resident in one member state of the European Union could be denied access to tax treaties if that company s places of management and control were divided between a number of different EU member states. Yours faithfully Whilst it might be said that the provision for an active conduct of a trade or business in paragraph 3(a) of the limitation of benefits provision would sometimes permit treaty access in such situations, in our experience this can be a very subjective and uncertain test. Treaty access can vary from year to year depending on the level of commercial activity being carried out in the relevant company. Furthermore, different tax authorities will take different views as to what constitutes an active trade or business, and different requirements of substance will likely be required by different tax authorities. The EU fundamental freedoms can be limited in certain circumstances, however, the ECJ in the Cadbury Schweppes decision of 2004 made it clear that anti-abuse provisions could only target the creation of wholly artificial arrangements by incorporating a company in another EU Member State only for tax reasons, without any business purpose or substance over form, thus lacking of an economic reality. Given the complex nature of the limitation of benefits provision and the potential for conflict with EU laws, we believe that the limitation of benefits clause should not be included and the focus in targeting double tax treaty abuse should be on developing the general and targeted anti-abuse provisions, and related commentary. Sent by email, bears no signature MATHESON 10