1. Carve-out of financial services from the scope of Amount A
According to the Report on the Pillar One Blueprint,[1] the proposed scope of Amount A is designed to capture multinational groups that are presumed to participate in a sustained and significant manner in the economic life of a market jurisdiction. To this end, the OECD has broken down such groups into the broad categories of Automated Digital Services (ADS) and consumer facing businesses (CFB).[2], [3] Yet, the OECD has also set out a list of activities to be specifically excluded from Amount A. When doing so, the OECD has proposed a carve-out for financial services (FS), i.e. banking, insurance and asset management.[4]
While it is acknowledged that the FS business should not generally involve business of the sort that is properly regarded as ADS, the OECD instead contends that the FS business may encompass CFB activities. Nevertheless, the OECD advocates an exclusion of the FS business based on its highly-regulated nature.[5] More specifically, the OECD stipulates that the applicable regulations generally require that appropriately capitalized entities are maintained in each market jurisdiction to carry on business therein. As such, in the OECD view, profits from CFB activities that arise in a market jurisdiction will generally be taxed therein to the effect that there is no further need for any Amount A re-allocation.[6]
2. The role of regulation and its implications
The goals and degree of regulation in the FS business differ depending on the jurisdictions. Among others, regulation is generally aimed at preventing market failures and maintaining the stability and integrity of the financial system. Additionally, regulators implement standards addressing costumer protection, thereby ensuring a fair treatment for costumers. In certain instances, regulatory authorities supervise all FS activities. In others, different bodies oversee specific industries in the light of their peculiar features.[7] Also, regulation may be – to a certain extent – centralized (like in the EU) or highly fragmented (like in the US).
Despite such local nuances, as the OECD has correctly acknowledged, the FS business is a highly-regulated business. It follows that the business activities are strictly interwoven with the applicable regulatory framework to the effect that much of the organizational structure of multinational groups operating in such business is eventually driven by what is required by regulators and what allows the most efficient use of the capital within such regulatory constraints.
3. Proposals for reform
Considering the political momentum, the proposed carve-out might represent a starting point for a few changes to some other basic rules regarding FS taxation without being tied up in the ongoing debate concerning nexus and profit allocation rules to taxing the digitalized economy. More specifically, if it is true – and indeed it is true – that the highly-regulated nature of the FS business underpins the carve-out from the scope of Amount A, a similar reasoning can be stretched beyond this exclusion since the regulatory requirements established under applicable supervisory laws and regulations determine more in general the way in which FS entities do business. Among others, this could be done in relation to the concept of permanent establishment (PE) and certain areas of the transfer pricing (TP) Guidelines.
3.1. The concept of PE
The prerequisites of local presence admitted for regulatory purposes do not currently match to the definition of PE for tax purposes. In several jurisdictions, local regulation requires the creation of either a local subsidiary organized under domestic laws or a local branch. In such cases, the concept of a branch might be reconciled with the definition of a fixed place of business, as set forth under the basic PE rule, whereas the concept of dependent agent PE is de facto not applicable. In certain countries, a local legal representative must be appointed with a view to fulfilling local compliance requirements. In the EU Single Market, the fine line between the right of establishment and the freedom to provide services is not clear-cut. This has raised some uncertainty as to what provisions should apply in specific circumstances, along with disputes between taxpayers and tax administrations regarding the existence of a PE.
This mismatch essentially stems from the fact that – under the existing framework – the tax and regulatory provisions are two legally independent sets of rules with their own purposes and standards to the effect that each of them should be carefully assessed separately. However, the substantive requirements laid down under regulatory provisions are incumbent on the way FS entities do business. It is indeed the local regulatory framework that ultimately determines whether (and to what extent) business activities are admitted in a specific jurisdiction, ahead of any relevant tax considerations.
Accordingly – even under the current circumstances, where the tax and regulatory provisions are two legally independent set of rules – a concrete assessment regarding the existence of a PE for tax purposes should be well-entrenched with a preliminary analysis regarding the applicable regulatory framework. In fact, any such assessment cannot be isolated from a proper understanding of the framework in which the relevant FS entity carries out its activities.
In that sense, there are good grounds for the tax and regulatory concepts to be ultimately aligned. More specifically, it appears suitable that – leaving aside the circumstances under which a FS entity has a physical presence in the source country, that would otherwise trigger the basic PE rule – a taxable presence in the form of a dependent agent PE should more consistently arise only in the case where the foreign FS entity is legally admitted to transacting business in the country through a person that gives rise to a regulatory presence therein.
This alignment would facilitate a more consistent assessment regarding the existence of a PE against the legal and economic environment in which FS entities operate, most importantly against the activities that can be lawfully carried out in compliance with applicable regulatory requirements. Any redundant complexity should indeed be kept at an absolute minimum in view of preventing interpretation disputes between taxpayers and tax administrations, which could eventually turn out to be a potential source of double taxation issues.
Such alignment could be achieved through a (partial) redefinition of the dependent agent PE concept enshrined in double tax treaty law (arguably, through an amendment to the multilateral instrument) with a view to making it correlated to the definition adopted for regulatory purposes, in combination with some adjustments to the commentaries and the guidance to the attribution of profits thereto.[8]
3.2. TP Guidelines
The proposed carve-out might represent an opportunity to emphasize the relevance of the arm’s length principle (ALP) and its interconnection with the regulatory framework of the FS business.[9] Insofar the clear intention is to deviate from the ALP with regard to the calculation and allocation of Amount A through a formula,[10], [11] it seems that – in the OECD view – the ALP is per se no longer able to recognize value creation in (certain) business activities. This would create a mismatch with the current position given within the TP Guidelines that needs to be addressed. Indeed, in the latest version of the TP Guidelines, no revision has been done on the position on any global formulary apportionment mechanism, i.e. that should be rejected and that is not seen as a “realistic alternative” to the ALP.[12]
In that sense, there are good grounds to revise the TP Guidelines by addressing the FS business specifically in Chapter 1, Section B, with regard to the statement of the ALP, but also in Section D, regarding the Guidance for applying the ALP and especially in the new Chapter X on Financial Transactions released last February.
Certainly, it is true that the OECD in 2010 had already acknowledged the key role of regulatory measures for FS enterprises when dealing with the attribution of profits to PEs within the AOA Report,[13] but with the introduction of the new Section D.1.2.1. in the 2017 revision[14] the specific needs of regulated sectors have been mentioned only in a footnote, simply stating that appropriate reference should be made to the AOA Report.[15] Moreover, the same footnote is cross-referenced in the new Chapter X.[16] However, in light of the scenario depicted above, this does not appear sufficient, so that appropriate sections should be introduced in the TP Guidelines themselves addressing with specificity the Risk Assumption and the Financial Transactions for regulated industries.
Within the Risk Assumption concept in Section D.1.2.1., the TP Guidelines could further elaborate on how the control over the risk and decision-making concept[17] is aligned and interplays with the risk management systems that regulated entities must adopt as a mandatory requirement. For example, for insurance companies, it is important that decision-making processes and the organizational structure are defined and established as a part of their governance and risk management system within an individual assessment of their risk and solvency situation (“O.R.S.A.” process),[18] which is at the core of Solvency II Pillar Two framework[19] for insurance entities.
With regard to the Chapter X, the introduction of specific sections for the FS business might also represent an opportunity to better align the TP Guidelines with the AOA Report where their actual application leads in a potential mismatch in the arm’s length result under the application of Article 7 vs. Article 9 of the OECD Model.[20]
For insurance, to seek alignment between the TP Guidelines and the AOA Report, one important revision pertains to the outcome of the attribution and recognition of investment income. In the case of a PE, according to the AOA, when the functional analysis has determined that the PE alone has performed the KERT function (i.e. underwriting), the PE will be attributed the newly created insurance risk, together with both the associated underwriting income and investment income from the assets required as surplus and reserves to support the insurance risk.[21] Thus, wherever (solely) underwriting is performed, investment income should be always and automatically attributed to the PE, even if there are no people therein having investment skills/capabilities exercising control over the risk associated with the investment assets on an actual conduct basis, as required by Chapter 1 of the TP Guidelines. Therefore, the insurance PE attribution would not have led to the same conclusion had it been a legal insurance associated entity performing only underwriting. In fact, based on the wording of the TP Guidelines, there is no automatic allocation of the investment return to the underwriting function. Rather, to attribute any additional return,[22] the guidance requires that: (i) the associated entity performing underwriting has the requisite skills, including investment skills, and experience at its disposal,[23] and (anyway) (ii) the return derived from the investment of the premiums is allocated to the member performing underwriting but “in accordance with the guidance in Chapter 1”.[24] As a consequence, the latter wording mentioning underwriting as the rationale for the attribution in line with Chapter 1 would create a clash in itself and a contradiction with the AOA outcome, because the application of Chapter 1 requires the investment return to be allocated where underwriting is performed only if the same associate entity also exercises[25] control over the risk associated with the investment assets and has the financial capacity to assume that risk[26] determined by the actual conduct of the parties.[27] Thus, not only should the AOA wording be aligned with Chapter 1 of the TP Guidelines, but also the reference to underwriting function as the (sole) rationale for the allocation of investment return for investment assets should be removed.
[1] See OECD/G20, Tax Challenges Arising from Digitalisation – Report on Pillar One Blueprint, 12 October 2020, OECD 2020.
[2] See OECD/G20, Op. cit., pp. 19 et seq.
[3] According to the Report on the Pillar One Blueprint, the activity test is to be combined with a threshold test, which is out of the scope of this contribution. For further comments, on the threshold test, see OECD/G20, Op. cit., pp. 61 et seq.
[4] See OECD/G20, Op. cit., p. 49.
[5] The OECD had already acknowledged the key role of regulatory measures for FS enterprises when dealing with the attribution of profits to permanent establishments. To this effect, see OECD, 2010 Report on the attribution of profits to permanent establishments, 22 July 2010, OECD 2010, Parts II-IV.
[6] See OECD/G20, Op. cit., pp. 52-53.
[7] With specific respect to the insurance industry, see Daniele Frescurato, The concept of permanent establishment in the insurance industry, not yet published, para. 2.1.
[8] A specific proposal for the insurance industry has been elaborated in Daniele Frescurato, The concept of permanent establishment in the insurance industry, not yet published, para. 6.6. Similar proposals could be extended to other industries in the FS business.
[9] In the FS business, it is therefore important to explicitly acknowledge that not only the exclusion of the FS business is based on its highly-regulated nature and however due to the nature of CFB capital-intensive activities that arise in a market jurisdiction, but especially because the ALP should be the only basis with which value creation is evaluated because it is the most reliable founding concept to represent the way in which FS entities do business.
[10] See OECD/G20, Op. cit., para. 496, p. 123. That should not be confused with the application of the transactional profit methods but rather represent a “simplified proxy” – i.e. a simplifying convention – of the portion of the residual profit of a business in the economy of a market jurisdiction. On this regard, see OECD/G20, Op. cit., para. 507, p. 126.
[11] As a result, only the Amount B will be clearly referring to the ALP. See OECD/G20, Op. cit., para. 651, p. 160.
[12] It is worthwhile remembering that this position is based on the fact that predetermined formulae are arbitrary and disregard market conditions, the particular circumstances of the individual enterprises and the management’s own allocation of resources – thus producing an allocation of profits that may bear no sound relationship to the specific facts surrounding the transaction – and also because they may in fact present intolerable compliance costs. On this regard, see OECD Transfer Pricing Guidelines, paras. 1.21, 1.25, 1.27 and 1.32, pp. 40 – 43. It is therefore advisable to acknowledge the mentioned mismatch on the purpose of the ALP in order to reduce complexity, enhance tax certainty and minimize burdens for tax administrations and taxpayers alike as the new Inclusive Framework intends.
[13] It is in particular acknowledged the tied relationship between functions, assets and risks according to which assets and risks follow functions and capital follows risks – and not the other way round – and due to this special relationship between risks and financial assets in those specific sectors, the Authorised OECD approach (AOA) uses the specific key entrepreneurial risk-taking function (KERT) function terminology in describing the functions relevant to the attribution of both risks and assets, and ultimately the capital necessary to support these risks. See OECD, 2010 Report on the attribution of profits to permanent establishments, 22 July 2010, OECD 2010, Parts II-IV, para. 51, p 75.
[14] This section relates to the analysis of risks in commercial or financial relations introduced in the TP Guidelines in light of the implementation of the BEPS Action 8-10.
[15] See 2017 OECD Transfer Pricing Guidelines, Chapter I, Section D.1.2.1, fn. 1, p. 53.
[16] See 2020 OECD Transfer Pricing Guidance on Financial Transactions, para. 10.15, fn. 3.
[17] See 2017 OECD Transfer Pricing Guidelines, Chapter I, Section D.1.2.1, para. 1.65, p. 56.
[18] The Own Risk and Solvency Assessment (ORSA) is an internal process undertaken by an insurer or insurance group to assess the adequacy of its risk management and current and prospective solvency positions under normal and severe stress scenarios. For further comments, see Naic Own Risk And Solvency Assessment (ORSA) Guidance Manual and Risk Management and Own Risk and Solvency Assessment Model Act (#505).
[19] See Delegated Regulation (EU) 2015/35, Art. 306.
[20] In particular, it should be borne in mind that the AOA is based upon the principle of applying by analogy the guidance found in the TP Guidelines for purposes of determining the profits attributable to a PE and that the functional and factual analysis performs the same role in the comparability analysis in a PE context under Article 7 as it does in situations involving associated enterprises under Article 9, therefore in principle aiming at granting the same results at arm’s length. See OECD, 2010 Report on the attribution of profits to permanent establishments, 22 July 2010, OECD 2010, paras. 10 and 13, pp. 9 and 14.
[21] See OECD, 2010 Report on the attribution of profits to permanent establishments, 22 July 2010, OECD 2010, para. 107, p. 195 and para. 166, p. 207 with which the OECD further states that: “In the case of a PE jurisdiction that has required the non-resident enterprise to place particular assets in trust, it would be appropriate to attribute the investment income earned with respect to those assets to the PE to the extent that key entrepreneurial risk-taking function is performed by a PE in that location”.
[22] Hereby it is of course assumed that the mentioned underwriting activity has been accurately delineated via a proper functional analysis, according to which the underwriting return has been appropriately recognized further to the application of Chapter 1 of the TP Guidelines.
[23] See 2020 OECD Transfer Pricing Guidance on Financial Transactions, para. 10.199, p. 36.
[24] See 2020 OECD Transfer Pricing Guidance on Financial Transactions, para. 10.212, p. 38.
[25] One should bear in mind that Control over a specific risk in a transaction focusses on the decision-making and the actual performance of such decision-making functions relating to a specific risk of the parties to the transaction in relation to the specific risk arising from the transaction. See 2017 OECD Transfer Pricing Guidelines, Chapter I, Section D.1.2.1, paras. 1.61, 1.65, 1.66, 1.76, pp. 55 – 62.
[26] See 2017 OECD Transfer Pricing Guidelines, Chapter I, Section D.1.2.1, para. 1.65, p. 56.
[27] See 2017 OECD Transfer Pricing Guidelines, Chapter I, Section D.1.2.1, para. 1.120, p. 77. Anyway, ultimately the return derived from the investment of the premiums would be allocated to the member(s) of the MNE group that are assuming the risk associated with the investment asset risk. See 2020 OECD Transfer Pricing Guidance on Financial Transactions, para. 10.212, p. 38.
________________________
To make sure you do not miss out on regular updates from the Kluwer International Tax Blog, please subscribe here.
