Oxford University Centre for Business Taxation
On November 29, the Centre for Business Taxation (CBT) convened a conference in London on The Future of the Arm’s Length Principle (ALP) with speakers drawn from business, the OECD and academia. This was also an opportunity to mark the publication of a new book by CBT Associate Fellow Richard Collier and a previous head of Transfer Pricing at the OECD, Joe Andrus, entitled Transfer Pricing and the Arm’s Length Principle After BEPS (OUP).
The subject of the conference and the book is topical. There has recently been increasing interest in ways to tax corporate income which may radically alter, or remove entirely, the current reliance on the ALP. Two examples of proposals that would entirely eliminate the ALP are the earlier US proposal for a Border-Adjusted Tax, and the European Commission’s proposal for the CCCTB. The current discussions on taxing digital business would also suggest displacing the current degree of reliance on the ALP. There is also the ongoing effort to complete work initiated in BEPS – on profit splits, financial transactions, and permanent establishment (PE) attribution – in respect of which there is no ready consensus.
A number of themes emerged from the discussion at the conference, both in relation to the current state of the ALP and the prospects for its future. The broadest – and unresolved – question was about the overall future of the ALP. Some had the ALP on life support, or hanging on a cliff by the fingertips. Perhaps not surprisingly, this view was not shared by the OECD, which argued that the ALP was generally working well
An early message in the conference from specialist transfer pricing practitioners was that, whilst there is in theory a single ALP rule, in practice it is subject to multiple – and therefore inevitably inconsistent – interpretations and compliance obligations. There is also a marked asymmetry in the approach of states according to whether the taxpayer in any particular case has generated profits or losses.
A recurrent concern was the complexity in practice of operating and administering the ALP system. It was generally acknowledged that levels of complexity had risen sharply following the new transfer pricing requirements from the OECD’s BEPS project (for example, it has significantly expanded compliance obligations relating to: the new approach to delineating a transaction for the purposes of a transfer pricing analysis; the newly-introduced 6-step risk framework for pricing risk within MNC groups; and the proliferation of dependent-agent PEs). The debate addressed whether the level of complexity meant that the system was approaching, or had even passed, the point at which all the various detailed requirements of the ALP could neither be complied with by taxpayers nor policed by tax administrations. The difficulties caused to the dispute and litigation process involving transfer pricing issues were also discussed.
There was a general acknowledgement that some parts of the BEPS work on the ALP had been helpful developments in the context of the current transfer pricing rules, for example: increasing the focus on two-sided transfer pricing analysis, rather than simply looking at the return due to a single tested party, with no regard to where and why any residual income is allocated within the group; the transparency agenda; and some positive steps on dispute resolution. However, some deep concerns on the viability of other parts of the BEPS package were expressed. For example, the new approach to PEs and the new BEPS risk framework came under recurrent criticism for a variety of practical, technical and policy reasons. Here it was also argued that there are at least two problems of principle. First, the new risk framework seems in conflict with the ALP itself, and second, the ALP simply cannot deal with the allocation of risk between two related parties, since in truth the risk is not borne by either party, but by the ultimate shareholders. It was also noted that the BEPS package had in some important respects failed to deliver on areas that in the run-up to BEPS had been identified as high priorities.
The discussion of the future prospects for the ALP was broken down into two broad directions. The first concerned priorities for future actions on the basis that the current ALP system is broadly retained. The second considered possible reforms of a more significant nature, adapting or even replacing the current reliance on the ALP.
On priorities for future actions to bolster the ALP, various points were discussed including: addressing perceived weaknesses in the existing rules (e.g. the BEPS risk framework); pursuing open significant issues (e.g. the ongoing work on profit splits); and resolving issues that lack clarity under the current rules (e.g. the relevance of any anti-avoidance theme under the ALP).
The discussion of the wider reforms began with some consideration of the accommodation of some form of “digital” tax. There was widespread agreement that the digital issue may trigger a fundamental re-appraisal of the current position and the future direction for taxing corporate income. The discussion of wider options for adapting the ALP included the use of a possible hybrid model under which routine functions within a MNC would be rewarded by the ALP, with residual profits taxed in the state of the relevant market sale.
Whilst acknowledging the evident differences in views expressed, it is hard to resist the conclusion that there remains a significant level of concern about the current ALP system. Yet – in this conference, at least – there appeared to be no consensus supporting any alternative system, or even modification to the current system. The future of the ALP is uncertain.
This blog is an account of the recent CBT conference on The Future of the Arm’s Length Principle.
Recent research from the Centre for Business Taxation relevant to this blog includes:
Back to top of article