BEPS THROUGH THE LOOKING-GLASS: WHERE DO WE STAND?
As regular followers of this blog will know, the OECD, a club of rich countries, has a mandate to reform international tax rules from the G20, with a project known as Base Erosion and Profit Shifting (BEPS). The project is due to end by December 2015: and work on the 15 points in the Action Plan is proceeding at a feverish pace, when compared to what’s usual in writing international tax rules.
The OECD is also making an unprecedented effort to consult widely on the proposals. Drafts are regularly released on most of the action points, and consultations are organised.
There are significant exceptions, however. The discussions on so-called Harmful Tax Competition are held behind firmly closed doors, as they concern the tax breaks governments offer to multinationals. The other aspect on which there has not yet been any consultation is the Multilateral Convention, which is intended to ensure rapid implementation of those proposals, since this needs changes in international tax treaties (which are agreements between countries as to how taxing rights are shared out between them, when there are cross-border investments).
Legislating for the world
Many of the changes can actually be implemented much more quickly, however. Some of the key proposals concern the rules on transfer pricing. These are contained in the OECD Transfer Pricing Guidelines, which began as a report in 1979, were reissued as Guidelines in 1995, and have been revised from time to time since then. They now run to over 350 pages.
These guidelines are not, formally at least, legally binding, but in many countries they are given legal status, as ‘interpretations’ of the tax treaty provisions. This is so in the UK under the Taxation (International and Other Provisions) Act (TIOPA) 2010, by which changes to the Guidelines need only be reported to parliament. This is not limited to OECD countries. Tax legislation in Tanzania also refers to the OECD Guidelines as an aid to interpretation, and this is so in many countries.
In Kenya, in 2005 the High Court even allowed Unilever to use a transfer pricing method relying on the Guidelines although those Guidelines were at that time not mentioned anywhere in Kenyan law. The judge stated ‘[w]e cannot overlook or sideline what has come out of the collective wisdom of tax payers and tax collectors in other countries’. Tax tribunals in India, which are snowed under with disputes caused especially by the dysfunctional and subjective nature of transfer pricing rules, regularly refer to the Guidelines. The Mumbai Tribunal in a recent decision even referred to one of the reports of the BEPS project as authority, even though its proposals have not yet been approved!
Obscurity and double-speak
Yet the BEPS project proposals are written in dense and technical language, accessible only to the hardened specialist. Those on transfer pricing are particularly obscure, because they put forward rewritten versions of the Guidelines, the implications of which are comprehensible only to geeks who already know the text backwards.
In fact, under the current proposals the Guidelines look set to becoming even more obscure, complex and contradictory. The term guidelines is a misnomer: they provide only a rambling and increasingly incoherent discussion of a variety of transfer pricing methods, with stylised examples. In addition to the five existing approved methods (see Box 3 on p7 here), the BEPS project proposes to add further ‘special measures’. So they offer a pick-and-mix range, allowing enough leeway for tax authorities such as those in Luxembourg to claim that their sweetheart deals are valid.
The problem is that the OECD is moving away from the ‘arm’s length’ principle, but does not want to admit this. Transfer pricing rules originally required strict application of the principle that the different parts of a multinational company group must be treated as if they are independent. Hence, they said that transactions between them must be priced on the basis of ‘comparables’ – prices actually charged in similar transactions by truly independent entities. This rule was introduced by the US in 1968, but they quickly found it did not work.
The reality, known to all tax practitioners, is that true comparables do not exist for transactions within an integrated multinational firm. The very good reason, understood by business economics, is that activities are internalised within a firm only if it is more profitable to do them in an integrated way than is possible by truly independent firms through the market. An integrated firm can generate higher profits because it controls unique technology, can produce more cheaply on a large scale or by sharing services and facilities over a range of products, and generally due to the synergy of combining related activities. Hence, its internal cost structures are quite different from those reflected in transactions between truly independent firms. Even true believers in the arm’s length principle accept that comparables cannot be found in countries with small and underdeveloped economies. In such cases, economic techniques are used to adjust for differences in market conditions, which at best produces an often wide range of possible prices.
How to apportion profits?
To deal with this, the US introduced a new method to tackle the problems, known as the ‘profit split’ method. The OECD reluctantly agreed to include it in the Guidelines in 1995. To convince the doubters, who rightly suspected that it entailed a rejection of the arm’s length principle method, it has been labelled a ‘transactional profit’ method. Instead of establishing clear and simple methods for apportioning the combined profits according to concrete factors reflecting the real activities in each country, it has continued to require an analysis of the functions performed, assets owned and risks borne by the separate legal entities. This means that the Big Four accounting firms and other specialist tax advisers still have plenty of lucrative work, both devising complex fragmented company structures, and preparing the documentation to justify the pricing policies. The profit split method is applied only in conjunction with other methods, or as a fall back. So companies dislike it, because it means the worst of both worlds. They still need to provide detailed explanations and documentation of all entities and transactions, but they find in practice that the final attribution of profits is arbitrary.
Increasingly, also, countries which are mainly hosts to multinationals have become dissatisfied at the low level of profits attributed to subsidiaries which have been located there to take advantage of factors such as availability of skilled but low-cost workers, or access to large potential markets. This has led China and India in particular to introduce the concept of ‘location savings’, to justify an additional adjustment of the attributable profits. The OECD countries don’t like this, but the BEPS project proposes to include such provisions in the Guidelines. The recent report on Special Measures also proposes measures ‘beyond the arm’s length principle’ to deal with some suggested ‘special cases’. These aim to deal with ‘cash box’ affiliates which can collect large profits under current rules, resulting from arrangements making them holding companies for ownership of intangibles, or financial management within a corporate group.
So the fragile consensus on transfer pricing is breaking down, as shown in the increase in disputes. For example, a large number of pending conflict cases between the US and India remain unresolved. Yet, these additional methods will obviously further complicate the transfer pricing rules. The Guidelines will become an even more tangled maze, which will need great familiarity to navigate. A careful reading of the extensive revisions proposed shows that there is likely to be a greater reliance on the profit split method. Yet the report on Profit Split produced under the BEPS project continues to base it on the ‘transactional’ approach.
A much easier way through the complications would be to extend, simplify and standardise the profit split method. This would particularly mean specifying the allocation keys for apportioning the profits. Tax advisers strongly oppose this, and assume that the OECD would not accept it, as the BEPS Action Plan rejected ‘formulary apportionment’ (FA), as does the OECD Guidelines. However, they define FA as apportionment by a formula ‘fixed in advance’. Profit split clearly is a profit apportionment method, but could be based on specific allocation keys suited to various business models, with clear principles for devising appropriate keys for new models.
The Looking-Glass world of international tax
The BEPS project is therefore now reaching its moment of truth, as the various proposals come together. Unfortunately, it is very difficult to properly understand the issues at stake even for experts, let alone the average person in the street. The discussions seem to be taking place in something like Alice’s looking glass world, presided over by Humpty Dumpty:
“When I use a word,” Humpty Dumpty said, in a rather scornful tone, “it means just what I choose it to mean – neither more nor less.” “The question is,” said Alice, “whether you can make words mean so many different things.” “The question is,” said Humpty Dumpty, “which is to be master – that’s all.” Alice Through the Looking Glass.
This weird outlook was on view recently at the OECD public consultations on 21-23 January on various BEPS proposals, now available online. For example, the first day concerned Artificial Avoidance of the PE Definition. This concerns the ways in which multinationals can fragment their functions among different affiliates, and attribute profits to those in low-tax countries — as revealed in publicised parliamentary hearings in many countries, on companies including Apple, Google, International Harvester, and Starbucks. Yet it would be very hard to grasp this was the issue from listening to the debate at the consultation.
This is not deliberate obfuscation, although specialists do like to protect their expertise behind a barrier of specialist language and obscure concepts. But here it seems to go further. Tax advisers seem to believe in the fantasy land that they have created of separate corporate entities and sham contracts. Hence, much of that discussion focused on the nature of so-called commissionaire contracts in civil law countries.
The other main problem is that the participants in the discussions mostly belong to the same cosy club. They know each other well, and speak the same language (English of course, but also taxspeak). This includes the government officials, since of course many advisers were formerly government (and OECD) officials. Yet this is considered to be an asset, since it proves that a person knows the right concepts and language. This was starkly revealed in the consultation on Friday 23 January on Dispute Resolution. Disputes over how profits should be allocated are dealt with under the so-called mutual agreement procedure (MAP) in tax treaties. This involves consultations between the responsible officials in each of the tax authorities concerned, who are referred to as the Competent Authority (CA). Although the taxpayer is not directly a party to the procedure, they can usually make representations, and usually employ advisers to do so. One of the proposals in the report is that the CA should be an independent office in the tax authority, to be free from influence from superiors and colleagues, enabling the dispute to be resolved in an impartial and objective manner.
Yet several participants in the discussion, now working as tax advisers, had previously served as CAs in tax authorities, notably Mike Danilack of the US, now with PwC and speaking on behalf of the US Council for International Business. Nobody seemed to see any problem with the revolving door, which has sent so many CAs into such lucrative jobs in the private sector. They were not coy, rather actually proud, of having sat at both sides of the table.
Our submission pointed out that a system that is ‘principled, fair and objective’, as suggested in the report, should require decisions to be based on reasoned explanations, which should be published. This suggestion was dismissed out of hand. At present, these conflicts are dealt with in secret, even though they often involve multi-million dollar differences. The details are known only to members of this closed world, tax officials and advisers. This is obviously very valuable information, which enables those individuals to earn substantial fees when they go into private practice. Yet they apparently have no difficulty in believing that they can act in a ‘principled, fair and objective’ manner when they sit on the government side of the table.