One of the most significant issues facing the transfer pricing community, and the wider business society, concerns the taxation of the digital economy. Generally, two challenges are raised regarding this issue: first, that digital companies supposedly pay low taxes due to the excessive use of loopholes and, second, more intricately, that the rise of the digital economy leads to a new dynamic between market states and the host countries of these companies. In particular, market jurisdictions claim that a high value is generated from the data provided by users in their countries and that this value is not captured fairly.
Regarding the first charge, over the past five years, many steps have been taken in the international tax framework under the BEPS approach to close perceived loopholes around non-residency status and, more importantly, on the valuation and recognition of intellectual property (IP) value contributions. It should, therefore, be expected that the crassest examples of low group effective tax rates will become a thing of the past.
OECD pillar one: The problem with a departure from the arm’s-length principle
However, even if the question of whether digital companies will be taxed more substantially has been addressed, a question remains as to where they will be taxed.
The OECD has tried to address this point via its new pillar one approach, which has been subject to public consultation. This approach stipulates that a certain percentage of super profits (i.e. profit beyond a defined threshold) should be taxable in market jurisdictions.
The proposed formulary approach is a clear deviation from the arm’s-length principle, which serves as the basis for long-standing international transfer pricing principles. As such, it will make transfer pricing more complex and costly, and will erode the international consensus that has done much to prevent double taxation.
In particular, even the pillar one proposal still foresees an allocation of the income remaining after the attribution of Amount A under the OECD pillar one proposal according to the arm’s-length principle. The inevitable contradictions between these two approaches will likely lead to increased controversy. Functions surrounding the development, maintenance, enhancement, protection, and exploitation of intangibles - such as the user base - still need to be remunerated. Since this remuneration ultimately needs to be drawn from the overall profit of the company, it will remain a contentious issue to determine how much profitability stems from various intangible contributions for the user base apportionment.
In view of the upcoming realisation that the pitfalls related to the proposed pillar one solutions are larger than initially expected, the question is, therefore, whether there could still be an economically based alternative to the proposed formulary apportionment.
Local data: Formulary apportionment or economic foundations?
Digital business models often do not require a local presence to conduct business. Whether the local users still give rise to a local nexus for income tax is, to some degree, a normative question that must be answered at a political level. However, there is also the technical question about how value generation of local user contributions can be assessed economically.
Through projects, NERA Economic Consulting has gained insights and developed objective techniques to assess value contribution by users and data, which could be considered in the regulatory debate. A simplified example of how an economic analysis can be technically conducted is presented below. This is not to say that local users or data should be considered a local nexus per se. However, it helps to attribute an arm’s-length profit to such intangibles.
Data-based economic valuation of user contributions
In the simplified example, a technology company that has a user base that uses the digital services of the company while being shown advertisements, is considered. Typically, the company will be largely driven by two factors: user participation and the underlying technology. From an economic point of view, both factors are unique value drivers that generate value jointly and should thus be remunerated via the profit split method. This shifts the question to what percentage of the profits should be attributable to each intangible category.
Economic analysis to determine an appropriate split of a jointly generated profit is rooted in game theory. In our experience, the Shapley value concept can be particularly useful for gaining insights. The concept is based on marginal contributions (i.e. the relative profit that both technology and users can bring to the business). If one party is relatively unimportant to the overall business, it should earn a lower share of the overall profit.
A detailed analysis of the technology and user base components aims to identify the parts that are ‘unique and valuable’ in the sense of the OECD Guidelines. Substantial portions of the technology and user base might turn out to be either relatively easy to replicate or not subject to effective protection (legal or otherwise). The precise definition of what constitutes unique features will depend on the individual case, but, as a guideline, analyses can be conducted regarding which features can be readily copied by competitors and what data is easily available on the one hand, and which features are strictly proprietary to the analysed company on the other hand.
When the unique and valuable aspects of the user base and technology are clearly identified and delineated, it becomes possible to conduct a meaningful marginal profit analysis for four combinations:
‘Normal’ technology and ‘normal’ user base;
Normal technology and comprehensive user base;
Unique technology and normal user base; and
Unique technology and comprehensive user base.
For each of these cases, the profit potential is separately identified through a simulation of the company’s (external) monetisation (i.e. how much revenue could be generated in each combination). Data analytics experts from technology companies regularly conduct such analyses for business reasons. The marginal profit increase (with and without) each unique and valuable element then reflects the relative value contributions for the two categories of intangibles.
A value for local user contributions
Overall, the methods discussed above allow for the identification of the proportion of a digital company’s profits that should be attributed to the user base as opposed to the technology. In practical terms, it is often necessary to consider other intangibles or activities as well, but the basic principle can nevertheless be used.
It follows that a completely formulary apportionment for remunerating the user base is not just a violation of the arm’s-length principle. It is not necessary to quantify local user contributions because economic methods to assess relative value contributions based on the multinational’s own big data exist.
If it is decided that a local user presence gives rise to a taxation nexus, the method above allows companies to determine an economically justified allocation to the overall user base, which can then be allocated to the various tax jurisdictions in which the users are located. Since this is economically justified in the first place, it would remain within the framework of the established arm’s-length principle and reduce the scope for international disputes.
Yves Hervé
T: +49 69 710 447 508
Philip de Homont
T: +49 69 710 447 502
