How could the TMT industry respond to digital tax proposals?

The TMT industry includes a broad range of companies, from digital technology giants to film studios, hardware manufacturers, and software-as-a-service companies. The industry participants, however, share a similar trait in that they generally rely heavily on research and development (R&D) and the use of intangible assets in driving profitability and growth.

BEPS Action 1 considered whether new international tax rules should be written for the so-called, but undefined, 'digital economy'. Interim proposals and unilateral measures (such as the UK's proposed and France's enacted digital services tax) are also generally confined to what might be traditionally considered as the digital space, targeting large technology companies that generate revenues from exploiting user contributions and data.

To avoid a multiplicity of differing unilateral measures by countries, the G20/OECD, working through the Inclusive Framework of more than 130 countries, is seeking multinational consensus on a long-term solution. The work-in-progress long-term solution, however, has a much broader scope and impact than what one might consider to fall under the scope of a digital company.

Jurisdictions, including the US, have steered the digital debate into a broader discussion of how to reallocate taxing rights, generally from resident jurisdictions to market jurisdictions, across all industries. Many TMT companies that are not in the digital space have a physical or non-physical presence in market jurisdictions. More taxing rights may be allocated, potentially under principles other than the arm's-length standard, to these market jurisdictions under the OECD's Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitisation of the Economy, which was released on May 31 2019.

Industries that have high profitability would likely experience the greatest impact from the proposals being considered.

There will still be much discussion among jurisdictions and stakeholders on the contours of these new profit allocation rules. This articles discusses how the proposed approaches outlined in the programme may impact TMT companies, especially in light of some key industry characteristics, such as lack of physical presence, significant start-up investments, fluctuation in profitability among business segments/products, and a higher degree of interdependence among technology and market intangibles.

Two of the three approaches being considered under the programme of work – the modified residual profit split (MRPS) method and the fractional apportionment (FA) method – would be based on the global, regional, or business unit (BU) profit of a multinational enterprise (MNE), as opposed to the local entity profit as determined under the arm's-length standard.

As described in this publication's article, entitled 'G20/OECD work: Maybe way beyond digital', these proposed methods go beyond allocating profits within a supply chain under the arm's-length standard and expand the profit pool by potentially subjecting a multinational's global profits to a new taxing right.

In determining total profits to be split, the programme of work identifies an issue that can be particularly pronounced in the TMT industry:

"The profitability of a MNE group can vary substantially across different business lines and regions… the programme of work will explore the possibility of determining the profits subject to the new taxing right on a business line and/or regional basis."

As with many other sectors, risk is prevalent in the TMT industry. Many MNEs hedge risks by differentiating, which leads to acquisitions and mergers. Furthermore, many acquisitions in the TMT industry consist of a profitable acquirer and a not-yet-profitable target. As a result, one TMT company can have many fundamentally different product offerings, and profits can fluctuate widely across business segments or products.

However, there is a question over how different profitability among business units should be addressed in the MRPS or FA methods. For example, a software licensing company has an established product (Product A) that makes a 50% operating margin across several markets outside its home country (Country X). It has also spent millions to develop Product B in Country X, but Product B is not yet on the market. Should only the excess profits associated with Product A be potentially subject to the new taxing right in the markets in which it is sold, while losses associated with Product B are absorbed in Country X because it does not yet have a market component?

In another example, assume the same software licensing company has an overall operating margin of 2%. If consolidated financials are used in the MRPS, the MNE will likely not have excess profits that are subject to the new taxing right. But if segmented data are used, the excess profits associated with Product A may be subject to the new taxing right.

Finally, if an MNE has many products or segments with varying profitability and geographical footprints, the need to perform multiple MRPS or FA analyses can become time- and resource-consuming. MNEs will need to consider how to collect the data necessary to make the determination, and how the data, once collected and entered into an income tax return, would be audited by tax authorities.

Start-up losses, in the form of initial investments in intangibles, are also common in the TMT industry.

Similar to the situation in the broader economy, 78% of technology-based firms survive their first year. However, according to a November 2017 report by J John Wu and Robert Atkinson from the Innovation and Information Technology Foundation, entitled How Technology-Based Start-Ups Support US Economic Growth, only 41% of technology companies make it through their first five years, which is 6% worse than the overall economy.

Many start-up companies focus on top line and market share growth rather than profitability in their early years, and building a platform to establish market share can be costly. While technology companies only make up 3.8% of all US businesses, they contribute to more than 70% of the business R&D investment to the US economy and provide more than 50% of the R&D jobs. Technology companies include TMT industries such as semiconductors, computers, software publishers, computer system design, and data hosting, but also include life science and aerospace products.

In particular, two industries within the TMT sector, the semiconductor machinery industry and the semiconductor and other electronic components industry, exhibit a disproportionate amount of R&D intensity, with R&D investment to revenue ratios of 28.4% and 18.5%, respectively. In comparison, and again according to Wu and Atkinson's research, companies in the broader economy only invest 3.3% of their revenues in R&D.

In addition to R&D investments, TMT companies also incur heavy regulatory costs in each market, such as obtaining and maintaining a licence to do business.

The OECD's programme suggests several potential ways to address losses, such as splitting losses symmetrically with profits, using an earn out approach, whereby profits are subject to the taxing right only when historical losses have been earned out, or using a hybrid of both.

The programme, however, does not specify under which set of facts and circumstances would one method be more appropriate over the other. It may, therefore, be tempting to split losses in the early years; but be aware that sharing losses means sharing the potential upside as well in future years.

Furthermore, if losses are not cleanly segmented between those that are aimed at developing traditional intangibles (for example, technology IP) versus those subject to the new taxing right (market IP), the splitting of losses in early years may give market jurisdictions a claim to excess profits associated with technology intangibles when a business becomes profitable.

There can also be difficulties in defining losses from an accounting standpoint. Will losses include interest expense and stock compensation deductions? Which jurisdiction's accounting or tax rules should be used? For how many years can losses be carried forward? How can one assign losses to individual business lines? What kind of record keeping is needed to keep track of historical losses under the earn-out approach?

Under the MRPS approach, it is expected that the programme will consider rules to quantify profits associated with market IP.

As it stands, the programme suggests first carving out a routine return and then splitting non-routine returns between traditional intangibles and market IP using proxies such as capitalised expenditures, projected future income, or fixed percentages. With a brick-and-mortar company, one may attribute R&D costs to technology intangibles and marketing costs to market IP. In the TMT industry, however, value is often derived from a single platform that can attract a significant number of users to create barriers to entry. Technology and users have little value without the other. Therefore, it is harder to find a proxy that can estimate the distinct effect of market vs non-market intangibles on profits.

For example, Company A operates a social media platform. Its R&D team has developed a platform that not only efficiently matches users with advertising providers but also has a top-notch interface that attracts millions of users worldwide. Should some R&D costs be attributed to market IP, even though the costs are not incurred in the market jurisdiction? Similarly, for a loss-making company, does investing in technology to attract more users count as developing market IP, and should some or all of the losses therefore be reallocated to the market jurisdictions?

The complexity of implementing rules that separate market IP from other intangibles could lead policymakers to deviate from the arm's-length standard and instead adopt a more formulaic way of calculating market IP as a fixed percentage of a multinational's excess returns, including the use of safe harbours. If a formulaic approach is adopted, should there be exceptions for MNEs or segments that do not typically rely heavily on market IP, such as business-to-business companies?

The business unit conundrum and the investment problem also apply in a formulaic world: should routine returns, returns associated with traditional IP, and returns associated with market IP be a fixed percentage (for example, percentage of sales, percentage of fixed assets, percentage of total non-routine returns), or should it vary among business units and regions? Furthermore, if the OECD decides that a routine return is to be 10% of sales, for example, and a company has a negative 5% operating margin, should the non-routine return, which is further split between traditional intangibles and market IP, be -15% or -10%?

The MRPS and FA methods may use proxies or factors to apportion profits subject to the new taxing right among jurisdictions. For companies in the TMT industry, the most common allocation keys will likely be revenues or users. For companies whose gross revenues, net revenues, and users are not perfectly correlated (for example, those with different pricing models among countries, or higher rebates in certain countries), it will be important to determine the most appropriate proxy to measure value contributed by the market jurisdiction.

Start-up companies that want to split losses but do not have significant users or revenues also need to search for other reasonable allocation metrics.

It is also unclear whether businesses will be able to choose the most appropriate allocation key. If a choice is given, businesses will likely need to prepare documentation to defend their selection of what is most appropriate in the event of audits.

The OECD's programme will consider a baseline return in market jurisdictions "for marketing, distribution and user-related activities". Fluctuating profitability among business segments should be taken into account when determining baseline profitability.

The format of distribution can also be so different among industry players that a flat percentage of sales may not be applicable to all. For example, content owners often use streaming service providers to distribute content. Would the market jurisdictions still subject content owners to a baseline return if distribution profits are already taxed via the third-party distributors? Another question is whether the baseline distribution return should be the same for a widget seller and a software licensor.

This work stream will also impact non-digital companies, such as semiconductor manufacturers, because it can result in "the allocation of a higher return under traditional TP principles to market jurisdictions".

Any global consensus on the digital initiative would require changes to domestic laws and treaties. One can also speculate that the initiative will result in the modification of the 2017 OECD transfer pricing guidelines (TPG).

Chapter I of the TPG assigns excess returns to entities that control and bear the risks associated with such returns. In TMT companies, the risks associated with the development, enhancement, maintenance, protection and exploitation (DEMPE) of market-type intangibles are generally not controlled at the market jurisdiction. Will Chapter I of the TPG be revised to reflect the new paradigm of profit allocation, and will this impact other TPG chapters?

This article highlights characteristics in the TMT industry that should be given due consideration as the OECD Inclusive Framework aims to achieve broad political agreement in the near future and as countries implement any agreement from a technical perspective.

At the time of writing, no formal agreement had been reached on how profits associated with market IP should be calculated. Many questions, similar to those raised in this article, still need to be addressed, and the suitable answer to these questions will likely vary across industries. Businesses, as industry experts, can take part in shaping the outcome by engaging in discussion with government officials in their jurisdiction and by reviewing – and when appropriate, commenting – on OECD/Inclusive Framework discussion drafts.

Sajeev Sidher

Partner Deloitte Tax LLP T: +1 415 401 5806 ssidher@deloitte.com Sajeev Sidher joined Deloitte in 1998 and has two decades of TP and tax experience. Sajeev is a tax principal and leader of Deloitte Tax LLP's Silicon Valley practice, specialising in tax and transfer pricing. He works across a wide range of industries including: technology; durable products; chemicals; engineering and construction; transportation; and the multilevel marketing business. He is also the Americas TP industry leader for technology and specialises in assisting companies operating in the semiconductor, memory, hardware, gaming, e-commerce, and medical products industries. Sajeev provides global TP planning and compliance services with a focus on post-acquisition integration, supply chain restructuring and intangible property optimisation. Sajeev holds a bachelor's of science degree in business administration with a major in finance from San Francisco State University; a JD (cum laude) from Albany Law School of Union University; and an LLM in taxation from Georgetown Law Center. He is a member of the New York and Connecticut state bars.

Kaidi Liu

Senior manager Deloitte Tax LLP T: +1 202 220 2734 kaliu@deloitte.com Kaidi Liu is a senior manager in Deloitte Tax LLP's national TP practice. Kaidi has more than 10 years of experience in transfer pricing, working in Washington DC, Chicago, Shanghai, and Kolkata. She specialises in strategic BEPS TP assessments, focusing on topics such as country-by-country reports, substance in light of the OECD risk-control framework, and impact of the digital discussion. Kaidi also has extensive experience in intangible asset valuation and transfer, business model optimisation, business restructuring planning, advance pricing agreements (APA), and global documentation. Kaidi has assisted clients across a variety of industries: consumer products; digital services; financial services; food and beverage; industrial products; and medical device. As a member of Deloitte's Washington national tax team, Kaidi has developed various resources for Deloitte's global TP practice, including sample documentation reports, BEPS implementation matrices, and country-by-country report FAQs and review guides. Kaidi is pursuing a master's in management from the London School of Economics and Political Science. She holds a BA in economics and biology from Northwestern University. Kaidi is fluent in English and Mandarin.

Copyright © 2019 Deloitte Development LLC. All rights reserved.

This communication contains general information only, and none of Deloitte Touche Tohmatsu Limited, its member firms or their related entities (collectively, the ‘Deloitte network’) is, by means of this communication, rendering professional advice or services. Before making any decision or taking any action that may affect your finances or your business, you should consult a qualified professional adviser. No entity in the Deloitte network shall be responsible for any loss whatsoever sustained by any person who relies on this communication.