Embedded IP

Embedded IP is intellectual property that contributes to the value of a tangible asset, financial asset or service, or personal service. Its value is charged in a bundled price along with that of the property in which it is embedded. Embedded IP is therefore not subject to a separate charge like in the case of an IP licence.

For transfer pricing (TP) purposes, we should look to the aggregation rules in the US Treasury regulations (and Inland Revenue Code (IRC) Section 482 after the US 2017 Tax Act), and the OECD TP guidelines for multinational enterprises and tax administrations (OECD TPG). Reference to these, combined with an application of the best method rule (US) or most appropriate method (OECD) will most often result in a conclusion that the party buying – or licensing a 'make and sell' licence – from a related party property with embedded IP and reselling such property is only entitled to a routine return for its routine buy-sell activities. This conclusion eliminates the need to analyse the value of the IP separately from the value of the property that embeds the IP, as one-sided methods such as the transactional net margin method (TNMM)/comparable profits method (CPM) allow benchmarking the arm's-length return for the buy-sell functions. Any residual value is related to the embedded IP.

Example 1: a manufacturer sells a TV set embedding various IP in the form of technology and trade marks/trade name to related and unrelated distributors performing comparable functions, using comparable assets, and assuming comparable risks. For TP purposes, both the manufacturer and related distributor must support the arm's-length nature of the price used between them. The selling price used for sales to unrelated distributors can be used as an arm's-length measure of the appropriate selling price for sales to the related distributors. Under this one-sided TP comparable uncontrolled price (CUP) method, there is no need to attempt to determine the arm's-length value of the embedded IP.

Example 2: a manufacturer sells a TV set embedding various IP to a related distributor only. For TP purposes, again, both the manufacturer and its related distributor must support the arm's-length nature of the consideration paid and received. In the absence of internal comparable uncontrolled price transactions – the price paid by unrelated comparable distributors in the previous example – the arm's-length consideration could be determined by reference to the profit generated by independent TV set distributors performing broadly comparable functions, using broadly comparable assets, and assuming broadly comparable risks as the related distributor. Note that it is irrelevant whether the TV sets distributed by the uncontrolled comparable companies embed IP of the same value as that embedded in the TV sets subject to the controlled transaction because the value of the distribution functions performed by the related distributor and comparable companies is the same (by assumption, in this example). Therefore, here again, under this one-sided transfer pricing method (TNMM/CPM), there is no need to attempt to determine the arm's-length value of the embedded IP.

In many cases, however, the manufacturer owning the IP may be selling finished TV sets to unrelated distributors, and licensing at the same time the make and sell rights to the IP to related manufacturers. In such cases, can the uncontrolled transaction be reliably used to determine the value of the controlled IP licence? The controlled and uncontrolled transactions are very different in that the controlled transaction is the licensing of make and sell rights to IP, while the uncontrolled transaction is a tangible property transaction that happens to embed the IP subject to the controlled licence. The extent and nature of the rights transacted in both cases are also different. In the controlled transaction, the rights transacted are make and sell rights to IP (no right to further develop the IP is transacted); in the uncontrolled transaction, the rights to sell a product are transacted (no right to make a product using the IP or right to further develop the IP are transacted).

A similar question can be examined when the property transacted is not a tangible product but is a financial or personal service that embeds certain IP. In such cases, the uncontrolled transaction may be the provision of a service with embedded IP to an unrelated party while the controlled transaction may be the licensing of the IP to a related party allowing that related party to provide comparable services as provided in the uncontrolled transaction to other related parties.

Given the very broad definition of intangible property and assets in US law and in the OECD TPG, embedded IP is likely to be found in a number of controlled tangible and services transactions. In these controlled transactions, consideration paid reflects the return to many economically substantial risks arising from assets used and functions performed by the participants in the controlled transactions. The use of assets and the performance of functions obligates the participants to incur costs such as material cost, labour costs (salaries and benefits), depreciation of fixed assets and amortisation of intangible assets (in some cases), interest on debt, research and development (R&D) expenses and expenses associated with the development of other intangible assets, and taxes. For this discussion, we will assume the existence of embedded IP in an otherwise tangible or services transaction. Whether or not a controlled transaction concerns embedded IP is a factual issue discovered during the functional analysis and the process of accurately delineating a controlled transaction. The exact nature of the embedded IP is not of particular relevance to this discussion as we focus our discussion specifically on conceptual considerations. Note, however, that insofar as the embedded IP is concerned, the economically substantial risks arise out of the performance of the so-called DEMPE functions (development, enhancement, maintenance, protection and exploitation). We will come back to DEMPE shortly. For the time being, we will assume that the same legal entity performs and controls all DEMPE functions – there is no fragmentation of DEMPE. The word 'control' is used within its meaning in Chapter I of the OECD TPG.

The value of IP is reflected in the profit-over-costs element captured by the controlled distributor redeemed to the manufacturer as the return to the risks it faced as a result of the costs it incurred. Different IP will have different market values and will require different levels of continuous costs in the form of maintenance and enhancement. Intellectual property may create barriers to entry for competitors. How long that barrier to entry is effective varies depending on the nature of the IP. Marketing intangibles for example are often believed to require more intensive and continuous maintenance than technology IP that is legally protected by patent. The value of IP therefore depends on: (i) the level of costs required to develop it and the systemic risk associated with such costs commitments; (ii) the level and frequency of maintenance and enhancement costs, and associated systemic risk, required to support the continued expectation of probability – weighted cash inflows reflecting a premium over marginal cost; (iii) the level of IP development competition in the market; and (iv) the availability of legal protection of such expectation of cash inflows. Thus, the level and duration of expected cash inflows, probability of success and failure, and legal barriers to entry are critical determinants of IP value.

The assessment of the value of embedded IP can be performed in a reliable way by reference to well-accepted valuation techniques found in the asset pricing and corporate finance literature. Such methods have authority under US Treasury regulations and the OECD TPG. In the corporate finance world, these approaches can be classified in three groups, each possibly containing different variants:

• The market approach seeks to evaluate the value of IP by reference to the price paid for comparable IP exchanged between unrelated parties. If that approach is appealing for its simplicity, its usefulness is limited by the requirement that the IP must be sufficiently comparable to provide a meaningful data point. Since IP is often unique – and hard-to-value-intangibles that have received enormous attention in the BEPS discussions at the OECD are by definition unique and valuable – it is often unreliable to adopt a market approach (only) to value IP, with the exception of cases where an IP owner would be licensing the IP, under similar circumstances, to both related and unrelated parties;

• The cost approach seeks to evaluate the value of IP by reference to the costs incurred in developing the IP. This cost approach may be reliable to value IP at the very early stages of development, but for more developed IP it is generally unreliable unless the IP is not unique or particularly valuable. Investments in IP are typically risky (especially for technology IP that carries technical risk), which means that typically no stable relationship exists between the costs of developing IP and the value of the IP when developed; and

• The income approach seeks to evaluate the value of IP by reference to the expected discounted net cash flows the IP is projected to produce. Given the type of IP subject to controlled transactions, this approach is often the most reliable. This is reflected in the guidance provided in US Treasury regulations and by the OECD that focuses on discounted cash flow (DCF) valuation to give effect to an income approach. The main challenge in using a DCF approach is the development of probability-weighted financial projections. A financial projection is different from a forecast. A financial projection is the probability-weighted average of all possible forecasts a company can envision for every year for which there is a reasonable expectation of cash flows associated with the subject IP.

One of the merits of the income approach is that it is relatively simple to implement. In addition, income approaches have a long history in other contexts (e.g. purchase price allocation valuation, and M&A), including non-tax contexts, are well understood and accepted, and are often the only reasonable valuation option available. The OECD TPG published in 2017 as a result of the OECD BEPS work recognised the usefulness of the income approach by providing new guidance on its use for TP purposes (see Chapter VI). The market approach is often analogised to the CUP method, and the cost approach is often analogised to the cost plus method. The income approach can be associated with the TNMM/CPM approach in that they all assess profitability of a tested party at the (net) profit level.

In summary, in the case where a manufacturer sells products with imbedded IP to a related party, one can evaluate the reliability of the derivation of the value of that IP obtained using the following three categories of methods:

• Market approach: the IP owner is licensing comparable IP separately to a third-party manufacturer, or comparable IP is licensed under comparable circumstances between two unrelated parties. Databases exist that may provide such market evidence, if reliable comparable IP transactions can be found;

• Cost approach: the IP owner capitalises all costs incurred in developing the IP using a discount rate reflecting its expected return. That return can be reliably estimated using financial metrics such as cost of debt and cost of equity (weighted average cost of capital); and

• Income approach: financial projections (probability-weighted) are developed and the value of the IP is calculated as the difference between the total net present value of the combined transactions and the net present value of the non-IP related expected cash flows (using CUP or TNMM/CPM).

Being able to identify and quantify embedded IP is of great importance to a taxpayer qualifying for IP regimes. Indeed, in the application of IP regimes, the taxpayer assumed to be performing R&D activities may have to be able to identify embedded IP income that will serve as a basis for the application of the regime.

It may be worthwhile considering, at this stage, if the embedded IP income measured covers a R&D function only or a broader set of functions associated to that IP, like a sales function of the proceeds of the R&D function. If such is the case, one can wonder if the full IP income needs to be considered as a basis for the application of the regime, or only the portion attributable to the 'pure' R&D function.

Embedded IP is everywhere. Its valuation – benchmarking – is not straightforward and resorts to approaches that, also under the OECD regulations, are akin to corporate finance approaches.

Although not always necessary, in quite a few cases, separate benchmarking of embedded IPs is necessary for TP purposes.

That same separate benchmarking is likewise of great importance in the application of national IP regimes allowing deductions on the basis of the IP income.

André Schaffers

Partner, transfer pricing Deloitte, Belgium Tel: +32 2 600 67 15 Mobile: +32 497 51 52 31 aschaffers@deloitte.be André Schaffers is a partner in the transfer pricing practice of Deloitte Touche Tohmatsu Limited's Belgium member firm. He is in charge of the economics practice and coordinates Deloitte's global innovation group. After working in the financial industry, André Schaffers joined Deloitte's transfer pricing group in 1998. He has worked on various engagements, including setting up of global transfer pricing policies, advance pricing agreements, audit defence of existing or new transactions, documentation, and business alignment projects. In addition to his role in Deloitte's Belgian firm, André has had various EMEA and global roles within Deloitte's transfer pricing practice. Representative experience André's projects include the analysis of all types of transactions: tangible goods and services, with a special emphasis on financing and intangible property. His projects cover a variety of industries, including the automotive, pharmaceuticals, chemicals, construction goods, consumer goods, electronics, energy, financial services, industrial equipment, medical equipment, mining, and retail sectors. André is Deloitte Belgium's primary contact for transfer pricing analysis of intercompany financial transactions including pricing of debt, cash pooling arrangements, and other complex financial instruments. In recent years, André has focused increasingly on the link between the Belgian patent/innovation income deduction regimes and transfer pricing. Indeed, the application of these regimes – primarily designed to create tax incentives for companies with significant research and development activities – requires the determination of an arm's-length return on self-generated intangible property; thus, OECD-inspired economic models and analyses are needed. André is in frequent contact with the tax authorities to discuss issues involving financial transactions, intangibles, the innovation income deduction, and transfer pricing documentation in general. He is a frequent speaker at external and Deloitte-sponsored events and seminars. Education 1991: Master in Applied Economics, ICHEC (Brussels) 1997: Master in Business Administration, Solvay Business School, University of Brussels

Filip Vanluydt

Director, global business tax, Belgium Deloitte Touche Tohmatsu Tel: +32 2 600 65 63 Mobile: +32 476 53 00 48 fvanluydt@deloitte.com Filip Vanluydt is a director in the transfer pricing practice of Deloitte Touche Tohmatsu's Belgian member firm. Experience Filip Vanluydt is an economist, specialising in the management of a wide range of transfer pricing projects. He has completed several local, regional (European), and global transfer pricing documentation studies and supply chain restructuring projects, involving, for example, the establishment of shared services and central procurement companies, the migration of intellectual property and the design and implementation of central entrepreneur structures. Filip has also been involved in a number of audit defence projects, competent authority, and unilateral and bilateral advance pricing agreements involving negotiations with revenue authorities in Europe and Japan. Filip has worked with many clients from a wide range of industries, including the manufacturing, technology, telecommunications, energy and resources, pharmaceuticals, chemicals, retail and services sectors. He has a master's degree in commercial enginerring from the University of Hasselt and a Master in Tax Law degree from Katholieke Universiteit Leuven. Filip speak Dutch and is fluent in English and French.