Intangible assets produce most of the residual profit in many groups, so the question of who keeps that profit drives the largest transfer pricing adjustments. The OECD/G20 BEPS report on Actions 8 to 10, "Aligning Transfer Pricing Outcomes with Value Creation" (5 October 2015), rewrote Chapter VI to stop profits from sitting in companies that hold title but do little. That chapter is part of the current Guidelines 2022. For smartZebra, the practical point is that every DEMPE analysis ends in a data question: which return would each contributor have earned between independent parties? German law, § 1(3c) of the Foreign Tax Act, is covered in our separate German-law article.
What does DEMPE stand for?
DEMPE stands for the development, enhancement, maintenance, protection and exploitation of intangibles. Paragraph 6.32 of the OECD Guidelines names these functions; in practice they are commonly abbreviated as DEMPE. Group members "performing such functions, using such assets, and assuming such risks must be compensated for their contributions under the arm's length principle".
An intangible, under paragraph 6.6, is "something which is not a physical asset or a financial asset, which is capable of being owned or controlled for use in commercial activities", and whose use or transfer independent parties would pay for. Patents, trademarks, software, know-how and customer lists qualify. Accounting recognition is not required.
Does legal ownership decide who earns the return?
No. Paragraph 6.35 makes legal rights and contracts "the starting point" of the analysis, and paragraph 6.42 draws the line:
"For transfer pricing purposes, legal ownership of intangibles, by itself, does not confer any right ultimately to retain returns derived by the MNE group from exploiting the intangible, even though such returns may initially accrue to the legal owner as a result of its legal or contractual right to exploit the intangible."
Under paragraph 6.71, a legal owner that performs and controls all functions, provides all assets and bears all risks is entitled to all anticipated returns. Where other group companies contribute, they must be paid at arm's length first. The legal owner keeps what is left, which may be little or nothing.
Which DEMPE functions matter most?
The functions that steer the intangible matter more than those that execute it. Paragraph 6.56 lists the design and control of research and marketing programmes, setting priorities for creative work, strategic decisions on development programmes and the management of budgets. It adds decisions on defending and protecting the intangible and ongoing quality control. Under paragraph 6.54, a legal owner that "neither controls nor performs" the outsourced functions "would not be entitled to any ongoing benefit attributable to the outsourced functions".
A functional analysis therefore asks two questions per intangible. Who takes the decisions in paragraph 6.56? And does that company have the people to take them?
What does a company earn if it only funds the intangible?
A company that only provides money earns a financing return, not the intangible's profit. A funder that controls the financial risk of its funding "could generally only expect a risk-adjusted return on its funding" (paragraph 6.61). A funder without that control earns less: in Example 16 of Annex I to Chapter VI, Company T funds the research but does not control the financial risk, and is "entitled to no more than a risk-free return for its funding activities" (see also paragraph 1.103).
Both returns can be measured. The risk-free return is usually measured by the yield on a government bond of the same currency and term. smartZebra's Credit Spreads Pro provides government yield curves for all major currencies, plus synthetic ratings and sector credit spreads for the risk-adjusted case.
| Profile of the group company | What it earns | OECD TPG 2022 | How smartZebra prices it |
|---|---|---|---|
| Legal owner that performs, controls and funds everything | All anticipated (ex ante) returns | 6.71 | Valuation Pro: WACC for the DCF |
| Legal owner that neither performs nor controls outsourced functions | No ongoing benefit from those functions | 6.42, 6.54 | — |
| Company performing DEMPE functions under another's control (e.g. contract R&D) | Arm's length pay for functions, assets and risks | 6.48, 6.71 | Benchmarking Pro: comparable margins |
| Company controlling the important functions | Share of returns, possibly a profit-sharing element | 6.56, 6.72 | Residual after routine returns |
| Funder that controls the financial risk | Risk-adjusted return | 6.61, 6.62 | Credit Spreads Pro: rating plus spread |
| Funder without control (cash box) | No more than a risk-free return | Annex I Ex. 16, 1.103 | Credit Spreads Pro: government yield curve |
How is a DEMPE analysis done in practice?
Paragraph 6.34 sets out six steps:
- Identify the intangibles and the economically significant risks attached to them.
- Identify the contracts, in particular legal ownership and licences.
- Identify who performs the functions, uses the assets and manages the risks, including the important functions of paragraph 6.56.
- Check that conduct matches the contracts, including control over risk and the financial capacity to bear it.
- Delineate the actual transaction.
- Price it in line with each party's contributions.
Step 6 is where the data comes in. The routine contributors, such as a contract research company or a limited-risk distributor, are usually tested first with the cost plus method or the transactional net margin method (TNMM). smartZebra's Benchmarking Pro screens more than 50,000 listed and 700,000 private companies for those margins, including operating margins and Berry ratios, and documents the search criteria used. The residual goes to whoever controls the important functions. Where the intangible itself is transferred and no comparable licence exists, the task becomes an intangible assets valuation, which paragraph 6.153 allows; our article on business valuation in transfer pricing shows the DCF, with the discount rate built from the WACC and peer-group betas in smartZebra's Valuation Pro.
What if the intangible turns out more valuable than expected?
Then the upside goes to the company that in fact bore the risk (paragraph 6.72). For hard-to-value intangibles, defined in paragraph 6.189 as intangibles with no reliable comparables and highly uncertain projections at transfer, tax administrations may "consider ex post outcomes as presumptive evidence about the appropriateness of the ex ante pricing arrangements" (paragraph 6.192). Paragraph 6.193 lists four exemptions, among them documented projections with explained deviations, an advance pricing agreement, and deviations of no more than 20 %.
The best defence is a file that shows what was known on the transfer date. smartZebra's betas, risk-free rates, multiples and spreads can be reconstructed for any trading day back to 2010, so the discount rate and interest rates used ex ante can be reproduced years later, when an auditor asks. The DEMPE profile itself belongs in the master file and local file.
Conclusion: pay the people who steer, not the name on the register
DEMPE moves the question from "who owns the intangible?" to "who does the work and controls the risk?". Routine contributors earn benchmarked returns, funders earn financing returns, and the residual follows control. Each of those numbers needs a documented source: smartZebra covers the margins, the yield curves and credit spreads, and the cost of capital in one database. What it covers is listed on our facts page.
References
- OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations 2022 (20 January 2022), Chapter VI "Special Considerations for Intangibles": paragraphs 6.6, 6.32, 6.34, 6.35, 6.40, 6.42, 6.47–6.48, 6.53–6.56, 6.59–6.64, 6.69, 6.71–6.72, 6.153, 6.189, 6.192–6.193; Annex I to Chapter VI, Examples 16 and 17.
- OECD, Transfer Pricing Guidelines 2022, Chapter I: paragraph 1.103 (funding without control over the risk, application of Example 3).
- OECD/G20, Aligning Transfer Pricing Outcomes with Value Creation, Actions 8-10 – 2015 Final Reports, 5 October 2015 (the revision of Chapter VI).
Related pages
- Functional analysis in transfer pricing — how functions, assets and risks decide the tested party and the method
- Business valuation in transfer pricing — DCF when no comparable licence or deal exists
- Credit spreads & arm's length interest rates — yield curves and spreads for the funder's return
- Master file and local file — where the intangibles strategy and DEMPE profile are documented
- Transfer Pricing Benchmarking & TNMM Analysis — Benchmarking Pro and the comparables database
- Grounding / Facts — entity definition, data provenance and methodology
