Loading...
Loading...
Borys Ulanenko
CEO of ArmsLength AI
![TP Documentation for Intangibles [2025]: DEMPE Analysis and Beyond](/_next/image?url=%2Fimages%2Fblog%2Fdocumentation-for-intangibles%2Fdocumentation-for-intangibles.jpg&w=3840&q=75)
Get the latest transfer pricing insights, AI benchmarking tips, and industry updates delivered straight to your inbox.
Strong intangibles documentation, consistent with OECD Chapter VI (intangibles) and Chapter V (documentation), typically includes: (1) all relevant intangibles—patents, trademarks, know-how, customer lists, and other IP, (2) the legal owner and DEMPE function performers, (3) arm's length compensation for each contributor based on functions, assets, and risks, and (4) valuation support at key transaction points. The DEMPE framework (Development, Enhancement, Maintenance, Protection, Exploitation) is central: it determines who is entitled to returns from intangibles based on substance, not just legal title.
Legal ownership is the starting point, not the conclusion. A parent company holding a patent registration but neither controlling nor performing DEMPE functions cannot claim all residual returns—the entities actually developing and exploiting the IP must receive arm's length compensation. Outsourcing DEMPE activities doesn't automatically strip returns from the legal owner, but entitlement depends on what it actually controls, performs, and assumes. Document the functional reality, not just the contractual form.
The OECD defines an intangible broadly as anything that is:
This definition focuses on economic value, not accounting labels. Many intangibles that create real value—self-developed technology, internal know-how, established customer relationships—never appear on a balance sheet but are crucial for transfer pricing purposes.
Key Distinction: An intangible may exist for transfer pricing even if it's not capitalized as an asset under accounting standards. The OECD cautions that TP intangibles analysis should not depend on accounting treatment or legal registration status.
| Category | Examples | TP Relevance |
|---|---|---|
| Trade Intangibles | Patents, trade secrets, know-how, proprietary processes | Core to manufacturing and technology licensing |
| Marketing Intangibles | Trademarks, brand names, logos, trade names | Central to distribution and licensing arrangements |
| Customer Intangibles | Customer lists, relationships, proprietary data | Often overlooked but valuable in acquisitions |
| Contract Rights | Exclusive supplier agreements, government licenses, franchise rights | Value derived from contractual exclusivity |
| Software | Internally developed applications, algorithms | Can constitute an intangible (copyright/know-how) depending on rights and use |
| Goodwill | Residual business value above identifiable assets | May need to be compensated when transferring businesses, depending on accurate delineation |
Intangibles rarely operate in isolation. A pharmaceutical company's value comes from the combination of patented compounds, regulatory approvals, manufacturing know-how, established distribution networks, and brand reputation. When valuing intangibles for transfer pricing, consider:
Independent parties pay for expected value, so documentation should acknowledge these combined effects—especially in business restructurings where the whole may exceed the sum of parts.
Important Distinction: The OECD treats group synergies and assembled workforce primarily as comparability factors affecting value, rather than "intangibles" in themselves (since they're not owned/controlled in the same way). However, they can significantly impact the pricing of transferred intangibles and should be addressed in valuation documentation.
Legal title to an intangible is not determinative of profit entitlement. Under OECD Chapter VI, entitlement to returns depends on the functions performed, assets used, and risks assumed (including control of those risks)—not just the name on the patent certificate.
The Guidelines explicitly state that identifying legal owners of intangibles "is only the first step" in determining which group member should receive returns. Even if one entity legally owns IP, other members may have performed the work that created its value.
| Scenario | Legal Owner | DEMPE Contributor | Result |
|---|---|---|---|
| Parent registers patent but subsidiary funds and directs R&D | Parent | Subsidiary | Subsidiary should receive returns commensurate with its contribution |
| Subsidiary develops brand locally but HQ holds trademark | HQ | Subsidiary | Subsidiary's marketing investment must be recognized |
| Holding company licenses IP but neither controls nor performs DEMPE | Holding company | Operating companies | Holding company not entitled to residual returns—limited to compensation for any functions actually performed |
A common structure involves one entity funding IP development while another performs the work. The OECD addresses this directly:
Documentation Risk: Assuming the legal owner captures all income without analyzing who actually performed and controlled DEMPE functions is one of the most common—and dangerous—documentation gaps.
DEMPE stands for Development, Enhancement, Maintenance, Protection, and Exploitation—the five functional categories used to analyze who creates value in an intangible. For each function, determine which entity:
What it covers: R&D, invention, initial creation of the intangible.
Typical supporting evidence:
Key question: Who conceived the intangible and executed its creation?
What it covers: Post-launch improvements, expansions, version upgrades.
Typical supporting evidence:
Key question: Who continually improves the intangible's value?
What it covers: Keeping the intangible operational and current.
Typical supporting evidence:
Key question: Who ensures the intangible remains useful?
What it covers: Legal and security measures to safeguard the intangible.
Typical supporting evidence:
Key question: Who defends and legally protects the intangible?
What it covers: Commercializing the intangible to generate returns.
Typical supporting evidence:
Key question: Who uses the intangible to generate revenue?
Practical Approach: Create a DEMPE matrix for each significant intangible—rows for each function, columns for group entities. Mark who performs each function, who controls it, and note supporting evidence (documents, personnel, costs).
Master File Requirements (Chapter V): The Master File specifically requires documentation of the group's strategy for development, ownership, and exploitation of intangibles; a list of important intangibles and their legal owners; and a summary of key agreements related to intangibles (including CCAs, license agreements, and principal research service arrangements).
For each DEMPE function, documentation should include:
Name the affiliate performing or controlling the function. This must match organizational reality, not just contractual assignments.
| Function | Personnel Examples | Supporting Documentation |
|---|---|---|
| Development | R&D scientists, engineers | Org charts, job descriptions, headcount by location |
| Enhancement | Product managers, developers | Project team lists, work assignments |
| Maintenance | Support staff, quality teams | Service level agreements, activity logs |
| Protection | IP counsel, patent agents | Legal team assignments, filing records |
| Exploitation | Sales teams, marketing staff | Revenue by entity, marketing budgets |
Document who makes decisions, not just who executes them:
Control is assessed using the OECD's risk control framework: does the entity have (1) the capability to make decisions about the activity, (2) actual decision-making authority exercised in practice, and (3) the financial capacity to bear the consequences if risks materialize?
Identify tangible and intangible assets deployed:
Document which entity bears each risk:
| Risk Type | What It Covers | Documentation |
|---|---|---|
| Development risk | R&D failure, delays | Project budgets, milestone contracts |
| Market risk | Commercial failure | Revenue forecasts, market studies |
| IP risk | Invalidity, infringement | Litigation reserves, insurance |
| Regulatory risk | Approval failures | Compliance documentation |
The OECD states that CUP and profit-split methods are "most likely to prove useful" for valuing significant intangible transfers. However, this is not a mandatory hierarchy—method selection remains fact-driven based on availability of comparables and transaction characteristics.
| Method | When to Use | Documentation Requirements |
|---|---|---|
| CUP | Comparable license or sale transactions exist | Comparable agreements, adjustment rationale |
| Profit Split | Unique intangibles, multiple DEMPE contributors | Contribution analysis, profit allocation basis |
When market comparables are scarce, income approaches are common:
Discounted Cash Flow (DCF):
Relief from Royalty:
The OECD explicitly states that "there rarely is any correlation between development cost and value." Cost-based methods should only be used:
Common Mistake: Using R&D expenditure as a proxy for intangible value. Breakthrough patents can cost millions to develop and be worth billions—or fail entirely. Document why a cost approach is appropriate if used.
Regardless of method, documentation must include:
An HTVI exists when, at transaction time:
Examples include:
The OECD permits tax authorities to use hindsight—looking at actual outcomes to assess whether the original price was arm's length. However, taxpayers can protect themselves with thorough ex-ante documentation:
What to Document:
| Element | Purpose |
|---|---|
| Detailed projections | Financial models, probability-weighted scenarios |
| Risk analysis | Technical, regulatory, commercial risks identified and priced |
| Assumption rationale | Why growth rates, success probabilities, discount rates were chosen |
| Comparability limitations | Document why no reliable comparables exist |
| Price adjustment clauses | Milestone payments or renegotiation triggers, if any |
Without proper documentation, tax authorities may treat actual outcomes as "presumptive evidence" of arm's length pricing. If a technology sold for $10 million ultimately generates $500 million in profit, the authority may adjust the price—unless the taxpayer can show:
The OECD HTVI guidance also sets boundaries on when tax administrations should not use ex post outcomes. Key exemptions include:
Taxpayers can rebut hindsight presumptions by demonstrating that the original pricing reflected reasonable expectations given information available at the time.
Protection Strategy: If you're transferring an HTVI, preserve all financial models, board presentations, and risk assessments used to set the price. These documents are your defense against hindsight adjustments—and may exempt you from ex post scrutiny if outcomes fall within documented ranges.
A Cost Contribution Arrangement is a contract among affiliates to share costs and risks of developing or acquiring intangibles (or other assets), expecting proportionate future benefits.
CCAs are common for:
For a CCA to be respected, it must satisfy several conditions:
| Requirement | What It Means |
|---|---|
| Mutual benefit | Only parties expecting proportionate benefits should participate |
| Defined interests | Each participant's share of future IP/benefits must be specified |
| No extra payments | Participants shouldn't pay separate royalties post-development—only contributions, balancing payments, and buy-ins |
| Arm's length contributions | Each party's contribution must be valued at arm's length |
A properly documented CCA includes:
Agreement Terms:
Contributions:
Benefits and Allocation:
Exit Provisions:
Balancing Payments:
Best Practice: While not strictly mandated, informal R&D cost-sharing—by email or implicit understanding—creates significant audit risk. Tax authorities strongly expect written CCAs with clear terms, consistent with what independent parties would document in a comparable arrangement.
IP transfers within an MNE require documentation covering:
Restructurings may require arm's length compensation where something of value is transferred and/or where an independent party would have been compensated for termination or renegotiation detriments. There is no automatic presumption that every restructuring triggers an indemnity—it depends on accurate delineation and what independents would have agreed.
Where compensation is appropriate, it may include:
| Element | What to Include |
|---|---|
| Rationale document | Board resolution, strategic plan explaining the move |
| Pre-transfer FAR | Functions, assets, risks before restructuring |
| Transfer agreement | Terms, date, rights conveyed |
| Valuation report | DCF, comparable transactions, or other methodology |
| Post-transfer FAR | New operating model, ongoing arrangements |
Cross-reference: See Documentation for Restructurings for comprehensive restructuring documentation guidance.
These gaps frequently arise during audits and can significantly weaken otherwise solid documentation.
Problem: Companies don't track what IP exists across the group—leading to overlooked value and inconsistent treatment.
Fix: Compile a comprehensive intangibles register: patents, trademarks, trade secrets, software, customer databases, key contracts.
Problem: Documentation assumes the legal owner performs all functions without analysis.
Fix: Create a DEMPE matrix for each significant intangible, documenting which affiliate performs each function with supporting evidence.
Problem: Paying nominal royalties to the holding company without considering who actually developed and exploits the IP.
Fix: Analyze substance—if R&D and marketing happen elsewhere, those entities deserve compensation.
Problem: Specifying a 5% royalty because "that's industry standard" without analysis.
Fix: Document the basis for the rate: comparable licenses, valuation models, or profit-split analysis.
Problem: Cross-border R&D cost-sharing without formal agreements.
Fix: Formalize all CCAs with the required OECD elements, including buy-in/buy-out provisions.
Problem: Transferring partially developed technology at a low value without contemporaneous forecasts.
Fix: Document all projections, risk assessments, and valuation assumptions at transaction time.
Problem: Subsidiary spends heavily building the brand locally but receives no recognition.
Fix: Document local marketing investment and ensure the subsidiary receives arm's length compensation for its Enhancement contribution.
Problem: CCA agreements lacking exit provisions or defined benefit shares.
Fix: Include all OECD-required elements, especially buy-in/buy-out formulas and contribution valuation methods.
Scenario: US Parent invented a patented manufacturing process. UK Subsidiary manufactures products using the patent and sells locally. Parent licenses the patent to Subsidiary.
DEMPE Analysis:
| Function | Entity | Evidence |
|---|---|---|
| Development | US Parent | R&D lab, scientist teams, project records |
| Enhancement | UK Subsidiary | Local process improvements, engineering modifications |
| Maintenance | US Parent | Patent renewals, ongoing validation |
| Protection | US Parent | Patent registration, enforcement |
| Exploitation | UK Subsidiary | Manufacturing, sales, customer relationships |
Pricing: Assume a CUP search identifies comparable patent licenses in this industry yielding 4-6% royalties. Based on this analysis, a 5% royalty on net sales is selected. UK Subsidiary's enhancement contribution supports positioning within the range.
Documentation includes: License agreement, comparables search report, DEMPE matrix, functional analysis narrative.
Scenario: German HQ owns a global trademark. Indian Subsidiary promotes and sells products under the brand locally, investing heavily in advertising.
DEMPE Analysis:
| Function | Entity | Evidence |
|---|---|---|
| Development | German HQ | Original brand creation, global identity |
| Enhancement | Indian Subsidiary | Local advertising, market adaptation |
| Maintenance | Both | HQ sets standards; India maintains local presence |
| Protection | German HQ | Trademark registration, enforcement |
| Exploitation | Indian Subsidiary | Local sales, customer relationships |
Key Issue: India's marketing investment creates local marketing intangibles. The OECD recognizes that local Enhancement contributions should be reflected in arm's length conditions—this may warrant adjustments beyond a simple royalty (e.g., profit-sharing, reduced royalty rates, or separate compensation for marketing services).
Documentation includes: Trademark license, marketing spend by entity, allocation of brand-building value, DEMPE analysis addressing local contributions.
Scenario: Japan and US affiliates jointly develop next-generation software. Each funds 50% of development and will share resulting IP equally.
CCA Documentation:
| Element | Detail |
|---|---|
| Participants | Japan Co., US Co. |
| Scope | Development of Software Platform X |
| Contributions | Japan: existing code base, 50% cash; US: development team, 50% cash |
| Interest | Each party owns 50% of resulting IP |
| Balancing | Annual true-up if contributions deviate from 50/50 |
| Exit Provisions | Buy-out at arm's length value per DCF methodology |
Documentation includes: Signed CCA agreement, contribution tracking schedules, benefit allocation formula, buy-in/buy-out calculation methodology.
Documentation Guides:
Technical Resources:
Glossary:
Under the OECD definition, intangibles are non-physical, non-financial assets that can be owned or controlled and for which independent parties would pay. This includes patents, trademarks, trade secrets, customer lists, software, goodwill, and know-how. The focus is on economic value, not accounting treatment—many valuable intangibles (self-developed technology, internal know-how) don't appear on balance sheets but are crucial for transfer pricing. If an asset provides competitive advantage and independent parties would pay for it, it's likely an intangible.
DEMPE stands for Development, Enhancement, Maintenance, Protection, and Exploitation—the five functional categories used to analyze who creates value in an intangible. DEMPE matters because the OECD requires that returns from intangibles be allocated according to who performs and controls these functions, not just who holds legal title. The entity doing R&D (Development), improving the asset (Enhancement), or commercializing it (Exploitation) must receive arm's length compensation aligned with its contributions. DEMPE shifts focus from contracts to substance—what each entity actually does, controls, and assumes.
It depends on who actually did the work. If the legal owner (e.g., a holding company) performs all DEMPE functions—funding, directing, protecting, and commercializing the IP—it keeps the bulk of returns. But if another affiliate did the development or marketing, that affiliate deserves compensation. OECD guidance is explicit: a legal owner who neither controls nor performs DEMPE functions is "not entitled to any portion of such returns." Document the functional reality—the affiliate(s) that economically created the intangible should get a fair share.
Use a functional analysis approach. For each significant intangible, create a DEMPE matrix with functions as rows and group entities as columns. Document for each function: which entity performs it, who controls decision-making, what personnel and assets are involved, and what risks are assumed. Support this with evidence—R&D project charters, patent filing records, marketing budgets, organizational charts. The OECD states this function-by-function analysis is "one of the key considerations" in pricing intangibles. Providing a clear DEMPE breakdown is essential for defensible documentation.
An HTVI is an intangible that, when transferred, has no reliable comparables and highly uncertain projections. Examples include partially developed technologies, assets with distant payoffs, or novel IP with no track record. HTVI matters because tax authorities may use hindsight—looking at actual outcomes as presumptive evidence of arm's length pricing. However, OECD guidance sets limitations (including a 20% deviation threshold concept) and allows taxpayers to rebut with robust ex ante documentation. To protect against adjustments, document detailed projections, risk analyses, and assumption rationales at the time of the transaction. If outcomes deviate from expectations, you'll need to show the deviation resulted from unforeseeable events or fell within projected ranges.
A properly documented Cost Contribution Arrangement should include: (1) all participants and their roles, (2) scope of the joint activity, (3) each party's contributions (cash, existing IP, services) and their arm's length valuation, (4) each party's expected interest in resulting IP and benefit share, (5) balancing payment procedures to equalize unequal contributions, and (6) buy-in/buy-out provisions with valuation methodology. The CCA should be a formal written agreement consistent with what independent parties would expect—informal cost-sharing by email creates audit risk. Keep contribution tracking records and annual true-up calculations.
Method choice depends on availability of data. The OECD recommends CUP (if comparable licenses exist) and profit-split methods (for unique intangibles with multiple contributors) as most reliable. Income approaches (DCF, relief from royalty) are common when market comparables are scarce—document all assumptions (projections, discount rate, useful life). Cost approaches are disfavored because R&D cost rarely correlates with value, but may apply for routine intangibles. Whatever method you choose, document the rationale, all assumptions, comparable searches if applicable, and sensitivity analysis.
Yes. If a local affiliate invests heavily in building a brand's local presence—through advertising, market adaptation, customer relationship building—this creates value that should be documented and compensated. The OECD recognizes marketing intangibles (trademarks, trade names, customer data) as a distinct category. Document the affiliate's marketing spend, local brand-building activities, and how its Enhancement function contributes to intangible value. Failing to account for local marketing investment can leave significant value unrecognized—and create exposure if tax authorities identify the gap.