How the DEMPE framework works for high-value intangibles: a practical guide beyond the glossary
TL;DR
DEMPE (Develop, Enhance, Maintain, Protect, Exploit) is the OECD Chapter VI test for who earns the return on an intangible. Legal ownership is the starting point; functional substance decides where the profit lands. A defensible DEMPE analysis maps every function, asset, and risk to the entity that actually performs and controls it.
What is the DEMPE framework in transfer pricing?
DEMPE is the OECD test from Chapter VI of the Transfer Pricing Guidelines that allocates intangible profits to the entities that perform and control the functions of Developing, Enhancing, Maintaining, Protecting, and Exploiting an intangible. Legal title sets the starting position; functional substance decides the arm's length share.
OECD Chapter VI (2017, refreshed in the 2022 consolidated Guidelines) introduced DEMPE to close the BEPS Action 8-10 gap between paper ownership and economic reality. The principle is simple: an intercompany IP licence cannot move profit to a low-substance principal if the actual DEMPE functions remain elsewhere. The execution is harder, because each of the five letters has to be evidenced separately.
In practice a DEMPE analysis becomes the spine of any IP-heavy structure — pharma R&D arrangements, tech principal models, brand licensing, post-acquisition IP migrations. It is also the most contested area of a TP audit, because tax authorities increasingly start from DEMPE and only reach for comparables once the function allocation is locked.
Why does DEMPE matter for intangibles specifically?
Intangibles can be transferred on paper in minutes — a licence agreement, an IP holding company, a cost-contribution arrangement. The OECD response was to disregard paper transfers that lack substance. DEMPE is the substance test. Without it, the legal owner gets only a routine return; with it, full residual profit.
The classic pre-BEPS pattern — IP holding entity in a low-tax jurisdiction, royalty stream out, no operating substance — fails DEMPE on its face. Even when the legal owner retains formal IP title, audit authorities will reallocate the residual return to the entities performing R&D, brand management, regulatory filing, and commercial exploitation. The legal owner is left with a financing return on the funding it actually provided, nothing more.
DEMPE applies whether the intangible is a patent (pharma), a software stack (tech), a trade name (consumer), customer data (FinTech), regulatory dossiers (med-tech), or process know-how (industrial). It is format-agnostic. The five functions and their control questions are the same.
How do you conduct a DEMPE analysis step by step?
Build five function maps — one for each DEMPE letter — and identify the entity performing AND controlling each. Then layer in the assets used and the risks borne. The entity that controls a function (not just executes it) and bears the related economic risk earns the residual return on that function's contribution.
- Inventory the intangible. Be specific: not 'the brand' but the trademark portfolio, marketing know-how, customer relationships, distribution agreements. OECD para 6.6 lists what counts.
- Map the Development function. Who designs the IP, runs R&D, files patents, owns the technical roadmap? Tag each subfunction to an entity AND name the senior decision-maker.
- Map Enhancement. Continuous improvements, version releases, brand campaigns, format extensions. Often spread across multiple entities.
- Map Maintenance. Renewals, defensive filings, infrastructure upkeep, regulatory monitoring. Routine but essential.
- Map Protection. Litigation strategy, anti-counterfeit work, trade-secret governance. Frequently centralised in legal HQ.
- Map Exploitation. Commercialisation, licensing decisions, market launch, pricing authority. Where the revenue conversion happens.
- For each function, identify the controller, not just the operator. Control = capacity to make decisions about whether to perform the function and how to perform it (OECD para 6.56).
- Allocate residual returns by control. Routine returns to operators; residuals to controllers. Document the linkage to comparables for each.
What are the most common DEMPE mistakes that trigger audits?
Three: confusing control with execution; treating R&D as if it ended at patent filing; and pricing intercompany licences without re-running DEMPE every two to three years. Audits also look for substance asymmetry — a principal entity with no functions documented under any of the five letters.
| Failure mode | Why it loses on audit | Fix |
|---|---|---|
| Principal entity claims residual return with no functions performed | Cannot evidence control of any DEMPE letter | Move functions to the principal OR move the residual to where functions sit |
| R&D centre paid on cost-plus, no upside | Enhancement + Development controlled there; residual mislocated | Re-characterise as DEMPE controller, share residual on a profit-split |
| Brand IP owner with no marketing function | Exploitation control absent; legal title without economic substance | Either internalise marketing or accept routine licence return only |
| Post-acquisition IP transfer without re-pricing | Functions migrate; DEMPE map stays frozen | Refresh DEMPE annually post-deal; document re-pricing trigger events |
How does DEMPE interact with hard-to-value intangibles (HTVI)?
HTVI rules (OECD Chapter VI Section D.4) let tax authorities re-price intercompany IP transfers ex-post if actual outcomes deviate materially from projections. DEMPE feeds the HTVI defence: a robust function map evidences that the projections reflected real risk allocation, reducing the authority's mandate to second-guess.
The interaction matters most for pharma, biotech, and software where the transferred IP is early-stage and the upside is genuinely uncertain. Documenting the DEMPE controller's risk-taking capacity (financial, technical, regulatory) at the time of the transfer is what turns a HTVI challenge from a re-price into a debate the taxpayer can win.
What does the case law tell us about DEMPE in practice?
Coca-Cola (US Tax Court 2020, 11th Circuit appeal 2024–2025), Medtronic III (US Tax Court 2022 remand), and several European MAP outcomes have all converged on the same point: principal entities without DEMPE substance get re-characterised. Authorities are willing to set aside the contractual structure when functions are demonstrably elsewhere.
On the European side the case law is more discreet because most TP disputes resolve in MAP rather than open court, but the published Italian and German decisions follow the same direction. The takeaway: documentation built around a clear DEMPE analysis is the single most important audit-defence asset for any IP-driven structure.
Domande frequenti
Domande comuni
Does DEMPE apply to financial instruments?
No — financial transactions are governed by OECD Chapter X (2020), not Chapter VI. DEMPE applies to intangibles. For intercompany loans, guarantees, and cash pooling, the equivalent framework is the accurate delineation analysis under Chapter I and the specific guidance in Chapter X.
Can DEMPE allocation change after a transaction?
Yes. If functions migrate post-deal (a common pattern after M&A integration), the DEMPE map must be refreshed and the intercompany pricing re-run. Authorities increasingly look for evidence of an annual DEMPE review, especially in pharma and tech structures.
Is DEMPE only relevant for large groups?
DEMPE applies whenever intercompany intangible transactions exist, regardless of group size. Mid-market groups with a single IP holding entity are now common audit targets, because the substance asymmetry is more visible at smaller scale than in a multinational with hundreds of legal entities.
Who at the firm typically runs the DEMPE analysis?
A defensible DEMPE map sits at the intersection of legal, tax, R&D, and commercial. The TP team coordinates, but the source evidence — who decides what, who signs off, who bears which risk — comes from interviews with the actual function owners. A purely tax-led DEMPE risks being a paper exercise.
Altri in questa categoria
Technical Foundations
Intra-group financial guarantees under OECD Chapter X: valuation, documentation, and defence
When a parent guarantees a subsidiary's external borrowing, the arm's length fee turns on three questions: how much credit uplift the guarantee actually delivers, how much of that is implicit support, and how the residual is priced.
8 giugno 2026 · 12 min
Transfer pricing benchmarking studies: methodology for defensible arm's length analysis
Most benchmarking studies fail audit not on the comparables themselves but on the comparability adjustments, the screening logic, and the rejection trail. Methodology is what an auditor reads first.
30 maggio 2026 · 10 min
FAQ