Transfer Pricing Policy Design

Policy design is the forward-looking exercise of deciding how related-party transactions will be priced — method selection, pricing bands, Board approval — before or as an arrangement goes live. Documentation, principally the annual Local File and Master File, is the backward-looking exercise of recording and evidencing pricing already applied during a completed financial year. A policy feeds directly into the documentation; documentation without an underlying policy tends to justify whatever happened rather than reflect a designed decision.

A Local File can describe the pricing method applied for a completed year without there being a standalone policy document that governs future years. A written, Board-approved policy is not itself a UAE statutory filing requirement, but it is the strongest practical evidence that pricing decisions were made deliberately and consistently, which materially strengthens the arm's length position the Local File later describes — and it gives finance a working reference for pricing the next transaction correctly the first time.

Method selection follows the Functional, Asset and Risk (FAR) analysis, not a default preference. We establish which entity performs which functions, owns which assets including intangibles, and bears which risks for each transaction category, then test whether Comparable Uncontrolled Price, Resale Price, Cost Plus, TNMM, or Profit Split fits those facts and whether reliable comparable data is available to support it.

A pricing band is a defensible range — typically an interquartile range drawn from a screened set of comparable companies or transactions — within which the group's actual pricing can sit and still be considered arm's length. A single fixed figure is more fragile: normal business fluctuation can push it outside a defensible range, whereas a band gives finance workable flexibility while still anchoring pricing to market evidence.

Yes, and often more urgently. Article 35's definition of Related Party turns on ownership and control, not on crossing a border — a mainland trading entity and its own Free Zone affiliate under common ownership are related parties for Corporate Tax purposes. Where the Free Zone entity holds Qualifying Free Zone Person status, mispriced intercompany dealings can threaten that 0% rate directly, making the policy design conversation higher-stakes than a purely cross-border one.

These fall under Article 36's Connected Person rules, not the Related Party rules, but they need the same arm's length treatment to remain deductible. We benchmark owner salary against comparable market compensation data, and director-owned premises rent against comparable rental data, then set a defensible figure — or confirm the current figure already sits within a defensible range — as part of the same policy engagement.

Yes, the related-party mapping and FAR analysis can run in parallel with Corporate Tax registration finalisation, since neither depends strictly on the other. The final policy document typically references the taxable person's Tax Registration Number once issued, but the substantive design work — method selection, pricing bands, Board approval — does not need to wait for the TRN.

For a mid-sized group with reasonably prompt access to financial records and the people who can explain each entity's function, a first-time policy design engagement typically runs five to six weeks from kick-off to Board-approved policy — covering related-party mapping, FAR analysis, method testing, pricing band design, and drafting. Groups with more entities or more transaction categories take proportionally longer.

PNPC recommends formal Board (or equivalent governing body) approval and a dated sign-off, even for smaller owner-managed groups, because that approval itself becomes part of the evidentiary record that pricing decisions were made deliberately and with proper oversight — not simply assumed by whoever happened to be handling finance at the time.

We treat this as a triggered mini-review — applying the same FAR and method-selection logic to the new arrangement specifically, rather than forcing it into an ill-fitting existing category or leaving it unpriced until the next annual refresh. The outcome is folded into the policy document as an addendum so the framework stays current.

The financial data underlying each band is generally refreshed annually to reflect current-year comparable results, while the underlying comparable company search itself is typically refreshed on a multi-year cycle, commonly every three years or sooner if the industry or functional profile changes materially, consistent with OECD guidance on documentation currency.

It can, but the immediate priority in an active request is producing the Local File and Master File within the stipulated window, generally 30 days, and engaging the FTA on the technical position. PNPC often runs an urgent documentation response and a policy design engagement in parallel in that scenario — closing the immediate gap while also fixing the underlying absence of a designed policy so the same issue does not recur next year.

Usually yes, at least in adapted form. A global policy can be a strong starting point and often sets the method and general approach correctly, but UAE-specific elements — the Connected Person rules under Article 36, the Qualifying Free Zone Person interaction, and comparable data appropriate to the UAE entity's actual market — are rarely captured accurately by a template built for another jurisdiction's rules.

Yes. PNPC scopes the policy design engagement based on the number of entities, related-party transaction categories, and connected-person payment types involved, and agrees a fixed, written fee before work begins. We do not bill open-ended hourly time for a deliverable with a defined scope and output.

A policy is inherently a forward-looking governance document — it cannot rewrite pricing already applied and reported for a closed financial year. What it can do is set the method and bands correctly for the current and future periods, and PNPC separately assesses whether the pricing already applied for the closed period can be supported by a best-available benchmarking rationale for that year's Local File, treating the two as related but distinct pieces of work.

Transactions with genuinely similar functional and risk profiles — several management fee arrangements with different sister entities that all receive the same bundle of services, for example — can reasonably share a single pricing band under the policy. Transactions that differ materially in function, risk, or the assets involved need to be treated as separate categories, even if they are superficially similar in nature, such as 'services' broadly defined.

The policy sets an allocation key — commonly headcount, revenue, or actual usage — that reflects how the benefit of the shared service genuinely accrues across the entities involved, then applies a cost-plus markup consistent with what an independent shared-services provider would earn. The allocation key itself is a FAR question: it has to match the real driver of benefit, not simply the easiest figure to pull from the accounts.

Yes. A related-party sale of property, equipment, or other fixed assets needs to be priced at what an independent buyer and seller would agree at the transaction date, generally supported by an independent market valuation or comparable market transaction data, rather than transferred at book value or original cost, which rarely reflects current market value.

Yes. A guarantee has value — it is the credit enhancement the guaranteed entity would otherwise have to pay an independent guarantor or accept a higher interest rate to obtain — and an arm's length guarantee fee should be set and documented, similar in principle to how an interest-free related-party loan needs a documented rationale for the absence of interest.

No. If the FAR analysis and benchmarking show that current pricing sits outside a defensible range, the policy is designed around the genuinely arm's length position going forward, not around preserving the status quo. We flag the gap explicitly, together with the practical options for closing it prospectively, so the group can make an informed decision rather than have the gap surface for the first time during an FTA review.

Yes. Where a company does not have a formal board, PNPC recommends a documented sign-off by the entity's governing body under its own constitutional documents — commonly the sole owner, the manager named on the trade licence, or the designated signatory — dated and recorded in the same way a board resolution would be, so the same evidentiary purpose is served regardless of the company's formal governance structure.

A royalty or licence fee is benchmarked against comparable independent licensing arrangements or industry royalty-rate data where available, testing whether the rate reflects what an independent licensee would pay for equivalent rights, taking into account the exclusivity, territory, and nature of the intellectual property involved. This is treated as its own transaction category in the policy, distinct from general management fees.

The policy itself focuses on the Corporate Tax arm's length position, but PNPC checks that the pricing it sets is not inconsistent with how the same transaction is valued for VAT purposes, since UAE VAT law applies its own market-value rule to supplies between related parties in certain circumstances. The two regimes serve different purposes and can produce different figures, but a management fee priced one way for Corporate Tax and invoiced a materially different way for VAT on the same underlying supply is worth reconciling rather than leaving unexamined.

A holding company typically has a narrower transaction profile — dividend flows (generally outside the arm's length pricing rules), intercompany loans to operating subsidiaries, and possibly a management or strategic oversight fee — while an operating company usually has a wider range of transaction categories including goods, services, and connected-person payments. The policy is scoped to each entity's actual transaction profile rather than applying a single generic template across every entity in the group.

PNPC's role in that scenario is to resolve the disagreement by reference to the FAR analysis and comparable data, not by internal negotiation between the entities' finance teams. An allocation key or fee level that emerges from negotiation between related parties, rather than from what an independent arrangement would look like, is exactly the pattern the arm's length standard is designed to test — so we anchor the resolution in the data, not in what each side is prepared to accept.

The policy document itself is an internal governance artefact — it is not a statutory filing and is not shared with counterparties or the FTA as a routine matter. It is retained by the group and produced to the FTA only if requested as part of a broader transfer pricing information request, typically alongside the Local File, and can also be shared selectively during due diligence for financing, investment, or acquisition purposes at the group's discretion.

Yes, typically. The policy defines a materiality threshold, informed by the disclosure and documentation thresholds under Ministerial Decision No. 97 of 2023, below which a transaction is monitored but not subjected to the full method-selection and benchmarking process, while transactions at or above the threshold are priced under the formal framework. Smaller transactions are still tracked in aggregate by category, since several immaterial transactions can collectively become material.

A benchmarking study is the technical comparable-data search and analysis that supports a specific pricing method for a specific transaction category — it is an input into the policy, not the policy itself. The policy is the broader governance document that records which method applies to which category, the resulting pricing bands, the Board approval, and the process for pricing new arrangements going forward. PNPC scopes benchmarking studies as a standalone deliverable for clients that already have a policy framework and simply need the underlying comparable data refreshed or extended to a new transaction type.

The FAR analysis, method selection, and much of the comparable-data benchmarking can proceed on management accounts and provisional figures, but final pricing bands are more reliably confirmed once the audited or finalised financial statements are available, since benchmarking outputs are typically expressed relative to costs, sales, or margins drawn from those figures. PNPC often runs the analytical work in parallel with year-end close and finalises the bands once the numbers are settled.

Yes, if it has related-party or connected-person transactions above the relevant thresholds — the Article 34 arm's length requirement applies to every taxable person, not only those that have elected QFZP status. A Free Zone entity taxed at the standard 9% rate above the AED 375,000 threshold on non-qualifying income still needs defensible related-party pricing, and the policy exercise also gives the group a clean basis for assessing whether a future QFZP election is viable.

The depth of the FAR analysis is proportionate to the complexity and materiality of the transactions involved, not fixed at one standard depth regardless of group size. A small group with two or three straightforward transaction categories — say, a management fee and a shareholder loan — needs a properly reasoned but appropriately scaled FAR analysis, not the same depth of interview and documentation a forty-entity multinational would require.

PNPC does not design a policy backwards from a predetermined tax outcome. The FAR analysis and method selection are run on the entity's actual facts, and the resulting pricing band is whatever that analysis genuinely supports — if that outcome does not match what management hoped for, the policy reflects the defensible position, and any tax planning discussion happens separately and transparently as a distinct decision, not disguised as a benchmarking result.

The document structure — method per category, pricing bands, review cadence, governance process — follows a consistent framework across engagements so the deliverable is recognisable and complete, but the content within that structure, including every method selection, band, and Connected Person figure, is derived entirely from the specific group's FAR analysis and benchmarking. Nothing in the substantive content is copied from another client's policy.

A tested party's own profitability is not, by itself, evidence that intercompany pricing is wrong — independent businesses can be loss-making for genuine commercial reasons unrelated to related-party pricing. PNPC tests whether the pricing method and band are appropriate given the FAR profile, and separately considers whether the loss reflects normal business risk the tested entity genuinely bears, rather than assuming a loss automatically signals mispriced intercompany transactions.

A single policy document can and typically does cover every related-party and connected-person transaction category for the group — goods, services, loans, royalties, and Connected Person payments — organised by category within one governance document, rather than as separate standalone policies for each transaction type. This keeps the review cadence, Board approval, and implementation handover consistent across the whole group.

Yes, where relevant. A group preparing for external investment or an IPO can expect investor and lender due diligence to specifically test whether related-party pricing is genuinely arm's length and properly governed, rather than assumed. PNPC builds the policy with that scrutiny in mind for clients already anticipating a transaction, including making sure the Board approval trail and benchmarking rationale are presentation-ready, not just internally adequate.

A full FAR re-analysis is generally only required when something material has changed — a new entity, a restructuring, a materially different function or risk profile for an existing entity. Where the group's structure and functions are unchanged, the annual review is a lighter refresh: updated financial data against the existing pricing bands, and a check for any new arrangement that needs adding to the framework, rather than rebuilding the FAR analysis from scratch each year.

Where a group has related-party flows between an Indian entity and a UAE entity, PNPC's India transfer pricing practice and Dubai desk work from a single agreed set of facts about each entity's functions, assets, and risks, so the method selected and the pricing band set for the UAE policy are consistent with the position being built for the Indian entity's own documentation under Section 92 of the Income-tax Act, rather than two disconnected positions arrived at independently.

The policy design engagement is delivered almost entirely remotely — document review, FAR interviews over video call with the people who run each entity, benchmarking, and drafting all proceed without a physical-presence requirement. The one genuine dependency is timely access to the right people for the functional interviews, since that conversation cannot be reliably reconstructed from documents alone, not physical location.

The most useful preparation is a current group ownership chart with percentages, a rough list of intercompany transaction types and approximate annual values, copies of any existing intercompany agreements, the most recent financial statements, and a note on which entities are Free Zone-registered and whether any has elected Qualifying Free Zone Person status. With those on hand, the scoping call can move straight to a reasoned view of the engagement's scope rather than spending the session gathering basic facts.