What Are Transfer Pricing Methods? The Five OECD Methods in UAE Practice

When two companies in the same group trade with each other — a Dubai parent selling to its Saudi subsidiary, a Sharjah manufacturer buying from a related supplier — the price they set moves profit between countries, and therefore between tax bases. Transfer pricing methods are the approved techniques for answering the only question tax authorities care about: what would unrelated parties have charged for the same transaction? That hypothetical market price is the arm's-length price, and the five OECD methods are the routes to it.

The methods split into two families. Traditional transaction methods compare prices directly: CUP compares the price itself; resale price works backwards from the resale margin; cost plus works forwards from the cost base. Transactional profit methods compare profitability: TNMM tests net margins against comparables; profit split divides the combined profit as independent parties would. The choice is not a hierarchy — the OECD and the UAE require the most appropriate method for the specific transaction, judged on comparability, data availability, and reliability.

In the UAE, this is not academic. The Corporate Tax Law's arm's-length principle applies to all transactions between related parties and connected persons, and the FTA can adjust non-arm's-length pricing — with penalties. Every UAE group with cross-border or domestic related-party dealings needs a defensible method selection.

Comparable Uncontrolled Price (CUP): the most direct method — compare the related-party price with the price in a comparable uncontrolled transaction. Internal CUP (the company sells the same product to third parties) is gold standard; external CUP (market prices for the same commodity) works for traded goods. Best for: commodities, standard products, intercompany loans benchmarked to market rates.

Resale Price Method: start from the price at which the related distributor resells to third parties, subtract an arm's-length gross margin (from comparable distributors), and the remainder is the arm's-length purchase price. Best for: distribution activities where the reseller adds limited value.

Cost Plus Method: start from the supplier's costs, add an arm's-length mark-up (from comparable manufacturers or service providers). Best for: contract manufacturing, intra-group services, semi-finished goods.

Transactional Net Margin Method (TNMM): test the net profit margin (operating margin, return on assets, Berry ratio) the tested party earns on the controlled transaction against comparable independent companies. The workhorse method — best for: routine manufacturing, distribution, and services where gross-level data is messy but net margins are comparable. It is the most used method worldwide.

Transactional Profit Split: divide the combined profit from the controlled transactions between the parties as independent enterprises would — via contribution analysis or residual analysis. Best for: highly integrated operations and unique intangibles where neither party can be benchmarked alone (joint development, shared IP exploitation).

The UAE Corporate Tax Law embeds the arm's-length principle for transactions between related parties, and FTA guidance follows the OECD framework — the five methods are the accepted toolkit. UAE groups face distinctive fact patterns: Dubai headquarters charging management fees to Gulf subsidiaries (cost plus or TNMM on services); JAFZA distributors buying from foreign parents (resale price or TNMM); UAE manufacturers in industrial zones producing for related distributors (cost plus or TNMM); and IP holding structures licensing trademarks into operating companies (CUP on royalties or profit split).

Documentation is where UAE practice is maturing. Large groups face master file, local file, and country-by-country reporting obligations; all taxpayers with related-party transactions disclose them in the Corporate Tax return. The FTA has been building transfer pricing audit capability, and method selection is the first thing challenged — a TNMM applied to a unique-intangible transaction, or a CUP asserted without genuine comparables, invites adjustment.

Practical UAE considerations: comparable company data for the Gulf is thinner than for Europe or the US, so benchmarking studies often use regional or global comparables with adjustments — a defensible approach if documented, but one the FTA scrutinises. Intercompany loan pricing needs CUP-style benchmarking to market rates (EIBOR-plus margins for comparable credit). And domestic related-party transactions are not exempt — the arm's-length principle applies within the UAE too, including transactions with free zone persons.

GulfStar Group: a Dubai parent provides IT support services to its three Gulf subsidiaries, costing the Dubai service centre AED 4,000,000 per year. The group must set an arm's-length service fee.

Method selection: the service is routine, low-value-adding, with no unique intangibles — CUP fails (no comparable uncontrolled service price exists); profit split is overkill. The choice is between cost plus and TNMM.

Cost plus application: benchmarking finds comparable independent IT support providers earning mark-ups on total cost of 8–12% (arm's-length range). Applied to AED 4,000,000 cost: fee range AED 4,320,000–4,480,000. The group sets AED 4,400,000 (10% mark-up, the median).

TNMM cross-check: the service centre's operating margin at that fee is 10/110 ≈ 9.1% — within the comparable net-margin range of 7–11% for routine service providers. The two methods converge, strengthening the defence.

Contrast with a distribution case: a JAFZA subsidiary buys finished goods from its German parent for AED 20,000,000 and resells for AED 26,000,000 (gross margin 23%). Comparable independent distributors earn 20–26% gross margins — the resale price method confirms the AED 20,000,000 purchase price is arm's length. Had the margin been 40%, the FTA could argue the purchase price was inflated to shift profit to Germany — and adjust it down, increasing UAE taxable profit.

One-method-fits-all is the classic failure: applying TNMM to every transaction — services, loans, IP licences, commodity sales — because the benchmarking database makes it easy. Each transaction needs its own most-appropriate-method analysis; an IP licence tested by TNMM when a CUP on royalty rates exists is indefensible.

Second, cherry-picked comparables. A benchmarking set of three hand-picked companies, or comparables from unrelated industries, collapses under FTA review. The comparability analysis — functions, assets, risks — must be genuine and documented.

Third, ignoring the tested party. TNMM tests the simpler, routine party — but groups sometimes test the complex entrepreneur by mistake, or test a loss-making distributor without asking why independent distributors would accept losses.

Fourth, set-and-forget pricing. An arm's-length price set in 2024 is not automatically arm's length in 2026 — markets move, functions change. Transfer pricing policies need periodic review, and the documentation must reflect the year under audit.

Fifth, UAE groups sometimes assume domestic transactions are safe. They are not — the FTA can adjust related-party pricing within the UAE, and the disclosure in the tax return puts every transaction on the record. Document domestic pricing with the same rigour as cross-border.

Arm's-length principle is the foundation the methods serve. Transfer pricing documentation (master file, local file) is where method selection is defended. Country-by-country reporting gives tax authorities the map on which method challenges are launched. Related parties are the transactions' participants. TNMM is the most-used method and deserves its own study. And benchmarking is the comparables analysis that powers CUP, resale price, cost plus, and TNMM alike.