Transfer Pricing Basics, Part 2 of 3. Part 1: what transfer pricing is and what it covers. You are here: how a price is proven. Part 3: who has to file documentation — tiers, thresholds and penalties.
Part 1 ended with the transaction delineated: functions, assets and risks on each side, and conduct taking precedence over the contract. That delineation is the input here. Change it and the method changes with it.
The five transfer pricing methods
Paragraph 2.1 names five methods in two families. The comparable uncontrolled price, resale price and cost plus methods are the traditional transaction methods; the transactional net margin method and the transactional profit split are the transactional profit methods.
| Method | What it compares | Typical use | What the analysis needs |
|---|---|---|---|
| Comparable uncontrolled price (CUP) | The price itself, against a comparable uncontrolled transaction | Commodities, quoted instruments, intercompany loans and guarantees | A genuinely comparable third-party transaction, or differences that can be reliably adjusted |
| Resale price | The gross margin a reseller earns | A distributor that buys and resells without transforming the goods | Gross margins measured consistently; a reseller that does not build marketing intangibles |
| Cost plus | The mark-up on the supplier's cost base | Contract manufacturing, contract R&D, routine intra-group services | A defensible cost base and a mark-up derived from comparables, not from experience |
| Transactional net margin (TNMM) | A net profit indicator — profit relative to costs, sales or assets | One party has the less complex functions and only net-level comparables exist | Financial data for independent companies with comparable functions, consistently computed |
| Transactional profit split | How the combined profit of the transaction is divided | Unique contributions on both sides, highly integrated operations, shared significant risks | Reliable financials from both parties and allocation keys that survive challenge |
Paragraph 2.2 sets the standard — selection "always aims at finding the most appropriate method for a particular case" — judged on the strengths and weaknesses of each method, its appropriateness given the transaction as delineated by the functional analysis, the availability of reliable information, and the degree of comparability including the reliability of any adjustments. Paragraph 2.3 adds two tie-breaks that bite only where two methods are equally reliable: the traditional method is preferable to the profit method, and the CUP is to be preferred above all. Paragraph 2.12 removes the belt-and-braces habit: the principle does not require applying more than one method to a transaction.
So the exercise is not elimination. You select one method and record why it fits the transaction and the data you actually have. Which method usually wins, and why the answer is the third selection criterion rather than any ranking, is the subject of our piece on choosing a transfer pricing method.
Three choices decide the result
- The tested party. Per paragraph 3.18 it is "most often the one that has the less complex functional analysis" — for a routine distributor or contract manufacturer, the subsidiary rather than the principal.
- The profit level indicator. An EBIT margin, a net cost-plus mark-up, a gross margin, a Berry ratio. It has to reflect how that party earns its return, and it has to be computed identically for the tested party and every comparable. Accounting differences, not economics, decide most arguments.
- The comparables. A search for independent companies whose functions and risks resemble the tested party's, screened for independence, loss-making years and activities that break comparability, with each rejection recorded. How much searching is enough is itself answered in the Guidelines, and the answer is proportionate rather than maximal — the subject of our note on how much search for comparables is enough.
The range, and why it matters more than the price
A search yields a set of results, not one number. Where comparability defects remain that cannot be identified or quantified, paragraph 3.57 allows statistical tools such as the interquartile range; the wording is permissive — such tools "might assist in enhancing the reliability of the analysis" — rather than the automatic trim many studies apply.
Two consequences follow, and they are the whole commercial point of the exercise. Under paragraph 3.60, a result inside the arm's length range needs no adjustment. Under paragraph 3.62, where an adjustment is required and comparability defects of the kind described at paragraph 3.57 remain, measures of central tendency such as the median may be used. Being inside the range is protection; being outside it exposes you to the middle of someone else's data set.
A worked example
A German manufacturer sells finished goods to its French subsidiary, which markets and resells them locally, holds inventory, carries the receivables, and runs no R&D and no brand of its own.
Delineation puts the routine functions in France, so the subsidiary is the tested party under paragraph 3.18. A CUP is unavailable — nothing comparable is sold to third parties. The functions are distribution rather than transformation, so a net profit indicator is the workable choice and the EBIT margin the natural one. A search for independent European distributors with comparable functions produces, on illustrative figures, an interquartile range of 1.8 % to 4.6 % with a median of 3.1 %. The subsidiary's actual EBIT margin is 0.6 %.
That sits below the range, so paragraph 3.60 gives no protection and an adjustment is in prospect — in Germany towards the median under section 1(3a) of the Foreign Tax Act, unless another value within the range better satisfies the arm's length principle. Note what carries the argument: not the method, which was uncontroversial, but the comparables. Change the search criteria and the range moves.
When is the price actually set?
There are two practices and the Guidelines recognise both at paragraphs 3.69 to 3.71: setting the price in advance, and testing the outcome after the year end. German law makes the timing explicit — section 1(3) sentence 4 of the Foreign Tax Act requires the circumstances at the time the transaction was agreed to be decisive — so a group that only calculates after the financial statements are closed has to record which data were available when the price was agreed.
One exception now bypasses the study altogether. Amount B, published in February 2024 and annexed to Chapter IV of the Guidelines, supplies a return on sales from a fixed matrix for in-scope baseline marketing and distribution activity, for fiscal years beginning on or after 1 January 2025 in adopting jurisdictions. Where it applies, the number comes from the matrix. Everywhere else, the benchmarking study still sets it.
Which makes this a data problem
If the method turns on the availability of reliable information, the data set decides which method you are entitled to use — and how defensible the answer is years later, in an audit, to someone who was not there.
Genuine comparables for a routine European entity are overwhelmingly private companies, and a listed-only peer set cannot support a TNMM twice. Benchmarking Pro, the transfer pricing database smartZebra launched on 27 July 2026, was built for that constraint: 700,000+ curated company data points across 15+ European countries, priced as a flat fee rather than per search, and included for existing Transfer Pricing Pro users. The transfer pricing module covers the ratios these studies actually use — EBIT margin, net and gross cost plus, adjusted net cost plus, gross margin, return on assets and the Berry ratio — screens comparables for independence, losses and R&D intensity, classifies them by SIC and NAICS code, computes the interquartile range and median, and opens the original source report behind every data point. Transfer Pricing Pro is the transfer pricing software layer around it, exporting the benchmarking appendix straight into the file; what makes such a database defensible in the first place is set out in our guide for tax consultants.
Financing is the exception that proves the rule. Interest rates have an observable market, so arm's length interest rates are a CUP exercise under Chapter X rather than a margin proxy.
The method choice stays with the adviser. What the next part covers is the other half of the obligation: writing it down, on time, in the form each country requires.
Next in this series: Part 3 — Who Has to File Transfer Pricing Documentation?
References
- OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations 2022 — Ch. II paras. 2.1 (the five methods in two families), 2.2 (most appropriate method and the four selection criteria), 2.3 (preference for traditional methods and for the CUP where equally reliable), 2.12 (no requirement to apply more than one method), 2.64 (TNMM definition), 2.98 (cost-based indicators), 2.106–2.107 (Berry ratio)
- OECD Transfer Pricing Guidelines 2022 — Ch. III paras. 3.18 (tested party), 3.57 (interquartile range where unquantifiable comparability defects remain), 3.60 (no adjustment within the range), 3.62 (measures of central tendency), 3.69–3.71 (price-setting and outcome-testing approaches), 3.80–3.83 (proportionality of search effort)
- OECD, Pillar One – Amount B, OECD/G20 Inclusive Framework on BEPS, February 2024, annexed to Chapter IV of the Guidelines as "Special considerations for baseline distribution activities" — fiscal years beginning on or after 1 January 2025 in adopting jurisdictions
- Germany: § 1 Abs. 3 Satz 4 Außensteuergesetz (circumstances at the time the transaction was agreed) and § 1 Abs. 3a Satz 4 (adjustment to the median unless another value within the range better satisfies the arm's length principle)
Related pages
- Transfer Pricing Benchmarking & TNMM Analysis — the module, the profit level indicators and the calculation log on live data
- Benchmarking Pro — transfer pricing database with 700,000+ company comparables — ratio coverage, country coverage and the flat-fee model
- Part 1 — What Is Transfer Pricing? — definition, scope and the arm's length principle
- Part 3 — Who Has to File Transfer Pricing Documentation? — the three tiers, thresholds and penalties
- Why Does the TNMM Win? Choosing Between the Five Methods — the selection rules applied to a real decision
- How Much Search for Comparables is Enough? — the OECD's proportionality standard for benchmarking effort
- Transfer Pricing Databases: A Guide for Tax Consultants — selection criteria and what auditors reject
