Loading...
Loading...
Worked examples for the five main transfer pricing methods: CUP, resale price method, cost plus, TNMM, and profit split. See how each method turns facts into an arm's length price or return.
Borys Ulanenko
CEO, ArmsLength AI
Continue exploring
Browse the full resource library or contact us if you want recommendations for your specific use case.
The five main transfer pricing methods answer different pricing questions:
| Method | Example question | Typical calculation output |
|---|---|---|
| CUP | What price did independent parties charge for the same or highly comparable transaction? | Arm's length price per unit |
| Resale Price Method (RPM) | What gross margin should a routine reseller keep? | Transfer price from resale price minus gross margin |
| Cost Plus | What gross mark-up should a routine supplier earn on production or service costs? | Transfer price from cost base plus gross mark-up |
| TNMM | Is the tested party's net profit in line with comparable companies? | Arm's length operating margin, net cost plus, or other PLI |
| Profit Split | How should combined profit be divided when both parties make unique contributions? | Allocation of combined profit between related parties |
Examples are simplified. In real documentation, the calculation sits behind the functional analysis, tested party selection, comparability analysis, data screening, and method selection narrative. Start with the full transfer pricing methods guide if you need the method-selection logic before the math.
Do not start with a formula. Start with the transaction.
| Question | Why it matters |
|---|---|
| What is being transferred? | Goods, services, IP, financing, or a bundle may require different methods. |
| Who performs the key functions? | Routine execution points toward one-sided methods; unique contributions may point toward profit split. |
| Which assets are used? | Manufacturing assets, inventory, marketing intangibles, and IP ownership affect expected returns. |
| Which risks are controlled? | Market, inventory, product, credit, and capacity risks affect the arm's length reward. |
| What reliable data exists? | A perfect-looking method fails if the comparables or segmentation are not reliable. |
The examples below assume the facts have already been delineated and focus on mechanics.
ParentCo sells the same commodity-grade input to:
The uncontrolled sale is close in product specification, timing, currency, delivery terms, and volume. The only material difference is freight: ParentCo pays freight in the related-party sale, while the unrelated customer pays freight separately.
| Item | Related-party sale | Uncontrolled sale |
|---|---|---|
| Product | Commodity input A | Commodity input A |
| Quantity | 10,000 units | 10,000 units |
| Third-party price | n/a | $102.00 per unit |
| Freight difference | ParentCo pays freight | Customer pays freight |
| Freight adjustment | $2.00 per unit | n/a |
If ParentCo charged SubCo $96.00 per unit, the tested price is below the adjusted CUP by $4.00 per unit.
CUP is the most direct method when a sufficiently comparable uncontrolled transaction exists. The hard part is not the arithmetic. The hard part is proving that differences in product, volume, contract terms, timing, market level, geography, and risk either do not materially affect price or can be reliably adjusted.
Internal CUPs are often more persuasive than broad external market observations because the taxpayer can usually explain product specifications, terms, volumes, and timing in more detail.
For deeper CUP guidance, see the Comparable Uncontrolled Price glossary entry and the transfer pricing methods guide.
ManufacturerCo sells finished goods to DistributorCo, a related limited-risk distributor. DistributorCo resells the products to independent customers without material transformation. DistributorCo performs routine sales, warehousing, and order fulfillment, but does not own valuable marketing intangibles.
Independent distributors performing similar functions earn gross margins between 22% and 28%, with a median of 25%.
| Item | Amount |
|---|---|
| DistributorCo third-party resale revenue | $1,000,000 |
| Selected arm's length gross margin | 25% |
| DistributorCo operating expenses | $180,000 |
RPM starts from the resale price and backs into the transfer price by leaving the distributor an arm's length gross margin.
At a $750,000 purchase price, DistributorCo keeps $250,000 of gross profit. After $180,000 of operating expenses, it earns $70,000 of operating profit.
RPM is a gross-margin method. It is strongest when the reseller's functions and accounting are comparable to the distributors used as benchmarks. It can be less reliable where gross profit is distorted by inconsistent classification of discounts, freight, warranty costs, rebates, or selling expenses.
Do not use RPM only because the tested party is called a distributor. If the reseller performs significant manufacturing, owns valuable local marketing intangibles, or bears major entrepreneurial risks, a simple gross-margin resale method may understate its contribution.
For method context, see resale price method discussion and tested party selection.
ContractManufacturerCo manufactures components for PrincipalCo under PrincipalCo's specifications. PrincipalCo owns product IP, controls market risk, and commits to purchase the output. ContractManufacturerCo performs routine manufacturing and does not own unique intangibles.
Comparable independent contract manufacturers earn gross mark-ups on production costs between 8% and 12%. The selected point is 10%.
| Item | Amount |
|---|---|
| Direct materials | $2,000,000 |
| Direct labor | $700,000 |
| Manufacturing overhead | $300,000 |
| Production cost base | $3,000,000 |
| Selected gross mark-up | 10% |
Cost plus works when the supplier's cost base is a reliable measure of its value-creating activity. The key judgment is defining the cost base consistently with the comparables.
| Cost item | Common treatment question |
|---|---|
| Direct materials | Included if the manufacturer bears procurement and production responsibility. |
| Pass-through costs | Often excluded or separately reimbursed if no value is added. |
| Idle capacity costs | Need fact-specific treatment; may not be fully chargeable to the principal. |
| Stock-based compensation or unusual charges | Require consistency with the benchmark and local rules. |
Cost plus is a gross method. A TNMM "net cost plus" PLI is different: it usually tests operating profit over total operating costs. The names sound similar, but the profit line and cost base are not the same.
For related guidance, see cost plus in the transfer pricing methods guide and Cost Plus vs TNMM.
ServiceCo provides routine IT support services to group companies. It uses no unique technology, bears limited commercial risk, and is selected as the tested party. Independent comparable service providers earn net cost plus returns between 5% and 9%, with a median of 7%.
| Item | Amount |
|---|---|
| ServiceCo operating costs | $4,000,000 |
| Actual intercompany service revenue | $4,200,000 |
| Actual operating profit | $200,000 |
| Actual net cost plus | 5.0% |
| Arm's length range | 5.0% to 9.0% |
| Selected point | 7.0% |
First calculate the tested party's actual PLI.
If the policy uses the median point of 7%, calculate the target operating profit and revenue.
The year-end true-up required to reach the selected point is:
TNMM does not usually price an individual invoice directly. It tests whether the tested party's net profitability is arm's length. The practical output is often a policy rate, a target PLI, or a year-end adjustment.
Use a PLI that matches the tested party's value driver. Net cost plus often fits routine services and contract manufacturing. Operating margin often fits routine distributors. Return on assets is more relevant where operating assets drive value.
For a fuller explanation, see TNMM vs CPM, tested party selection, and the benchmarking study guide.
USCo and EUCo jointly develop and commercialize a software platform. USCo contributes core engineering IP and product architecture. EUCo contributes market-specific product design, major customer relationships, and regional commercialization know-how. Both parties make unique and valuable contributions, and reliable one-sided comparables do not capture the combined value creation.
The group applies a residual profit split:
| Item | USCo | EUCo |
|---|---|---|
| Routine function benchmark | $1,500,000 | $1,000,000 |
| Unique contribution score | 60% | 40% |
Combined operating profit from the controlled business is $10,000,000.
Step 1: Allocate routine returns.
Step 2: Calculate residual profit.
Step 3: Split the residual profit.
Step 4: Add routine and residual returns.
| Party | Routine return | Residual share | Total profit allocation |
|---|---|---|---|
| USCo | $1,500,000 | $4,500,000 | $6,000,000 |
| EUCo | $1,000,000 | $3,000,000 | $4,000,000 |
| Total | $2,500,000 | $7,500,000 | $10,000,000 |
Profit split is not just TNMM with a different label. It is a two-sided method that requires reliable combined profit data and a defensible allocation key. It is most useful when each party's contribution is too important to leave as an untested residual.
The OECD guidance treats profit split as potentially appropriate where parties make unique and valuable contributions, operations are highly integrated, or parties share economically significant risks. The method can be powerful, but it is data-intensive.
For a dedicated walkthrough, see the profit split method guide.
Use this table as a starting point, not as a substitute for method selection.
| Fact pattern | Method to consider first | Why |
|---|---|---|
| Same product sold to related and unrelated parties under comparable terms | CUP | Direct price evidence may exist. |
| Routine distributor resells finished goods without transformation | RPM or TNMM | RPM if gross-margin comparability is strong; TNMM if net-level data is more reliable. |
| Contract manufacturer produces under principal specifications | Cost plus or TNMM | Cost plus if gross cost mark-ups are reliable; TNMM if net margins are more reliable. |
| Routine intra-group service provider | Cost plus or TNMM | Cost base is usually the value driver. |
| Limited-risk distributor with reliable public company data but weak gross margin consistency | TNMM | Net margins may be more reliable than gross margins. |
| Commodity transaction with quoted market prices | CUP | Market price data may support direct price testing. |
| Both parties own valuable intangibles or jointly control key risks | Profit split | One-sided methods may leave too much value as an untested residual. |
| Bundled transaction with goods, services, and IP | Depends on delineation | You may need to separate components or test the bundle if separation is unreliable. |
The "best-looking" calculation is not always the most appropriate method. The method has to follow the facts, available data, and reliability of comparability adjustments.
| Mistake | Why it matters |
|---|---|
| Jumping straight to TNMM | You may miss a reliable CUP or gross-margin method. |
| Using RPM for a distributor with major local intangibles | Routine distributor margins may not reward the real value contribution. |
| Calling TNMM "cost plus" | Net cost plus under TNMM is not the OECD gross cost plus method. |
| Ignoring pass-through costs | Including large pass-throughs can dilute or distort mark-ups. |
| Applying profit split without combined profit data | The method needs reliable financial segmentation across parties. |
| Treating an example as a safe harbor | Worked examples illustrate mechanics; they do not replace benchmarking or local law. |
TNMM is often easier to apply in practice because public company net margin data is more available than exact transaction-price or gross-margin comparables. That does not mean TNMM is always the most reliable method. If a reliable CUP exists, it may provide a more direct measure.
No. The OECD cost plus method is a gross profit method that applies a mark-up to an appropriate cost base. TNMM net cost plus is a net profit indicator, usually operating profit over total operating costs. They can produce similar-looking mark-up percentages but test different profit levels.
RPM is stronger when the tested party is a routine reseller and reliable gross-margin comparables are available. TNMM may be more reliable when gross margins are distorted by accounting differences but net operating results are comparable.
Yes, but documentation should identify the most appropriate method. A secondary method can be used as a corroborative check, especially where the primary method has data limitations.
No. A defensible local file or master file needs functional analysis, transaction delineation, method selection, comparable search logic, financial data, adjustments, and jurisdiction-specific disclosures. These examples show calculation mechanics only.