Loading...
Loading...
How to apply the OECD cost plus method for property and services: defining the cost base, gross mark-up vs margin, internal and external comparables, service-specific caveats, pass-through costs, the LVAS simplified approach, and when to fall back to TNMM.
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 cost plus method is one of the traditional transactional methods in the OECD Transfer Pricing Guidelines (2022 edition, Chapter II, Part II, Section C). It sets an arm's length price by starting from the costs the supplier incurs for the controlled supply of property or services, then adding a gross profit mark-up that reflects the functions performed, assets used, and risks assumed by that supplier.
describes cost plus as probably most useful where semi-finished goods are sold between associated enterprises, where joint facility agreements or long-term buy-and-supply arrangements exist, or where the controlled transaction is the provision of services. The method's scope is broader than manufacturing alone: services are an equally important application.
US terminology alert: In US regulations, "CPM" usually means the Comparable Profits Method, a net margin method aligned with OECD TNMM. In OECD language, "cost plus" is the gross mark-up on costs method. This article uses cost plus for the OECD method throughout. For TNMM vs US CPM, see CPM vs TNMM. For a full methods tour, see Transfer pricing methods: the complete guide.
Cost plus sits alongside CUP and resale price as a traditional transactional method. The difference is the economic "perspective" each method tests.
| Method | Perspective | Profit indicator | Often strongest when |
|---|---|---|---|
| CUP | Transaction price | None (direct price benchmark) | Commodities, some financial/royalty benchmarks |
| Resale price method (RPM) | Buyer / distributor | Gross margin on resale | Limited-risk distribution without transformation |
| Cost plus | Supplier / manufacturer / service provider | Gross mark-up on costs | Routine manufacturing, many routine services |
| TNMM | Tested party | Net profitability (OM, NCP, etc.) | Gross-level data is unreliable or unavailable |
| Profit split | Both parties | Combined profit allocation | Unique contributions on both sides |
The OECD framework does not rank cost plus above TNMM in the abstract. Cost plus is often preferable when reliably comparable gross cost and mark-up information exists for genuinely similar suppliers. TNMM is frequently the better fit when cost classification or accounting diversity across comparables makes gross mark-ups hard to compare.
Traditional transactional methods (CUP, RPM, cost plus) generally require a higher degree of product or service comparability than transactional profit methods such as TNMM. A functional match alone may not be enough for cost plus if the products or services are materially different in complexity, specification, or risk.
Conceptually:
Transfer price = Cost base x (1 + Arm's length gross mark-up %)
Or equivalently:
Transfer price = Direct costs + Indirect costs in the base + (Cost base x Mark-up %)
The precise contractual mechanics (pass-through treatment, billing cadence, budgeted vs actual cost true-ups) should reflect what independent parties would accept for comparable arrangements.
The cost base under cost plus is not a fixed concept. It depends on the nature of the controlled transaction and the supplier's role:
For manufacturers, the anchor is typically cost of sales / COGS:
For service providers, the cost base may be broader than production-style COGS. The OECD defines the cost plus mark-up by reference to the supplier's direct and indirect costs of providing the service. Where relevant, certain operating expenses (supervisory, general, and administrative costs associated with service delivery) may need to be included to achieve meaningful comparability.
This distinction matters: a cost base limited to "COGS only" may be too narrow for services, while a cost base that includes all operating expenses starts to approximate a net-level analysis. Where the cost plus base approaches total costs in practice, the boundary with TNMM blurs, and practitioners should be transparent about which method they are actually applying.
Cost base boundary issues are audit-prone: what your ERP calls "COGS" and what a comparable files as COGS can differ even within the same industry. For services, the divergence is often wider because service providers may have no meaningful COGS line at all. Cost plus studies often live or die on cost mapping, not on database exports alone.
The gross mark-up is gross profit expressed relative to cost of sales (or, for services, relative to the defined cost base):
Gross mark-up % = Gross profit / Cost base
That is not the same as gross margin (gross profit / revenue). Most databases and annual reports emphasize margin. The conversion is:
Gross mark-up = Gross margin / (1 - Gross margin)
Example: a 10.7% gross margin implies roughly a 12% gross mark-up on costs, since 0.107 / (1 - 0.107) = 0.12.
Many financial databases (Orbis, TP Catalyst) allow you to pull gross mark-up as a discrete ratio (gross profit / COGS) directly, rather than converting from margin. Where available, pulling the ratio directly is safer than converting aggregate quartile margins manually.
Cost plus is not universal. It tends to fit routine suppliers with defined functions, limited risk, and no unique intangibles that would make a cost-only benchmark misleading.
A classic fact pattern is contract manufacturing: specifications, materials, and IP supplied by a principal; the manufacturer provides processing capacity and routine execution. If inventory and market risk sit with the principal, a gross mark-up on production costs is often a clean analytical story.
Toll manufacturing goes further: the principal supplies inputs and the toller charges for conversion. Cost plus is frequently an excellent match, subject to cost base consistency.
Cost-based methods are a common fit for intra-group services where the supplier performs a routine function and the arrangement is essentially cost recovery plus a routine return. However, services method selection has its own hierarchy:
Before pricing any service, you still need the benefits test: does the service confer an economic benefit the recipient would have been willing to pay for (or would have performed itself)? Shareholder activities and duplicated services must be filtered before the cost base and mark-up question arises.
For low-value-adding services (LVAS), the OECD's Chapter VII simplified approach is a separate pathway, discussed in its own section below.
Where a related manufacturer sells intermediate products for further processing, CUP may be missing and RPM may not fit the buyer's transformation economics. Cost plus can anchor pricing to the seller's production economics if the seller is truly the right tested party.
The OECD recognizes that in some transactions, the supplier acts only as an agent or intermediary for certain costs, adding no value to those specific inputs. In such cases, the pass-through costs should be excluded from the mark-up base, and the mark-up applied only to the supplier's own service or agency-function costs.
Examples: third-party software licenses procured on behalf of the group, travel expenses booked on behalf of another entity, subcontracted inputs where the supplier has no procurement or coordination role.
Getting pass-through treatment wrong in either direction is costly. If genuine pass-throughs are marked up, the transfer price overstates the arm's length result. If costs where the supplier adds real coordination, procurement, or management value are passed through at zero margin, the price is too low.
Document the rationale for treating specific cost items as pass-throughs. Authorities expect to see a clear link between each item's classification and the supplier's actual role.
Before locking in cost plus, document functions, assets, and risks. The tested supplier should look routine:
If the entity owns valuable IP, absorbs large warranty risk on finished goods, or behaves like a principal, cost plus may understate the appropriate return and invite method challenges. See tested party selection.
This is the most operationally critical step:
The OECD indicates that the supplier's cost plus mark-up should ideally be established by reference to the mark-up that the same supplier earns in comparable uncontrolled transactions (internal comparables). If the same entity sells similar services or goods to both related and unrelated parties under comparable terms, that internal evidence can be highly reliable.
External comparables serve as a guide where internal evidence is absent or insufficient.
Isolate routine manufacturers or service providers with similar FAR and similar products or services, not merely the same NACE/SIC code. Key filters:
Run quantitative screens (size, activity, data years), then conduct manual review. Document accept/reject rationale for every company reviewed; authorities often scrutinize exclusions more than selections. See quantitative screening filters and benchmarking study guide.
For each comparable, compute gross mark-ups for available years. Multi-year data is often useful for dampening single-year distortions, but it is not a systematic requirement in all cases. The OECD discusses averaging approaches (see et seq.) without mandating a single method.
Where residual comparability concerns remain after screening and adjustments, practitioners commonly use the interquartile range as one statistical tool to narrow the range. IQR is not an automatic default; it is a response to imperfect comparability. The tested party's mark-up should be explained relative to the resulting range.
If material differences remain between the tested party and comparables, consider adjustments. The central issue in cost plus is ensuring a comparable mark-up on a comparable cost base: if cost bases differ in composition, even a "correct" mark-up comparison can mislead.
Common adjustments include working capital differences (see working capital adjustments) and sometimes capacity utilization narratives (with quantitative support where possible).
This boundary causes the most confusion in practice.
| Feature | Cost plus (OECD) | TNMM |
|---|---|---|
| Profit indicator | Gross mark-up on the relevant cost base | Net profitability vs an appropriate base |
| OECD bucket | Traditional transactional method | Transactional profit method |
| Data needs | Reliable gross profitability inputs with comparable cost bases | Reliable operating profit and denominator data |
| Sensitivity to COGS/OPEX split | Very high | Often lower: net profit can absorb some classification noise |
| Product/service comparability | Higher degree required | More tolerant of functional-only similarity |
| Often preferred when | Clean, comparable gross cost structures exist | Gross comparisons are contaminated by accounting diversity |
A common source of confusion is the term "net cost plus" (operating profit / total costs). This is a TNMM PLI, not a variant of the OECD cost plus method. The OECD services chapter itself refers to CUP, cost plus, or cost-based TNMM as distinct alternatives, which supports the practical point but also shows the need to label methods correctly.
Practical takeaway: if comparables mix capitalization policies, freight treatment, or plant depreciation presentation in ways you cannot normalize, TNMM (a separate method with its own PLI) is often more reliable than forcing a gross mark-up set. For method selection context, see transfer pricing methods guide.
Think of the income statement vertically: cost plus stops at gross profit; TNMM looks through to operating profit after below-the-line operating items. If you cannot defend where the "line" sits for each comparable, you cannot defend a gross mark-up range.
If your tested party capitalizes certain production costs into COGS but a comparable expenses them below gross profit, mark-ups will lie. This is the classic depreciation / QC / inbound freight / plant supervision wage problem: same economics, different presentation.
Comparing a US GAAP tested party against comparables reporting under German HGB or other local European statutory GAAPs amplifies this problem. Explicitly note the accounting standards each comparable reports under.
Mitigation: read notes, normalize where quantifiable, and if the problem is endemic, switch methods (commonly TNMM with a carefully chosen PLI).
Absorption costing loads fixed production overheads into unit costs; marginal costing may leave fixed costs out of the cost base you think you are comparing. Confirm conventions before you compare. In practice, most publicly reported financials use absorption costing, but intra-group reporting may differ.
The OECD () notes that if an associated supplier is less efficient than independent comparables, the additional costs from that inefficiency should not automatically be passed on to the buyer with a mark-up. The cost base should approximate the costs an efficient comparable supplier would incur.
When input prices spike, a fixed multi-year mark-up may diverge from what independents negotiate (pass-through clauses, re-openers, commodity referencing). Document the pricing policy in the intercompany agreement, including how and when mark-ups adjust.
Many groups label internal recharge mechanics "cost plus." OECD cost plus still requires arm's length evidence through comparability analysis (or a valid simplified regime where applicable). A mark-up picked administratively is not a method conclusion by itself.
In manufacturing, marking up standard (budgeted) costs with variances handled separately is common. In services, the cost pool may be actual costs. The choice affects how risk is allocated between supplier and buyer and should be documented explicitly.
For qualifying low-value-adding intra-group services (LVAS), the OECD Chapter VII ( et seq.) describes an elective simplified approach. Key characteristics:
The 5% is not a general benchmark. It applies only to services that meet the LVAS definition (supportive, not core business, not requiring unique intangibles) and only in jurisdictions that have adopted the simplified approach. Where a tax administration has not adopted it, the group follows local requirements. The 5% should not be used as a benchmark for services outside the LVAS definition or outside the elective simplified scheme.
Qualifying services typically include: IT support, HR administration, accounting and finance functions, internal communications, and general back-office support.
Services that typically do not qualify: R&D, manufacturing, procurement functions, treasury activities involving significant risk, and any service where unique intangibles drive value.
For deeper services documentation patterns, see documentation for services.
If classification reconciliation fails, TNMM (a separate method) is often next. If intangibles or risk are substantial, cost plus alone is unlikely to carry the file.
Strong cost plus work is still comparables work: the right routine profile, transparent screens, and documented rejects. ArmsLength AI helps teams run comparable searches and maintain accept/reject rationales and documentation trails suited to audit review, covering both internal and external comparable identification.
If you want to explore the workflow, start from AI benchmarks or Talk to us.
For qualifying low-value-adding services under the OECD simplified approach (), 5% on the relevant cost pool (excluding pass-throughs) is the specified outcome, where the local jurisdiction has adopted the approach. For services that do not qualify for LVAS or where the jurisdiction has not adopted it, the arm's length mark-up depends on a full comparability analysis reflecting the specific functions, risks, and assets of the tested party. Do not extrapolate the 5% to non-qualifying services.
Yes. The OECD indicates the supplier's cost plus mark-up should ideally be established first by reference to mark-ups the same supplier earns in comparable uncontrolled transactions. If the supplier also sells to unrelated parties under comparable terms, that evidence can be more directly comparable than external database searches.
Rarely as a first-line method. IP is often unique by nature, pushing analysis toward CUP (true royalty comparables), profit split, or Chapter VI approaches for hard-to-value intangibles. Cost plus might appear only in narrow routine hosting/administration fact patterns and even then CUP or TNMM may be more reliable.
Sometimes, when CUP and RPM are genuinely unavailable or inferior. Expect heavier scrutiny: you must explain why a supplier-side gross mark-up is more reliable than alternatives given how the product is sold and priced in the value chain.
Billing mechanics allocate costs internally. The OECD method compares your mark-up to independent benchmarks (or fits within an accepted simplified regime). One can exist without the other. Auditors care about arm's length evidence, not labels.
Multi-year data is often useful to smooth volatility, but the OECD does not mandate a fixed number as a systematic requirement. Three to five years aligned to available financials and business cycles is common in practice. OECD guidance discusses averaging approaches ( et seq.) without prescribing a single formula.
When gross-level data is untrustworthy after classification review, or when comparables' cost base boundaries cannot be harmonized without heroic adjustments. TNMM (with a carefully chosen PLI) is a separate method and frequently the more defensible path. Note that a "net cost plus" PLI (operating profit / total costs) is TNMM, not a cost plus variant.
Costs where the supplier acts as a pure agent or intermediary (adding no value) should be passed through without a mark-up. The mark-up applies only to the supplier's own costs where it contributes functions, assets, or risk. Document the classification rationale for each pass-through item.