Loading...
Loading...
This guide explains what RPM tests, how to apply it step by step, and when TNMM, CUP, or cost plus is more reliable than a gross-margin resale-price method.
Borys Ulanenko
CEO, ArmsLength AI
Continue exploring
Browse the full resource library or contact us if you want recommendations for your specific use case.
This guide explains what RPM tests, how to apply it, and when to use a different transfer pricing method.
After reading, you will be able to:
For how RPM sits alongside CUP, cost plus, TNMM, and profit split, see the transfer pricing methods guide. The OECD rules are in Chapter II: Transfer pricing methods.
The resale price method, or RPM, starts with the price a related-party reseller charges an independent customer and works backward to infer an arm’s-length purchase price from its affiliate. In simple terms, if you know what the distributor sold the product for in the market, RPM asks: what gross margin should that distributor have kept, given what it actually does and the risks it actually bears? The remainder points to the arm’s-length transfer price.
states that RPM begins with the price at which product purchased from an associated enterprise is resold to an independent enterprise, reduces that resale price by an appropriate resale price margin (gross margin) that covers selling and operating expenses and an appropriate profit for the reseller’s functions (including assets used and risks assumed), and treats the remainder—after adjustment for other purchase-related costs such as customs duties—as an arm’s length price for the original controlled purchase.
RPM is designed for a specific fact pattern: a group company purchases goods from a related supplier and resells them to third parties without significant transformation. It is most commonly used for buy-sell distribution arrangements involving wholesale distributors, importers, and sales entities that mainly handle warehousing, logistics, and customer-facing selling.
Within the broader transfer pricing framework, RPM is one of the traditional transaction methods, alongside CUP and cost plus—see OECD Chapter II. Under OECD-based systems, it is a recognized method; in the United States, the counterpart appears in the Section 482 regulations as the resale price method.
Related guides: Cost plus in transfer pricing (supplier-side traditional method), CPM vs TNMM (US “CPM” is the net-margin method, not RPM), and tested party selection.
A few terms matter from the start:
At a high level, RPM is a way to test distribution-side controlled purchases by reference to third-party resale activity. Where the facts fit, it can be a direct and intuitive method. Where they do not, gross-margin comparisons become fragile quickly.
The legal basis for RPM is well established, but the method is not appropriate merely because a taxpayer resells goods. The key question is whether a gross-margin method is the most reliable way to price the controlled transaction given the facts, the available comparables, and the applicable local rules.
Under the OECD Transfer Pricing Guidelines, Chapter II, RPM is one of the traditional transaction methods (together with CUP and cost plus). See .
The critical OECD point is method selection. The question is not whether RPM is available in principle, but whether it is the most appropriate method on the facts. requires weighing:
A practical clarification matters here. OECD guidance does not rank methods in the abstract, but it does establish preferences when reliability is equal: provides that where a traditional transaction method and a transactional profit method can be applied with equal reliability, the traditional method is preferable—and where CUP and another method can be applied with equal reliability, CUP is preferred. Conversely, notes cases where transactional profit methods may be more appropriate (for example, limited reliable gross margin data on third parties, or highly integrated / unique contributions where profit split may fit). The controlling question remains the most appropriate method for the specific controlled transaction, evidenced in line with .
In the United States, RPM appears under Treas. Reg. §1.482-3(c) and is evaluated under the best method rule. The practical idea is similar to the OECD standard: choose the method that produces the most reliable measure of an arm’s-length result under the facts.
In a US analysis, the key questions usually include:
US practice often focuses explicitly on the reliability of gross profit comparability, which is where many RPM analyses succeed or fail.
In the UK, HMRC often refers to RPM as resale minus. The substance is the same: start from the independent resale price and subtract an arm’s-length gross margin for the distributor’s functions and risks.
In India, Rule 10B expressly recognizes the Resale Price Method. Indian disputes often turn on whether the local entity is really a routine reseller and whether gross-margin comparability is robust enough to support the method. For India-facing files, a useful action point is to document segmentation and the reseller profile with unusual care—especially where the local entity also performs marketing, after-sales support, or channel-development activities.
Across OECD-aligned systems, RPM is widely recognized. But formal recognition does not make it the right answer. Tax authorities usually expect taxpayers to explain why RPM is more reliable than CUP, TNMM, or another method in the actual fact pattern.
Before running any calculations, four method-selection principles matter.
First, do not choose RPM because it looks simple. Choose it because the facts support a gross-margin method.
Second, select the tested party carefully. RPM usually works best when the reseller is the less complex entity and its gross return can be benchmarked against comparable independent distributors. On choosing the tested party under one-sided methods, see in Chapter III and our tested party selection guide.
Third, do not rely on nominal method preference to cure weak comparability. If gross margins are being driven by accounting differences, channel differences, or functional gaps, a different method may be more reliable.
Fourth, treat comparability adjustments as part of method selection, not an afterthought. If the necessary adjustments are not realistically available, that is often a sign that RPM should not be used.
RPM is built for a specific commercial pattern: a related-party purchase of goods followed by resale to independent customers, with limited value added before sale.
The classic RPM fact pattern looks like this:
That profile matters because RPM starts from the resale price to independent customers and works backward to an arm’s-length gross margin. The method is most reliable when the reseller’s contribution is mainly commercial execution.
RPM is often a good fit where the reseller performs standard distributor functions, such as:
These are typical routine distribution functions. The business may be operationally important to the group, but it is not usually the entity expected to earn returns associated with unique intangibles or non-routine strategic control.
RPM is naturally a distributor method because its logic depends on a visible resale to independent customers. That resale gives the analysis a market-facing starting point.
Manufacturers are different. Their value often comes from transforming inputs, managing production assets, and controlling manufacturing risks. Service providers are further removed still because there may be no resale of goods at all.
That is why RPM tends to be strongest for finished-goods distribution and weaker for businesses where the tested party materially alters the product or provides economically significant bundled services.
Some structures resemble RPM cases at first glance but need closer classification.
Commissionaire versus buy-sell distributor
A commissionaire may interact with customers, but may not take title to the goods. If the local entity is really acting as an agent, the pricing issue may be better framed around a commission or service return than a resale gross margin.
Distributor with embedded services
Some distributors also perform installation, technical support, training, or warranty services. If those functions are minor and separable, RPM may still work. If they are economically significant, the gross margin may reflect more than distribution.
Entrepreneurial distributor versus routine reseller
Not every distributor is routine. Some local entities control strategy, fund major advertising, or assume significant market risk. Those businesses may be legal distributors but poor RPM tested parties.
At a mechanical level, RPM starts from the price the reseller charges an independent customer and works backward to an implied arm’s-length purchase price.
Isolate the controlled buy-sell transaction in which a related supplier sells goods to a related distributor. Then choose the tested party, which is usually the reseller if it performs the simpler set of functions.
If the reseller performs substantial non-distribution functions, the method becomes harder to apply reliably from the outset.
Determine the price at which the product is resold to the first unrelated customer. That resale price is the starting point.
Confirm that the resale being used:
The core RPM question is: what gross margin would an independent distributor earn in comparable circumstances?
contemplates determining the resale price margin by reference to the margin the same reseller earns on comparable uncontrolled purchases and resales (internal comparable), or by reference to margins earned by an independent enterprise in comparable uncontrolled transactions (external comparable).
Establish that margin in this order:
The relevant benchmark is a gross margin on resale, not a net margin.
Once you determine the arm’s-length gross margin, subtract it from the resale price to derive the arm’s-length transfer price.
If the margin is expressed as a percentage of sales:
If an arm’s-length distributor should earn a 25% gross margin on a resale price of 100, the implied arm’s-length transfer price is 75.
If the comparable distributor is not perfectly similar, make adjustments only where they can be supported with reasonable reliability.
Typical areas to examine include:
In many analyses, multiple comparables produce a range of gross margins. In that case, compare the tested result to the arm’s-length range derived from those margins. For how ranges and median / percentile choices are used in practice, see IQR vs full range and IQR calculation.
If the tested result falls outside the range, an adjustment may be required. If it falls within the range, the result is generally more defensible—subject to the reliability of the comparability analysis.
Assume:
Then:
Under RPM, the controlled purchase price would be tested against $700.
Now assume the economic facts are the same, but the tested party books $50 of inbound freight in operating expenses while the comparable distributors include equivalent freight in COGS.
That changes the implied transfer price:
A single accounting classification issue changes the implied transfer price by $50. That is exactly why RPM can be analytically strong in the right case but fragile when gross-profit accounting is inconsistent.
RPM tests a narrow but important question: after buying from the related supplier and before considering below-gross-profit items, did the reseller retain a gross margin consistent with what an independent distributor would have earned?
That means RPM tests:
Comparability is often the hardest part of applying RPM well. Gross margins are usually more sensitive than net margins to differences in functions, channel, accounting treatment, and market conditions. The OECD framework for comparability—including the comparability factors used to select and review comparables—is set out in Chapter III; in practice, also see our benchmarking study guide.
The core question is not simply whether two companies both resell products. It is whether they earn a similar gross return for performing a similar distribution role under similar conditions.
Under RPM, the tested party’s resale price is reduced by an arm’s-length gross margin. That margin must compensate the reseller for its distribution functions, cover its operating expenses, and leave an appropriate profit for the role performed.
Because the measure sits at the gross level, RPM is highly sensitive to differences that may be less disruptive under TNMM when a net profit level indicator is more reliably comparable (see on situations where transactional profit methods may be more appropriate).
Functions are usually the most important driver under RPM. Review whether the reseller:
Relevant risks include:
Useful distinctions include whether the reseller employs:
Review both written terms and actual conduct, including:
RPM comparability also depends on commercial conditions, such as:
This is one of the most important practical issues under RPM. Gross margin depends on what is included in cost of goods sold versus operating expenses.
Before relying on a gross-margin benchmark, assess whether major cost items are treated similarly across the tested party and comparables.
Inventory intensity can materially affect gross margins. A distributor with long holding periods or significant markdown exposure may need a different gross return from a fast-turnover reseller.
Check:
Gross margins vary sharply by channel. A reseller to large retailers may not be comparable to a reseller operating in a service-heavy retail environment, even if both are “distributors.”
Geography matters because gross margins are influenced by local competition, logistics, tariffs, and consumer behavior.
Internal comparables are usually stronger under RPM. They are more likely to reflect:
External comparables can still be acceptable, but they require more caution because public data often lacks the detail needed to confirm true gross-profit comparability.
Product comparability is generally less strict under RPM than under CUP, because RPM focuses more on the reseller’s gross reward for distribution. notes that fewer adjustments are normally needed for product differences under RPM than under CUP, because minor product differences may have a less material effect on margins than on prices. Still, product differences cannot be ignored: requires comparing the property transferred, and explains when broader product differences may be acceptable if functions and economic circumstances align—while emphasizing that closer product comparability still produces a better result.
A distributor of branded luxury cosmetics is not an obvious comparable for a distributor of basic industrial supplies, even if both are technically resellers. Where valuable or unique intangibles are involved, calls for particular attention to product comparability.
Some differences can be adjusted for, but not all. Adjustments may be possible where there is reliable evidence on:
accounting reclassification issues
freight treatment differences
clearly quantifiable rebates or discounts
certain working capital differences, where a working capital adjustment improves comparability and can be made with reasonable accuracy—the OECD Annex to Chapter III illustrates mechanics and cautions against automatic adjustments
limited and well-documented functional differences
When the differences are broader—such as major channel, marketing, or risk differences—the right answer is often to refine the comparable set or reconsider the method.
highlights that RPM depends on comparability of functions (and assets and risks): material differences affecting gross margins may require adjustments, and the extent and reliability of those adjustments drives how reliable RPM is in the case.
RPM is usually strongest where the tested party is a routine reseller buying products from an associated enterprise and reselling them onward without materially changing the product.
RPM is commonly the most appropriate method where the following features are present:
These are the cases where gross profit is a meaningful lens and may provide a more direct answer than a net-margin method.
RPM frequently appears in industries such as:
The common theme is the operating model, not the industry label by itself.
When RPM fits, it may be preferable to TNMM for two reasons.
First, RPM is closer to the controlled transaction. It starts from the resale price to third parties and works back to the transfer price.
Second, RPM can be analytically stronger where a reliable traditional transaction method is available. TNMM remains common in distribution cases, but it is broader and more removed from the transfer price itself.
RPM is often a good candidate if you can answer yes to most of these questions:
Method selection should follow the actual transaction and the quality of the available data, not a default preference for one method.
| Dimension | RPM | CUP | Cost Plus | TNMM | Profit Split |
|---|---|---|---|---|---|
| Starting point | Resale price to third parties | Price charged in comparable uncontrolled transaction | Supplier’s costs plus arm’s-length mark-up | Net profit relative to sales, costs, or assets | Combined profits from the controlled transaction |
| Typical tested party | Distributor/reseller | Either side, if a reliable direct price comparison exists | Manufacturer, service provider, or contract supplier | Usually the less complex party | Both parties together |
| Profit level examined | Gross margin | Direct price | Gross mark-up on costs | Net margin | Allocation of combined profit |
| Primary use case | Routine buy-sell distribution with limited value-added | Highly comparable transactions with strong direct pricing evidence | Routine manufacturing, assembly, or services | Routine activities where gross-margin comparables are weak | Integrated operations with unique contributions on both sides |
| Primary weakness | Sensitive to gross-margin comparability and accounting classification | Very demanding product and transaction comparability | Cost base definition can be contentious | Further removed from the actual transfer price | Complex and data-intensive |
| Sensitivity to accounting classification | High | Lower | High | Moderate to high | High |
| Best-fit transaction type | Routine distribution | Commodity-like or directly comparable transactions | Routine supplier-side functions | Broad set of routine one-sided cases | Highly integrated, multi-intangible cases |
If a reliable CUP exists, CUP will often be preferred because it compares prices directly—see and .
Choose:
RPM and cost plus are often mirror-image methods for traditional transactional testing.
RPM tests gross profitability. TNMM tests net profitability.
That leads to a practical tradeoff:
A related point: the Berry ratio is not a version of RPM. It is a TNMM-style profit level indicator used only in limited fact patterns, usually where the tested party acts more like an intermediary and bears little inventory risk.
RPM is a one-sided method. Transactional profit split is generally used where both parties make unique and valuable contributions, or where the business is too integrated for a one-sided method to capture reliably—consistent with .
If the reseller is routine, does not own unique intangibles, and reliable gross-margin comparables are available, RPM deserves serious consideration. If those conditions are missing, another method—often TNMM—may be more defensible.
RPM becomes fragile when the reseller’s gross margin no longer reflects a relatively routine buy-and-resell function.
RPM is usually strongest where the tested party is a routine distributor. It becomes much weaker when the reseller has created or controls substantial local marketing intangibles, such as brand-building capability, market-specific promotional assets, or premium customer-facing concepts.
In those cases, the reseller’s gross margin may reward more than basic distribution.
RPM is not well suited to entities that do much more than resell. Reliability falls when the reseller performs significant value-added functions, including:
The more the reseller looks like a service provider or integrated commercial operator, the less comfortable RPM becomes. flags that RPM is easiest where the reseller does not add substantial value before resale, and may be hard to apply reliably where goods are further processed or incorporated so that their identity is lost—or where the reseller contributes substantially to intangibles such as trademarks owned by an associate.
RPM can also be unreliable where inventory dynamics materially affect results, including:
notes that a resale price margin is more accurate where realized within a short time of purchase; longer gaps increase the likelihood that market, FX, cost, or other changes confound the comparison.
Gross-margin comparability depends on accounting consistency. Common problem areas include:
Where accounting treatment differs materially and reliable adjustments cannot be made, RPM often loses credibility.
Even if RPM is conceptually attractive, it may fail in practice because reliable gross-margin data is unavailable. Public databases often provide operating-margin information more consistently than true usable gross-profit data.
Warning signs include a distributor that bears:
RPM works best where products, customers, and channels are reasonably consistent. It becomes much less reliable when the tested party mixes very different product categories or route-to-market models without clean segmentation.
A common reason to reject RPM is that distribution cannot be cleanly isolated from other functions, such as:
When RPM breaks down, the analysis usually shifts to a method that better matches the facts:
In practice, RPM works best when applied in a disciplined sequence.
Define exactly which related-party transaction you are testing.
At this stage:
Test whether the reseller’s profile actually fits RPM. Map functions, assets, and risks in line with Chapter I (arm’s length principle) and the comparability framework in Chapter III. For interview-style collection of FAR detail, see functional interviews.
Document:
Also identify early whether the reseller performs broader functions that may disqualify RPM or require tighter segmentation.
Search for comparables in a clear order: internal first, external second. For database screening discipline (NACE/SIC, independence, loss-makers), see quantitative screening filters and loss-making comparables where relevant.
If internal comparables are unavailable or unreliable, document the external search carefully, including:
Normalize the data before calculating margins.
Key steps include:
Then:
Explain not just the outcome, but the reasoning behind it.
Document:
Even where RPM is conceptually the right method, execution often fails.
RPM tests gross margin, not EBIT margin, operating margin, or Berry ratio. If the available data only supports a net-margin analysis, that may indicate TNMM is the more reliable method.
Two companies can perform similar functions but report very different gross margins if one books key items in COGS and the other books them in SG&A.
Common trouble spots include:
RPM usually works best when applied to the least complex reseller. A distributor that performs strategic, value-adding, or risk-heavy functions may be a poor RPM tested party.
A label like “distributor” is not enough. Major differences in channel, customer base, service content, and risk profile can make gross margins incomparable.
Installation, technical support, customization, and bundled after-sales services can all distort gross-margin comparisons if they are embedded in the same revenue stream.
Broad benchmark sets are especially risky under RPM. A portfolio of loosely screened distributors does not become reliable merely because it produces a range.
A gross-margin computation is not enough. A defensible RPM analysis should explain:
An RPM file usually succeeds or fails on whether you can show that the reseller’s gross margin was calculated correctly and that the selected comparables are genuinely comparable at the gross-profit level. Position the analysis within your wider local file narrative using the transfer pricing documentation guide; OECD documentation principles appear in Chapter V. For typical documentation failure modes, see common documentation weaknesses.
A defensible RPM analysis should include:
The most persuasive evidence is often operational and accounting evidence, including:
If the reseller performs mixed activities, an entity-wide gross margin may be misleading. Authorities often want to see that:
Recurring weaknesses include:
Strong RPM documentation is less about volume than traceability.
Use this quick screen before committing to a full RPM analysis.
Before finalizing an RPM conclusion, check whether local authority practice makes a theoretically sound analysis hard to defend in the actual file.
In the US, RPM is assessed under the best-method rule, not as a default choice.
Practical implications:
In the UK, the method is commonly described as resale minus.
For UK files:
In India, RPM is a prescribed method, but disputes often recur over whether the local entity is truly a simple reseller and whether the gross-margin comparison is made on a like-for-like basis.
For India-facing files:
Across jurisdictions, local practice may still push taxpayers toward TNMM even where RPM is conceptually available, especially when gross-margin comparables are difficult to defend.
Before finalizing the report, ask:
Yes. In practice, Resale Price Method (RPM) and resale minus refer to the same transfer pricing method. The label “resale minus” describes the mechanics: start with the resale price to an independent customer, subtract an arm’s-length gross margin, and work back to an arm’s-length purchase price.
Sometimes. RPM can still work where the activities are routine, do not materially change the product, and do not meaningfully alter gross-margin comparability. If the assembly or packaging becomes economically significant, another method may be more reliable.
Internal comparables are usually preferred because they often reduce differences in market conditions, accounting treatment, customer profile, and operating model. They are not automatically reliable, but where they are genuinely comparable, they are often stronger than external distributor benchmarks.
Usually no. RPM is designed primarily for the purchase and resale of tangible goods. It is generally not a natural fit for services, licensing transactions, or transfers of valuable intangibles.
RPM tests whether the reseller earns an arm’s-length gross margin. TNMM tests a net profit indicator, such as EBIT margin or, in limited cases, the Berry ratio. Berry ratio is not a form of RPM.
That can complicate or undermine RPM. You need to determine whether the service element is minor, separately compensated, or significant enough to distort gross-margin comparability. If the functions cannot be isolated reliably, TNMM or a segmented analysis may be more appropriate.
Yes. A result inside the arm’s-length range does not cure weak method selection, poor comparables, unsupported adjustments, or unreliable gross-profit data. Authorities can still reject RPM if the foundation of the analysis is weak.
Document the transaction, tested-party selection, FAR analysis, method-selection reasoning, comparable search, gross-margin calculations, accounting consistency checks, segmentation, and final conclusion. For more detail, see the Documentation and Evidence Tax Authorities Expect section above and Chapter V (documentation).