The arm’s length principle is the single idea that underlies every transfer pricing system in the world. The Income Tax Act in India, the Internal Revenue Code in the United States, the General Anti-Avoidance Rule in Canada and the United Arab Emirates Corporate Tax law all share this one premise: when two members of the same multinational group trade with each other, the price they charge each other must approximate the price that two independent enterprises would have charged in the same circumstances. This article explains what the principle is, where it comes from, how Indian tax authorities apply it, and what it looks like in practice across the most common types of intra-group transactions. It is written for tax heads, CFOs and finance leaders who want a working understanding before commissioning or reviewing a transfer pricing study.
Where the Arm’s Length Principle Comes From
The arm’s length principle has its modern origin in Article 9 of the OECD Model Tax Convention, which states that where conditions are made or imposed between two associated enterprises in their commercial or financial relations that differ from those which would be made between independent enterprises, then any profits which would have accrued to one of the enterprises but, by reason of those conditions, have not so accrued, may be included in the profits of that enterprise and taxed accordingly. India incorporates this principle into its tax code through the transfer pricing provisions of the Income Tax Act (formerly Chapter X with Section 92 as the operative provision, and now mapped into the Income Tax Act, 2025 framework).
Why the principle exists is straightforward: without it, a multinational group would be free to set intra-group prices so that profits accumulate in low-tax jurisdictions and losses in high-tax ones. With it, each country gets to tax the profits that economic activity in its territory would have earned at market terms. The arm’s length principle is therefore the line between legitimate tax planning that respects market terms and base erosion that artificially shifts profit. The OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations, last updated in 2022, provide the authoritative practitioner-level guidance on how the principle is to be applied.
The principle is now embedded in the domestic tax law of more than 100 countries. The wording differs, but the substance is the same. In India, Section 92 (and its successor in the Income Tax Act, 2025) requires every international transaction or specified domestic transaction between associated enterprises to be computed having regard to the arm’s length price. The associated enterprise definition is broad and covers control through equity, management, debt or contractual rights. The international transaction definition is similarly broad and covers tangible goods, intangibles, services, loans, guarantees and any other transaction that has a bearing on the profits, income, losses or assets of the enterprise.
How the Principle Is Applied: The Five OECD Methods
The OECD Guidelines prescribe five methods for determining the arm’s length price, and Indian law adopts the same five methods plus a sixth Other Method that the Central Board of Direct Taxes has prescribed for use where none of the first five is the most appropriate. The five methods are Comparable Uncontrolled Price (CUP), Resale Price Method (RPM), Cost Plus Method (CPM), Transactional Net Margin Method (TNMM) and Profit Split Method (PSM). Choosing the most appropriate method is the central judgment of any transfer pricing study and is documented in the Local File.
CUP: Comparable Uncontrolled Price. Compares the price charged in the tested transaction directly with the price charged in a comparable uncontrolled transaction. Works well where the product or service is sufficiently homogeneous and where a comparable third-party price is observable. Most often used for commodities, standard intermediate goods and royalty rates.
RPM: Resale Price Method. Starts from the resale price to an independent customer and works backwards by deducting an appropriate gross margin to arrive at the arm’s length transfer price. Best suited for distributors that do not add significant value to the product before reselling.
CPM: Cost Plus Method. Starts from the costs incurred by the supplier in a controlled transaction, marks up by an appropriate gross profit, and treats the result as the arm’s length price. Best suited for manufacturers and service providers performing routine functions.
TNMM: Transactional Net Margin Method. Examines the net profit margin relative to an appropriate base (such as cost, sales or assets) that a taxpayer realises from a controlled transaction. The most widely used method in India, especially for IT services, ITES, contract manufacturing and routine distribution.
PSM: Profit Split Method. Splits the combined profit (or loss) from controlled transactions among the associated enterprises in a way that approximates what independent parties would have agreed. Best suited for highly integrated operations where neither party performs only routine functions, such as global trading or co-developed intangibles.
The Arm’s Length Range and Why It Matters
Real-world comparables rarely produce a single price. They produce a range, and the OECD Guidelines as well as Indian law accept that any price falling within the arm’s length range is acceptable. Indian law specifies the use of the median when the tested transaction falls outside the interquartile range of the comparables; the difference between the median and the 35th-65th percentile range is the focus of much TP litigation.
Practitioners need to be careful about three things when building the arm’s length range. First, the comparables must be genuinely comparable: same type of business, similar functional and risk profile, comparable scale and comparable accounting treatment. Second, comparability adjustments must be made for material differences such as working capital, capacity utilisation and risk differentials. Third, the range must be tested for outliers, and the rationale for including or excluding any borderline case must be documented contemporaneously.
Indian assessing officers have rejected benchmarking studies on all three grounds. The most common rejection reason is the inclusion of comparables that fail a qualitative screen on functional comparability. The second most common is the absence of a working-capital adjustment where the tested party and comparables have materially different receivables and payables profiles. The third is the use of stale comparables that have changed business mix since the year of testing.
Worked Examples — Five Common Intra-Group Transactions
Example 1: Software development services. An Indian subsidiary develops software for its US parent on a cost-plus basis. The most appropriate method is typically TNMM with operating margin on operating cost as the profit level indicator. The Indian comparable set comprises listed Indian IT services companies of similar size and service mix. The arm’s length operating margin for routine software development services in India has historically ranged between 12 and 18 percent on cost, with the exact result depending on the comparability set and the year of testing.
Example 2: Intra-group royalty for trademark use. An Indian distributor pays a royalty to its overseas parent for the right to use the parent’s trademark. The most appropriate method is typically CUP, with comparable third-party trademark license agreements drawn from databases such as RoyaltyRange or ktMINE. The arm’s length royalty rate depends on the industry, the strength of the trademark and the geographic exclusivity granted. Indian benchmarking studies typically produce rates in the 1 to 4 percent of net sales range for consumer goods trademarks.
Example 3: Intra-group loan from parent to Indian subsidiary. The most appropriate method is typically CUP, with the arm’s length interest rate built up from the relevant risk-free rate (sovereign benchmark in the lender’s currency) plus a credit spread that reflects the subsidiary’s standalone credit rating, plus any liquidity premium for the maturity of the loan. The OECD Financial Transactions Guidance (2020) provides the framework, and Indian assessing officers have been increasingly rigorous in testing each component of the build-up.
Example 4: Intra-group guarantee. The Indian parent guarantees a loan taken by its overseas subsidiary from a third-party bank. The arm’s length guarantee fee can be benchmarked using either the yield approach (the savings to the borrower compared with what it would have paid without the guarantee, split between guarantor and borrower) or comparable uncontrolled guarantee fees. Indian Tribunal jurisprudence has consistently supported a guarantee fee in the 0.5 to 1.5 percent of guaranteed amount range, depending on facts.
Example 5: Management fees for headquarters services. The Indian subsidiary pays its overseas parent a management fee for headquarters services such as treasury, legal and HR. The arm’s length test has two limbs: (a) was the service actually rendered and did it confer an economic or commercial benefit on the Indian subsidiary (the benefit test), and (b) was the fee charged at arm’s length (the pricing test). Most disputes are lost on the benefit test rather than the pricing test, so contemporaneous documentation of the services received is more important than the markup chosen.
Documenting the Arm’s Length Position
The arm’s length position is documented in the Local File, supported by the FAR analysis (Functions performed, Assets used and Risks assumed by each party), the benchmarking study, the comparability adjustments and the Form 3CEB attestation. The Master File describes the group-wide TP policy framework. The CbCR reports the country-by-country split of revenue, profit, taxes paid, employees and tangible assets so that tax authorities can flag mismatches between the documented TP positions and the actual outcomes.
The arm’s length principle is not a calculation; it is a position supported by analysis. Documentation is the evidence of that analysis. The best documentation tells a coherent story that links the FAR analysis to the choice of method, the method to the comparables, the comparables to the range, and the tested transaction to its position within that range. The story has to be the same whether the reader is the CFO, the statutory auditor, the assessing officer or the Tribunal.
Common Errors in Arm’s Length Documentation
Six errors account for the majority of audit losses in our practitioner experience. First, method selection without a documented rejection rationale for the alternative methods. Second, comparable set assembled from a database screen without qualitative filtering. Third, missing or skipped working-capital adjustment where the tested party and the comparables have materially different working capital profiles. Fourth, FAR analysis that does not match the intercompany contracts or the actual day-to-day conduct. Fifth, interquartile range computed without consideration of outliers that distort the range materially. Sixth, internal inconsistency between the Local File, the Master File, the CbCR, the intercompany contracts and the audited financial statements.
Avoiding these six errors does not eliminate audit risk, but it shifts the audit conversation from documentation quality to economic substance. That is the conversation in which a partner-led, well-prepared firm has the best chance of an acceptable outcome. Our quality control discipline checks each of these six dimensions before any Local File leaves the firm; documentation that fails on any of the six is sent back for rework before sign-off.
Why the Arm’s Length Principle Will Outlast Pillar Two
The OECD Pillar Two reforms have raised questions about whether the arm’s length principle remains relevant in a world with a 15 percent global minimum tax. The honest answer is yes, and the case is stronger than the question suggests. Pillar Two sets a floor on the effective tax rate for in-scope groups in each jurisdiction; it does not allocate income between jurisdictions. The arm’s length principle is the rule that decides where income arises in the first place, and Pillar Two then tests whether the tax on that income meets the 15 percent threshold. The two operate in sequence, not in competition.
For transfer pricing practitioners, the practical implication is that the Local File, the Master File and the contemporaneous documentation remain the foundation of cross-border tax positions. Pillar Two adds a layer of calculation and reporting on top, with the GloBE Information Return and the QDMTT return drawing data from the same accounting and tax base that already feeds the TP documentation. Groups that have strong arm’s length documentation are well positioned for Pillar Two; groups that have weak documentation will struggle on both fronts simultaneously.
Where the Arm’s Length Principle Goes Next
Three forces will shape the next decade of arm’s length practice. First, Pillar Two implementation will create new data and reporting demands but will not displace the arm’s length rule itself; income still has to be allocated before the minimum tax can be tested. Practitioners should expect the documentation cycle to grow rather than shrink as Pillar Two reporting integrates with existing TP documentation.
Second, intangibles characterisation will continue to be the highest-stakes area. The OECD’s DEMPE framework from BEPS Actions 8-10 has reshaped how value is attributed to the entities that create, enhance and protect intangibles, and the litigation cycle on these issues is still in its early years. Expect more case law, more guidance updates and more documentation focus on the substance of intangibles activity rather than the legal title.
Third, financial transactions will see continued tightening. The 2020 OECD Financial Transactions Guidance set the framework; the next decade will see jurisdictions refining their administrative practice on credit spread benchmarking, guarantee fees and treasury arrangements. Documentation that meets the 2020 standard today may need substantial refresh by 2030.
Innobrant’s engagements are built to adapt to these forces. The same FAR-based documentation framework supports Pillar Two readiness, DEMPE analysis and financial transactions benchmarking; the work expands within the framework rather than being rebuilt for each new requirement.
Frequently Asked Questions
Is the arm’s length principle the same in India and the OECD framework? In substance, yes. Indian transfer pricing law adopts the OECD framework and the same five methods (plus the Other Method). Where Indian law differs is in the use of the arithmetic mean for old years and the median for current years when the tested transaction falls outside the arm’s length range, and in specific Indian procedural provisions such as Form 3CEB attestation, Master File and CbCR notifications.
What happens if my transaction price is outside the arm’s length range? The assessing officer will propose a transfer pricing adjustment to bring the price to the median of the range (under current Indian rules), and the additional income will be taxed in India. Whether the adjustment will hold depends on the strength of the comparability analysis: a sound Local File substantially reduces the risk of adjustment.
How often should benchmarking be refreshed? Every year. Comparable companies move in and out of the database over a 12-month window; financial results change; new comparable peer companies become listed and old ones de-list. Rolling forward unchanged is a common audit trigger and Indian Tribunals have rejected stale comparable sets multiple times.Who at Innobrant signs off on transfer pricing positions? Every Local File is reviewed and signed off by Director, CA Jashwanth Pasupuleti, before filing. The financial consulting team prepares the workpapers and the draft documentation; the partner reads every document before it leaves the firm.