Summary: The Profit Split Method (PSM) is one of the six methods recognised under Indian transfer pricing regulations for determining the arm’s length outcome of international transactions. Unlike one-sided methods such as TNMM, PSM considers the combined profits arising from a transaction and allocates them between associated enterprises according to their relative contributions, functions performed, assets employed and risks assumed. It becomes particularly relevant where both parties make unique and valuable contributions, transactions are highly integrated, or valuable intangibles make reliable one-sided benchmarking difficult. PSM can broadly be applied through contribution analysis or residual profit split. Under contribution analysis, the combined operating profit is divided using an appropriate allocation key reflecting relative value creation, such as R&D expenditure, capital employed or another FAR-based measure. Under the residual approach, routine functions are first rewarded using an appropriate benchmarking method such as TNMM or Cost Plus, after which the remaining profit is divided according to the parties’ contributions to the relevant intangibles or other value-driving activities. Indian transfer pricing disputes show that the success of PSM depends less on mathematical calculations and more on whether the method is appropriate for the underlying facts and whether the allocation mechanism is properly supported. Decisions involving Star International Movies Ltd., Synergy Maritime Pvt. Ltd., Toyota Boshoku Automotive India Pvt. Ltd. and PepsiCo India Holdings Pvt. Ltd. illustrate circumstances in which PSM has either been accepted or rejected. Businesses considering PSM should therefore identify integrated value creation early, document functions, risks and intangible-related contributions carefully, establish a defensible allocation key and consistently support the methodology with contemporaneous evidence.
Dividing the Pie Fairly: When and How to Use the Profit Split Method
Transfer pricing, at its heart, deals with a fairly simple question dressed up in complicated language: when two related entities transact with each other (one selling goods, providing services, or licensing technology etc to the other), how do we make sure the price charged between them is fair, and not just a convenient number picked to shift profits into a lower-tax jurisdiction. The answer that tax law across the world has settled on is the Arm’s Length Principle (ALP).
To operationalise this principle, Indian transfer pricing regulations recognise six methods, one of which is Profit Split Method (PSM). PSM comes up often in conversation, but its real-world application remains comparatively rare, largely because it is called for only in a fairly narrow set of situations. Let’s see how this works, how the split is worked out on the ground, and where PSM works and where it does not.
What Is PSM, and Where It Fits
Most methods, including the widely used TNMM, work by testing one side of a transaction. You pick the entity doing the relatively simpler or routine job, such as contract manufacturer, low risk distributers, back-office support service providers, benchmark its margin against independent comparable, and move on. This one-sided testing holds up well for exactly as long as that assumption holds, one party genuinely routine, the other genuinely carrying the value and the risk. The method starts to strain the moment both sides are doing something more than routine, and it is no longer obvious who should even be picked as the tested party.
PSM is built for exactly that gap. Instead of looking at one party in isolation, it looks at the transaction as a whole. It aggregates the combined profit from the transaction and divides it between the parties based on their relative contribution – the functions performed, the assets employed (intangibles in particular), and the risks assumed. It asks, in other words, how a jointly created outcome should be fairly shared between the parties who created it together.
This is precisely why PSM tends to surface in situations that are, by nature, deeply intertwined. In today’s business models, especially where group entities jointly develop valuable intangibles, drive product innovation together, or share strategic decision-making across borders, tax authorities have started pushing back on the blanket use of TNMM. Their argument is fairly intuitive: when value is genuinely being created on both sides of a transaction, testing the profitability of only one side simply does not tell the full story. This is where the PSM steps in and evaluates who is contributing what, and whether each party is being compensated fairly for it, earning too little, too much, or just about right.
Two Ways to Split the Pie
Once it is established that PSM is the right method for a given transaction, the next question is how the split is actually determined, and there are two recognised approaches to doing this.
Contribution Analysis (Combined Profit Split)
The combined operating profit is divided based on a FAR-driven allocation key – relative cost, headcount, capital employed, or some other measure that genuinely tracks value creation. The key is rarely a desk exercise; it usually needs a structured attribution study built with real input from the business, not a number reverse-engineered to look defensible. In essence, to arrive at the arm’s length profit, there are four steps:
- Determine the combined net profit of the international transaction
- Evaluate each party’s relative contribution through a functional analysis and on the basis of reliable external market data (e.g., 30:70)
- Split the combined profit in that proportion
- Treat the profit so apportioned as the arm’s length outcome.
Let’s understand this through an example. Say USA Inc. designs and manufactures the key component, India Ltd. manufactures the final product, and XYZ Ltd. distributes it. Assume the combined net profit attributable to USA Inc. and India Ltd. is 100.
| Steps | Working |
|---|---|
| Step 1: Determine the combined net profit from the transaction | USA Inc. + India Ltd. combined net profit = 100 |
| Step 2: Evaluate each party’s relative contribution via FAR | Let’s say no reliable CUP exists for either side’s tech; relative R&D spend is accepted as a fair proxy –
– USA Inc.: 30 – India Ltd.: 10 (combined 40) |
| Step 3: Split the combined profit in that proportion | USA Inc. = 100 × 30/40 = 75
India Ltd. = 100 × 10/40 = 25 |
| Step 4: Apportioned profit = arm’s length outcome | USA Inc.’s ALP profit = 75
India Ltd.’s ALP profit = 25 |
Residual Profit Split
Used where one party performs routine functions while the other owns the intangible that actually drives the business. Step one gives each party a routine return (typically via TNMM or Cost Plus). Step two attributes whatever profit is left over – the residual – to the intangible, and splits it based on relative contribution to that intangible. In essence, to arrive at the arm’s length profit, there will actually be five steps:
- Step 1: Gives each party/ deserving party a routine return (typically via TNMM or Cost Plus)
- Step 2: Determine the residual combined net profit of the international transaction (after deduction of profit allocated in step 1).
- Step 3: Evaluate each party’s relative contribution through a functional analysis and on the basis of reliable external market data (e.g., 30:70)
- Step 4: Split the residual profit in that proportion
- Step 5: Treat the profit so apportioned as the arm’s length outcome.
Let’s understand this again through the above example with slight changes. Say USA Inc. designs and makes a unique formula for pharma products. India Ltd. buys it from USA Inc. and designs and manufactures the rest of the product. XYZ Ltd. distributes it, and XYZ Ltd.’s margin is already confirmed to be at arm’s length via RPM, so it is out of the picture.
The real question is splitting what is left between USA Inc. and India Ltd., since both hold something genuinely unique (USA Inc.’s intangible and India Ltd.’s own R&D), and there is no comparable price for the USA Inc.- India Ltd. leg. Assume the combined net profit attributable to USA Inc. and India Ltd. is 100.
| Steps | Working |
|---|---|
| Step 1: Reward the routine function first, using an external benchmark | Let’s say manufacturing cost of USA Inc. is 100, and India Ltd. is 200, and 10% is the routine arm’s length return for the manufacturing function.
USA Inc. = 10 (10% of 100) India Ltd. = 20 (10% of 200) Total routine return allocated = 30 |
| Step 2: Determine the combined profit left over after the routine reward (Residual profit) | 100 – 30 = 70 (residual) |
| Step 3: Evaluate relative contribution to the residual via FAR | Same R&D proxy –
USA Inc.: 30 India Ltd.: 10 (combined 40) |
| Step 4: Split the residual in that proportion | USA Inc. = 70 × 30/40 = 52.5
India Ltd. = 70 × 10/40 = 17.5 |
| Step 5: Add back the routine reward = arm’s length outcome | USA Inc. = 10 + 52.5 = 62.5
India Ltd. = 20 + 17.5 = 37.5 |
The arithmetic is never where these cases are actually contested. The dispute is almost always over the two judgment calls baked into it was the routine return in Step 1 correctly benchmarked and does the Step 3 allocation key genuinely track who created the residual. This is precisely the fact pattern that recurs in Indian TP litigation: a captive unit set up for routine service work gradually starts co-developing IP, filing joint patents, influencing product design while still being paid a flat cost-plus margin as though nothing had changed.
What the Tax Authorities Have Actually Said
PSM disputes rarely turn on the arithmetic; they turn on whether the method was even the right one to reach for and whether it has been correctly benchmarked. A few cases are listed below to briefly explain the PSM applicability, some accept PSM and some reject it, depending on the facts of the case.
| Case | PSM Outcome | Core Reason |
|---|---|---|
| Star International Movies Ltd.
[TS-1003-ITAT-2019(Mum)-TP] |
Combined PSM upheld | Assessee and its channel-owning AEs had already adopted PSM by consolidating third-party revenue and cost; ITAT refused to let the AO carve out and separately tax a slice of the same pool. |
| Synergy Maritime Pvt. Ltd.
[TS-545-ITAT-2019(CHNY)-TP] |
PSM rejected | TPO built the split from Sec. 44C’s head-office expense cap, not Rule 10B(1)(d); no unique intangibles or interrelated transactions shown. Internal CUP upheld instead. |
| Toyota Boshoku Automotive India Pvt. Ltd.
[TS-452-ITAT-2022(Bang)-TP] |
PSM rejected | Assessee only used AE’s technology; contributed no unique intangible of its own. Aggregating royalty with manufacturing for TNMM doesn’t make the two so interrelated that PSM is needed. TNMM restored. |
| PepsiCo India Holdings Pvt. Ltd.
[TS-1250-ITAT-2018(DEL)-TP] |
PSM rejected | TPO “neither applied PSM correctly nor analysed the contribution… on the relative value of FAR”; DRP’s fallback was an “other method” that was really a disguised Bright Line Test. |
If we analyse such orders together, the pattern is fairly consistent. PSM survives scrutiny where both parties genuinely contribute something unique, where the transactions are interlinked enough that one-sided testing would be unreliable, and where the split ratio is backed by an actual, evidenced FAR exercise not a statutory expense cap, not a Bright Line Test wearing a different name, and not the assumption that aggregating transactions for TNMM automatically makes them ‘interrelated’ for PSM.
From Analysis to Action
- Start early: Identify potential PSM situations during the business and transaction structuring stage rather than waiting for a TP audit.
- Map value creation: Document how functions, decision-making, risks and intangible-related activities are actually carried out across entities.
- Build the allocation key carefully: The chosen factor should have a clear connection with the value contributed by each party and should be capable of being supported with contemporaneous evidence.
- Keep the methodology consistent: The basis used to determine the profit pool and allocation should be capable of being applied consistently across years, unless there is a genuine change in facts or circumstances.
- Stress-test the approach: Before adopting PSM, consider whether the outcome remains reasonable under different assumptions, allocation keys or business scenarios.
- Think beyond the TP file: A well-supported PSM should be capable of being explained consistently across tax, finance and business teams, particularly where the arrangement spans multiple jurisdictions.
PSM is not, and was never meant to be, an everyday tool. But where the facts call for it, applying it well, and applying it early, can make all the difference.
