Profit Split Method (PSM) in Transfer Pricing: A Practical Guide

Most Indian transfer pricing files settle on TNMM or CUP because a one-sided comparable can usually be found. PSM is different: it does not ask what an independent party would have charged for a similar transaction, it asks how the combined profit from a genuinely joint effort should be divided between the parties that created it. That makes it the most conceptually demanding of the five Rule 10B methods, and also the one most often misapplied when treated as a default rather than a deliberate choice. This guide builds on our overview of all five transfer pricing methods, which introduces PSM briefly alongside CUP, RPM, CPM, and TNMM.

How PSM Works Under Rule 10B(1)(d)

Rule 10B(1)(d) sets out PSM in four steps:

1. Determine the combined net profit earned by the associated enterprises from the international transaction (or transactions) being tested.
2. Evaluate relative contributions of each enterprise to that combined profit, based on the functions performed, assets employed (including intangibles), and risks assumed by each party, ideally using reliable external market data to benchmark the value of each contribution.
3. Split the combined profit between the enterprises in proportion to their evaluated relative contributions.
4. Arrive at the arm's length price for the tested transaction by taking into account the profit apportioned to the Indian entity.

The rule explicitly allows the combined profit to be split using either a direct contribution analysis or a two-stage residual analysis, discussed below, and permits factoring in market data on how independent enterprises performing comparable functions under comparable circumstances would have divided a similar combined profit.

When PSM Is the Most Appropriate Method

PSM is generally the right choice, rather than a fallback, where:

• Unique, valuable intangibles are transferred or exploited by one or both parties, such as jointly developed technology, proprietary processes, or brand value that has no independent market comparable.
• Operations are highly integrated, so that neither party's contribution can be meaningfully isolated and tested on a standalone basis; for example, a global product where R&D, manufacturing, and marketing intangibles are all held or developed jointly across group entities.
• Multiple related transactions are interlinked such that evaluating any single transaction separately, as CUP, RPM, CPM, or TNMM would require, does not produce a reliable result on its own.

PSM is not appropriate where the transaction involves only routine functions and reliable one-sided comparables exist; in that scenario TNMM, CPM, or RPM will generally give a more objective and defensible answer with less room for dispute over the allocation key.

Contribution Analysis vs Residual (Two-Stage) Analysis

The residual approach is generally preferred where it is feasible, because the first stage anchors part of the analysis to observable market comparables (for example, a contract manufacturer's routine return benchmarked the same way a CPM or TNMM analysis would), leaving only the harder-to-benchmark residual profit to the more judgment-driven allocation key. This reduces the portion of the total profit split that rests purely on subjective weighting.

Choosing an Allocation Key

Whichever variant is used, the credibility of a PSM analysis rests heavily on the allocation key chosen to divide profit (or residual profit) between the parties. Common keys include relative revenue, relative operating costs, relative headcount or compensation of key personnel, relative asset base (including capitalised intangible development spend), or a weighted combination of these, depending on which best reflects the actual economic drivers of value in the transaction. The key should be chosen based on the facts of the business, not selected retroactively to produce a particular allocation, and the reasoning should be documented as thoroughly as the profit computation itself, since tax authorities scrutinise the allocation key at least as closely as the profit pool.

Illustrative Example: Jointly Developed Software Platform

Assume an Indian subsidiary and its overseas parent jointly develop and commercially exploit a software platform: the parent contributed the original core architecture and owns the brand, while the Indian entity built substantial new modules, contributes ongoing R&D, and operates key customer-facing functions. Neither party's standalone contribution can be reliably benchmarked against an independent comparable, because the platform itself is unique.

A residual analysis might first allocate each entity a routine return for functions that can be benchmarked (for instance, a modest cost-plus return to the Indian entity for defined support functions, using CPM or TNMM comparables), then split the substantial remaining residual profit between the two entities using a key such as relative qualified R&D headcount or relative development spend over the relevant period, reflecting each party's actual contribution to the intangible that drives the residual profit. The specific split ratio in any real case depends entirely on the facts and the comparability analysis performed; there is no standard percentage that applies generically.

Documentation Expectations

A PSM study should document the transactions and parties covered, how the combined profit pool was computed and reconciled to financial statements, the functional analysis supporting each party's relative contribution, why a contribution or residual approach was chosen, the allocation key and the reasoning for it, and why PSM was selected over a one-sided method. This reasoning is typically the first thing examined in a transfer pricing audit, and where applicable the transaction still needs to be reported in Form 3CEB alongside the supporting documentation.