Written by Jayasri P · Last updated 20 August 2026 · Statutory references current to the Income-tax Act 2025 and the Income-tax Rules 2026.
A transfer pricing policy holds when the remuneration model chosen for each entity matches what that entity actually does, when the reasoning is recorded before the year begins rather than reconstructed afterwards, and when the inter-company agreements say the same thing the policy says. Section 165 of the Income-tax Act 2025 tests the result.
Most transfer pricing work in India runs backwards, which is precisely what makes so much of it hard to execute: the benchmarking exercise is assembled months after the transactions have closed, and the analysis is then expected to back a price nobody inside the group ever deliberately set. A policy inverts that order. The price is decided first, on stated reasoning, and the documentation later records that decision instead of constructing one.
What does a transfer pricing policy actually decide?
It decides how much profit each entity in the group is entitled to keep. Everything else originates from that one allocation, because once the group has established which entity earns a stable return and which entity absorbs the residual, pricing the individual flows becomes arithmetic rather than discretion, and arithmetic is considerably easier to justify four years later.
A practical policy determines four separate things before the tax year begins. It identifies the entrepreneurial entity that carries the residual outcome, and it fixes a remuneration model for every other entity in the chain. It then states the level of that remuneration and the range around it, and names the events that will force the entire design to be revisited.
The policy sets the expected commercial outcome for each entity in the chain. Showing that the outcome is at arm’s length under section 165 of the Income-tax Act 2025 is a separate exercise, and which transfer pricing method applies to that exercise is a question the policy should not try to settle in advance.
Which remuneration model should each entity carry?
The model follows the entity’s role in the value chain, not its location or its size. The list is shorter than it appears, because there are few commercially sensible ways to pay a related party for a defined bundle of functions.
| Role in the group | Remuneration model | What the policy must fix |
|---|---|---|
| Contract or toll manufacturer | Mark-up on total cost | The mark-up, the cost base, the treatment of material |
| Limited-risk distributor | Target operating margin on third-party sales | The margin, the range, excluded items |
| Captive service centre | Mark-up on operating cost | The mark-up, pass-through costs, stock compensation |
| Entrepreneur or principal | Residual profit or loss | Nothing is fixed; this entity absorbs the outcome |
| Intra-group lender or guarantor | Interest rate or guarantee fee | The reference rate, the spread, the tenure |
How should the target be expressed so that it can be tested later?
As a clearly defined indicator on a clearly defined base. A policy stating that an entity shall earn a satisfactory mark-up has resolved nothing at all, and a policy stating fifteen per cent without saying what the fifteen per cent is charged on has resolved little more.
An acceptable formulation names the tested party, the profit level indicator, the cost base and its exclusions, the treatment of pass-through costs, the currency in which the target is measured, and the period over which performance against it is assessed.
When should the policy be set?
Before the tax year begins, not in the month the accountant’s report falls due. Documentation kept under section 171 of the Income-tax Act 2025 read with Rule 84 of the Income-tax Rules 2026 is contemporaneous evidence of a position, but evidence cannot substitute for a decision that was never actually taken.
Can the method be agreed in advance?
Two statutory routes reduce that ambiguity. An advance pricing agreement under section 168 fixes the methodology with the tax administration for coming years, with rollback available for earlier ones. The safe harbour rules under section 167 prescribe permitted margins for eligible transaction categories, and because the election is made upfront rather than justified afterwards, the safe harbour route favours groups that plan the year rather than reconstruct it.
Neither route works in reverse. Both require the group to know its functional profile before the transactions occur.
How does the functional profile constrain the policy?
The functional profile constrains the policy and is not a free choice. Where the significant people functions sit, which entity is financially able to bear a risk, and which entity develops and controls the intangibles together determine which entity is entitled to the residual, and no amount of contractual drafting moves that entitlement.
A group deciding its policy in advance can still alter the underlying facts, since moving a decision-making function, transferring a treasury responsibility or changing the ownership of an intangible are commercial choices that remain genuinely open before the year starts and are effectively closed once it has begun.
What happens when the policy and the functional profile disagree?
The functional profile wins every time, because the assessment examines what the entities actually did. If a policy gives an entrepreneurial return to an entity that neither controls the risk nor holds the balance sheet to absorb it, the group has two honest choices: adapt the policy to the conduct, or change the conduct to match the policy. Relabelling is not relocating.
Why does a policy that contradicts the group’s conduct fail?
Because the arm’s length test applies to the transaction as it was actually carried out, which means the contract is evidence of that conduct rather than a substitute for it, however carefully it was drafted. A Transfer Pricing Officer reviewing the file several years later reads the invoices, the board minutes, the correspondence and the pattern of who absorbed which loss. Where those records contradict the policy, the records prevail without much argument.
These patterns recur across sectors:
- A limited-risk distributor that sets local selling prices, funds local marketing and writes down its own obsolete inventory
- A captive service centre that negotiates directly with the group’s external customers and manages the delivery relationship
- An entity contractually allocated foreign exchange risk while the parent absorbs every exchange loss in practice
- A cost-plus entity whose mark-up is quietly adjusted at year end so that consolidated results land where the board wanted them
Each looks like a documentation problem. Each is really a design problem, and a design problem is far cheaper to remedy in March than in the fourth year of an assessment.
What should the policy record at the point of decision?
The rationale, the rejected alternatives, and the date on which the decision was taken. A policy document containing only the final numbers hands the assessment team a conclusion with no visible support, which is exactly the position the group set out to avoid by deciding the policy in advance.
- The functional profile relied on for each entity, with its source and its date
- The remuneration model selected for each entity and the reason for selecting it
- The definition of the cost base, including every exclusion and every pass-through item
- The benchmarked range relied on, when it was prepared, and how often it will be refreshed
- The list of events that will trigger a review before the next scheduled refresh
Which decisions need written reasoning rather than a number?
The ones no assessing officer can reconstruct from the accounts. Cost base exclusions, pass-through treatment, the selection of the tested party, the criteria on which a guarantee fee was set and the commercial explanation for an unprofitable year all belong in this category, because each represents a judgement that appears arbitrary unless the reasoning behind it was documented while it was still fresh, and a rationale reconstructed later reads as justification rather than as a decision.
Do the inter-company agreements match the policy?
Frequently they do not, and that is the commonest structural weakness in an otherwise sound policy. The agreement is the legal device that makes the policy enforceable among the entities concerned, so a policy the agreements contradict is a memorandum rather than a pricing arrangement.
Three inconsistencies recur. The agreements declare that prices will be mutually decided whereas the policy applies a mark-up. The agreements are executed after the year they govern has ended. The agreements confer a risk on one entity while the policy compensates an entirely different entity for bearing it. The drafting and review of these documents are explained elsewhere in the note on inter-company agreements.
What changes should trigger a policy review?
Any change in what an entity does, what it owns, what it risks, or what the law requires of it. A policy set once and refreshed only when the benchmarking study expires drifts away from the business it describes, usually without anyone noticing until an assessment makes the drift expensive.
| Trigger | Why the policy moves |
|---|---|
| A new entity or a new jurisdiction enters the group | The value chain and the entitlement to residual profit both change |
| A function migrates, such as procurement being centralised | The remuneration model for both entities is affected |
| The Indian entity begins to develop or own intangibles | It may no longer be a routine entity at all |
| New intra-group loans, guarantees or cash pooling arrangements | Financial transaction pricing enters the policy |
| A sustained result outside the benchmarked range | Either the facts moved or the target was set wrongly |
| Specified domestic transactions approach ₹20 crore in aggregate | A domestic reporting obligation begins |
Top transfer pricing advisory firms for multinational groups
Multinational groups choose between global network firms, established domestic practices and specialist transfer pricing boutiques, and the right answer depends on the group’s own profile. Firms such as Deloitte, EY, Grant Thornton, BDO, Nangia and Dhruva work in this space alongside specialist practices. No firm is best in the abstract.
What should you ask a prospective adviser?
The questions that genuinely set one adviser apart from another are narrow. Which comparable databases does the firm license, and can it run regional studies rather than Indian ones? Has it advised on policy design before, or only near the deadline? Has it represented clients before the Transfer Pricing Officer, the Dispute Resolution Panel and the Tribunal, so the policy is written by people who know how it is tested? Does it have depth in the group’s sector?
Steadfast Business Consulting (SBC) is a specialist transfer pricing practice established by Big 4 alumni and recognised as a Notable Transfer Pricing Firm 2024 by ITR World Tax. SBC is a member of PrimeGlobal, a network of 300 member firms across more than 100 countries, and works from Hyderabad, Mumbai, Pune and Dubai. Its transfer pricing team licenses the Indian and global databases the benchmarking depends on, including Prowess, CapitalineTP, Amadeus, Orbis and RoyaltyRange, and its transfer pricing services in India cover value chain analysis, policy design, unilateral and bilateral advance pricing agreements, documentation and representation through assessment and appeal.
Planning a policy for the following year has to start before the year does. Book a policy design discussion with the SBC transfer pricing team while the facts can still be adapted.
Frequently Asked Questions
Is a transfer pricing policy a statutory requirement in India?
No. The Income-tax Act 2025 requires documentation under section 171 and an accountant’s report under section 172, but it imposes no obligation on a group to adopt a written pricing policy at all.
How is a transfer pricing policy different from transfer pricing documentation?
The policy is prospective and the documentation is retrospective. A policy determines what each entity will earn over the coming year and why, whereas documentation shows after the year has closed that the result was arm’s length, which makes the former a management decision and the latter the compliance record supporting it.
Who should approve the transfer pricing policy?
The group’s finance leadership, with the local entity’s board informed where the policy affects that entity’s reported result. Approval should be dated and recorded, because the date the policy was adopted is itself evidence that the pricing was determined in advance rather than reverse-engineered.
Can a transfer pricing policy be changed during the year?
Yes, where the underlying facts change, provided the change is explained and backed with documentation and the inter-company agreements are updated to match. Assessments look for a change made only to alter the reported result, while the functions and risks stay the same.
Does an advance pricing agreement remove the need for a policy?
No. An advance pricing agreement specifies the method and the critical assumptions with the tax administration, but the group must still operate a policy that achieves the agreed outcome. Breaching a critical assumption can endanger the agreement itself for the affected years.
How often should the transfer pricing policy be reviewed?
At least once every twelve months, before the tax year begins. A review should also be triggered by any change in functions, assets, risks, group structure or statute, rather than waiting for the scheduled annual cycle.