Transfer pricing methods in India: the six methods, with examples

Section 92C(1) of the Income-tax Act, 1961 lists the methods and Rule 10B sets out how each works. Rule 10AB, added from AY 2012-13, allows any other method that gives the price in comparable uncontrolled conditions.

CUP compares the price charged in the controlled transaction with the price in a comparable uncontrolled transaction, adjusted for differences such as volume, credit terms or freight. It is the most direct method, but it needs close product and contract comparability.

Example: an Indian steel company sells hot-rolled coils to its UAE affiliate at ₹62,000 a tonne. In the same month it sells the same grade, in similar quantities and on the same terms, to an unrelated buyer in Kenya at ₹65,000. After adjusting for a ₹500 freight difference, the arm's length price is ₹64,500, and ₹2,500 a tonne is the shortfall.

RPM starts from the price at which the product is resold to an unrelated customer, deducts the normal gross margin a comparable reseller would earn, and deducts expenses such as customs duty on the purchase. What is left is the arm's length purchase price from the group.

Example: an Indian distributor buys watches from its Swiss parent and resells them at ₹10,000. Independent watch distributors earn a 30% gross margin. The arm's length purchase price is ₹10,000 less ₹3,000 = ₹7,000, before customs duty. If the distributor paid ₹7,800, its purchase cost is too high by ₹800 a watch.

RPM suits distributors that do not change the product. It does not work well where the reseller adds substantial value, owns the brand locally or carries out manufacturing.

CPM takes the direct and indirect costs of production and adds the normal gross markup that a comparable independent producer earns.

Example: an Indian unit assembles components for its German parent. Its direct and indirect production cost is ₹80 per unit. Independent contract assemblers earn a gross markup of 20%, so the arm's length price is ₹96. Selling at ₹90 leaves a ₹6 shortfall per unit.

CPM works at the gross profit level, so it needs consistent accounting between you and the comparables on what sits in cost of production. That data is hard to find in public databases, which is one reason TNMM is used more often.

PSM adds up the profit that two or more associated enterprises make from a set of integrated transactions and splits it according to each party's contribution, measured by functions, assets and risks, or by what independent parties would have agreed.

Example: an Indian R&D centre and its Japanese parent jointly develop a medical device, each contributing unique know-how. The combined profit from sales is ₹50 crore. If the Indian centre's share of the relevant R&D effort and capability is 30%, its arm's length share is ₹15 crore.

PSM is used for unique intangibles and deeply integrated operations where no one-sided comparable exists. It needs reliable group-wide profit data and a defensible split key.

TNMM compares the net profit margin the tested party earns from the transaction, relative to cost, sales or assets, with the net margins of comparable independent companies. Net margins are less affected by small product differences than prices or gross margins, and data on them is available in Indian databases, so TNMM is the most used method in India.

Example: a captive software company earns 12% on its operating cost. Comparable independent software service companies earn between 13.2% and 17.4% (the 35th to 65th percentile). Its margin is outside the range and is adjusted to the median. See the full calculation in our arm's length price guide.

The other method allows any method that takes into account the price which has been charged or paid, or would have been charged or paid, for the same or a similar uncontrolled transaction in comparable circumstances. In practice it supports valuation reports, third-party quotations and bids, and approaches for intangibles or business transfers where none of the five standard methods fits.

Rule 10C(1) says the most appropriate method is the one best suited to the facts and that gives the most reliable measure of the arm's length price. Rule 10C(2) lists what to weigh:

• the nature and class of the transaction, and of the associated enterprises
• the functions performed, assets employed and risks assumed by each party
• the availability, coverage and reliability of data on uncontrolled transactions
• the degree of comparability and how reliably differences can be adjusted
• the reliability of the assumptions made

The TP study has to record why the chosen method fits and why each of the others was rejected. A Transfer Pricing Officer who disagrees with the method usually rejects the whole analysis, so this reasoning matters as much as the numbers.

The same six methods continue from tax year 2026-27 under section 165 of the Income-tax Act, 2025, with the arm's length range now in Rule 81 of the Income-tax Rules, 2026. For a method choice and benchmarking on your own transactions, see our transfer pricing services.