TNMM in India: How to Select the Right PLI

The Transactional Net Margin Method (TNMM) is the workhorse of Indian benchmarking studies. It compares a net profitability indicator — the Profit Level Indicator (PLI) — of the tested party with the same PLI of independent comparables, after adjusting for differences that materially affect the margin.

Getting the PLI wrong is the single most common reason an Accept-Reject matrix fails before the Transfer Pricing Officer (TPO). The PLI must isolate the tested party’s routine contribution and be measurable with reliable data from comparable financial statements.

PLI options under TNMM

PLI	Formula	Best suited for
OP/OC	Operating profit ÷ Operating cost	Contract service providers, low-asset-intensity functions
OP/Sales	Operating profit ÷ Operating revenue	Distributors and resellers
ROCE	Operating profit ÷ Capital employed	Capital-intensive manufacturers
ROA	Operating profit ÷ Total assets	Asset-heavy, low-margin operations
Gross margin	Gross profit ÷ Sales	Tested parties with little operating-cost variance

OP/OC (Operating Profit on Operating Cost) is the default choice for most Indian service entities — software, KPO/BPO, R&D and management consulting — because their margin is earned on cost-plus pricing. OP/Sales fits trading and distribution where revenue is the correct base. ROCE/ROA suits entities whose returns are driven by the assets they employ.

Decision rules

1. Follow the function. A contract software developer earns its return on cost; use OP/OC. A distributor earns a margin on turnover; use OP/Sales.
2. Prefer the PLI with the least variance across the comparable pool. A PLI that fluctuates widely between comparable companies is a weak indicator.
3. Use the same PLI for tested party and comparables. A common error is a mixed PLI — OP/OC for the tested party but EBIT margins from the database.
4. Exclude non-operating items. Interest income, extraordinary items and prior-period adjustments distort the PLI and should be normalized.

Step 1: Lock the NIC code before you pull data

Before opening Prowess or Capitaline, you must map the tested party to its 5-digit NIC-2008 code. The wrong code produces a pool that is either too broad or too narrow to support a reliable range.

Worked example: contract software developer

Take an Indian captive that writes code for its overseas parent and earns a cost-plus remuneration. Its FY 2025-26 accounts show:

Item	Amount (INR lakh)
Operating revenue	3,000
Operating cost (incl. employee cost)	2,400
Operating profit	600
OP/OC	25.0%

Four comparable companies screened under NIC 62011 return OP/OC of 14.2%, 19.8%, 27.4% and 33.1%. The arm’s-length range is the inter-quartile range — say the 25th to 75th percentile of 19.8% to 27.4%. The tested party’s 25.0% falls inside the range, so no adjustment is warranted. If it fell outside, you would need to justify why a functional difference explains the deviation before applying a working capital adjustment.

Common pitfalls

• Data-year mismatch. Comparables must reflect the tested party’s financial year; Prowess and Capitaline report calendar-year data for many entities.
• Related-party revenue leakage. Companies with significant related-party revenue distort the PLI and must be rejected with a documented reason.
• Persistent losses. Loss-making entities are often non-comparable unless the tested party itself is loss-making in the same cycle.

Choosing the right method overall

TNMM is rarely appropriate when one party holds unique intangibles or the transaction is too integrated to price separately — that is the Profit Split domain. See the methods guide for a full comparison of CUP, RPM, Cost Plus, TNMM and Profit Split.