OECD Method Selection: Tested Party, PLIs and the Best Method

Chapter 2 of the OECD Guidelines is the working heart of the document: it is where the arm’s length principle from Chapter 1 becomes a price. It lays down the tested-party selection, the profit level indicators (PLIs), the five methods and the best-methods rule. Most of the defensible decisions in a benchmarking study are made in this chapter, and its most-cited anchors — 2.48, 2.50-2.53 and 2.62-2.80 — appear in nearly every method-selection debate.

The three anchors practitioners cite

Anchor	What it fixes
2.48 — tested party selection	The entity whose transaction is least complex and most reliably benchmarkable is the one-sided tested party
2.50-2.53 — PLI list and logic	The list of profit level indicators and the rule that the PLI must reflect the tested party’s function and be reliably measurable from comparable data
2.62-2.80 — the methods set	The five methods (CUP, resale price, cost based, transactional net margin, profit split) and the best-methods rule

2.48 — the tested party is the least complex entity

In paraphrase, the provision directs that the one-sided method should be applied to the entity whose functions, assets and risks are least complex and whose profitability can be measured most reliably against independent comparables. Three consequences follow.

• The selection is functional, not structural. The tested party is chosen on the FAR profile, not on who is the seller or the buyer in the contract.
• The choice is the foundation of the method. A tested party that is not the least complex entity makes the benchmark unreliable no matter how good the comparables set is.
• The routine party earns the routine return. This is the regulatory basis of the routine-return logic that runs through resale price, cost plus and TNMM alike.

2.50-2.53 — the PLI list and selection logic

The provision supplies the list of profit level indicators and the logic for choosing among them. In paraphrase: the PLI must reflect the functions performed, the assets employed and the risks assumed by the tested party, and it must be reliably measurable from the comparable data available. The working mapping of profile to PLI:

Tested party profile	Usual PLI
Limited-risk distributor	Net cost plus (RPM) or gross margin on sales
Contract manufacturer	Cost-based mark-up
Routine service provider	Cost-based mark-up, or a labour-based PLI where costs track labour
Routine operating entity	Operating margin (OP/S or OP/OC) — the TNMM default

The full working list lives in the PLI reference; the selection logic at 2.50-2.53 is why a PLI choice is defensible only when tied to the tested party’s function, and why a PLI the comparables data cannot measure reliably is a method error, not a data problem.

2.62-2.80 — the five methods and the best-methods rule

In paraphrase, the methods set covers five routes — the comparable uncontrolled price, the resale price, the cost based route (cost plus and comparable profits), the transactional net margin method and the profit split — and the best-methods rule requires the method that, under the circumstances, provides the most reliable measure of an arm’s length result on the most reliable data with the least adjustments. The five families each carry their own reference page:

Method family	Reference
CUP	The CUP provisions
Resale price	The resale price provisions
Cost based / comparable profits	The comparable profits and cost-based provisions
Profit split	The profit split provisions
Uncontrolled goods/services and uncontrolled resale/conversion	The uncontrolled comparability routes

The best-methods rule is the provision cited whenever a jurisdiction challenges the chosen method: the answer is not “this method is permitted” but “on this data, this method is the most reliable”.

How Indian practice mirrors it

The reader’s map maps Chapter 2 onto Indian law directly:

OECD concept	Indian provision
Methods (Ch. 2)	s.92A-C (CUP, RPM, CPM, TNMM, PSM)

In practice, the tested-party selection at 2.48 and the best-methods rule at 2.62-2.80 are the same arguments the TPO and the taxpayer make in India: the method must suit the tested party, the PLI must suit the function, and the choice must be the most reliable on the available data.

Where this takes you

• The tested-party decision: tested party selection.
• The method decision: how to choose a method.
• The PLI decision: the PLI reference.
• The data the methods run on: OECD comparability and building the range.