India's transition from a strict arithmetic mean standard to a statistical range concept, introduced through Rule 10CA, changed how a benchmarking study's comparable set translates into an accepted arm's length price. Combined with Rule 10B's provisions on the use of multiple year data, the two rules together determine not just which comparables qualify, but which specific figure within an accepted set a taxpayer can actually rely on.
When the Range Concept Applies
The range concept under Rule 10CA is only available where the comparable set contains six or more entries after the most appropriate method has been applied, and where the method used is one of the price or margin-based methods, CUP, resale price, cost plus, TNMM, or PSM's specified variant, that Rule 10CA lists as eligible. Where the comparable set contains fewer than six entries, the older arithmetic mean plus a permitted variation band continues to apply instead, which makes the size of the final comparable set a threshold determination with real consequences for which computational rule governs the outcome.
Where the Taxpayer Can Land Within the Range
Once the thirty-fifth to sixty-fifth percentile range is established from the qualifying comparable set, a taxpayer's actual transfer price is accepted without adjustment if it falls anywhere within that range, and where it does not, the adjustment is computed by reference to the median of the range rather than the nearest edge, a materially more conservative outcome than simply nudging the price to the closer boundary. This median-based adjustment mechanic is frequently overlooked in preliminary risk modelling, which tends to assume a boundary-based true-up.
The Multiple Year Data Requirement Under Rule 10B
Rule 10B permits, and in the range-concept context effectively requires, the use of data for the tested party's current year together with data for the two preceding years for the comparables used, rather than testing against single-year comparable data alone. The rationale is that a single year's results can be distorted by short-term cyclical or one-off factors, and averaging or otherwise incorporating multiple years smooths out that noise to produce a more reliable range, but it also means the benchmarking study needs multi-year financial data availability for every comparable retained, which is not always straightforward for smaller or less transparent comparable companies.
Current Year Data and the Contemporaneous Documentation Tension
A persistent practical friction is that current-year comparable data is frequently unavailable at the time the taxpayer must prepare contemporaneous documentation, since comparable companies' financial statements for the same year are often not yet filed or publicly available when the taxpayer's own transfer pricing study is due. Rule 10B addresses this by permitting the use of data available up to the date of filing where current year data is genuinely not available at the time of preparing documentation, provided the taxpayer updates the analysis before the return is filed if better data becomes available.
Combining the Two Rules in Practice
In a properly constructed Indian benchmarking study, the search first identifies a qualifying comparable set of at least six entities, then pulls current and two preceding years' data for each, computes the resulting range using the weighted or averaged approach the rules prescribe, and finally tests the tested party's actual margin against that computed range, applying a median-based adjustment only where the price genuinely falls outside it. Skipping the multi-year data step, or applying the range concept to a five-comparable set that does not meet the six-entry threshold, are the two most common technical errors that unravel an otherwise sound study on review.
Conclusion
Rule 10CA's range concept and Rule 10B's multiple year data requirement operate as a single integrated computational mechanism, not two independent rules, and both the threshold entry conditions and the median-based adjustment mechanic materially change the practical outcome of a benchmarking study compared to the older single-year arithmetic mean standard. Groups preparing Indian TP documentation should confirm both threshold conditions are met before assuming the more favourable range-based outcome applies.
