Transfer pricing comparability filters must be applied at the outset, and oversized comparables may be excluded for turnover mismatch.

Transfer pricing comparability filters must be applied at the initial search stage, not introduced after the comparable set has been finalised; a later R&D filter was therefore rejected as inconsistent with the search process. Turnover is also a relevant comparability factor under the transactional net margin method, because scale, market position, economies of scale, assets and risks can materially affect margins. Comparables with substantially higher turnover than a captive service provider were directed to be excluded, and an upper turnover filter of 10 times the tested party's turnover was accepted for recomputation, resulting in partial relief.

---

---

Issues: (i) Whether the transfer pricing search process was vitiated by applying the R&D filter only after selection of the initial set of comparables. (ii) Whether comparables having significantly higher turnover than the tested party were liable to be excluded, and whether an upper turnover filter of 10 times the assessee's turnover should be applied.

Issue (i): Whether the transfer pricing search process was vitiated by applying the R&D filter only after selection of the initial set of comparables.

Analysis: The Tribunal held that once the search process had been completed and 24 comparables identified, a fresh filter could not be introduced at the later stage to alter the final set of comparables. A filter intended to be applied for comparability must form part of the initial search criteria itself; otherwise the exercise becomes inconsistent and the comparability analysis is undermined. The principle against selective application of filters after completion of the search process was accepted.

Conclusion: The later application of the R&D filter was not upheld as a valid method of comparability selection.

Issue (ii): Whether comparables having significantly higher turnover than the tested party were liable to be excluded, and whether an upper turnover filter of 10 times the assessee's turnover should be applied.

Analysis: The Tribunal treated turnover as a relevant comparability factor under the transactional net margin method because scale, market position, economies of scale, assets and risks materially affect margins. It held that companies with substantially higher turnover could not be compared on a like-to-like basis with a much smaller captive service provider. On the facts, the selected comparables had turnovers far exceeding the assessee's turnover of about Rs. 114.15 crores. The Tribunal therefore directed exclusion of the very large comparables and accepted the application of an upper turnover filter of 10 times the assessee's turnover for recomputation.

Conclusion: The higher-turnover comparables were directed to be excluded and the turnover filter plea was accepted in substance.

Final Conclusion: The transfer pricing adjustment was required to be recomputed after excluding the identified high-turnover comparables, resulting in partial relief to the assessee.

Ratio Decidendi: Filters in transfer pricing comparability must be applied at the proper stage of the search process, and entities with materially different scale of operations and turnover are not reliable comparables under the transactional net margin method.

---