TNMM comparables require functional and economic similarity; accounting-year differences and non-persistent losses alone do not justify exclusion.
Transactional net margin method comparables for captive data-processing and software-development services must be functionally and economically similar. Rule 10B(4) permits use of data from the two preceding years; a company is not a persistent loss-maker merely because it incurred losses in some relevant years, and a differing accounting year alone does not require exclusion where quarterly data can be extrapolated. Companies with materially different business models, substantial outsourcing, high-end services, software products, diversified licensing activities, significant brands or intangibles, entrepreneurial scale, or unreliable segmental data should be excluded. Repair and maintenance provisions require verification of supporting evidence, subsequent crystallisation and reversal, consistent with DRP directions.
Issues: (i) Whether the selected companies were comparable for determining the arm's-length price of captive data-processing and software-development services under the transactional net margin method for assessment years 2009-10 and 2010-11. (ii) Whether the disallowance of repair and maintenance expenditure for assessment year 2009-10 required fresh verification.
Issue (i): Whether the selected companies were comparable for determining the arm's-length price of captive data-processing and software-development services under the transactional net margin method for assessment years 2009-10 and 2010-11.
Analysis: Rule 10B(4) of the Income-tax Rules, 1962 permits consideration of data for the two preceding years. Companies having a profit in one of the relevant three years could not be treated as persistent loss-makers solely because they incurred losses in the other years. A different accounting year was also not, by itself, a ground to exclude a functionally similar company where relevant quarterly data could be extrapolated.
Analysis: Companies following materially different business models, including substantial outsourcing, high-end healthcare services, software-product development, diversified product and licensing operations, or having no reliable segmental information, were unsuitable comparables. Large entrepreneurial companies possessing substantial brands, intangibles, scale and end-to-end business operations were not comparable to the assessee's captive, limited-risk service operations. The export-revenue filter also warranted exclusion where it was not met.
Conclusion: The inclusion of KPIT Cummins Global Business Solutions Ltd., NIIT Smartserve Ltd., Allsec Technologies Ltd. and R-Systems International Ltd. was directed where applicable, while the functionally dissimilar or otherwise unsuitable companies identified for exclusion in the data-processing and software-development segments were directed to be excluded. The issue was decided in favour of the assessee.
Issue (ii): Whether the disallowance of repair and maintenance expenditure for assessment year 2009-10 required fresh verification.
Analysis: The claim relating to payment for elevator maintenance required examination in accordance with the Dispute Resolution Panel's directions. The year-end provision was supported by material showing subsequent crystallisation of a substantial part of the provision and reversal of the balance; its allowability was therefore dependent on verification of the evidence, with an opportunity of hearing.
Conclusion: The repair and maintenance claim was restored for verification and fresh consideration in accordance with law. The issue was decided in favour of the assessee for statistical purposes.
Final Conclusion: The arm's-length-price computation must be recomputed after giving effect to the directed changes in the comparable set, and the repair and maintenance claim requires fresh verification.
Ratio Decidendi: Under the transactional net margin method, comparables must be functionally and economically similar; a different accounting year or non-persistent losses alone do not justify exclusion, whereas material differences in functions, business model, scale, intangibles or availability of segmental data do.