‘Turnover Filter’ in ‘Comparability Analysis’ for Benchmarking

The term “arm’s length price (ALP)” is defined u/s. 92F(ii) as a price which is applied or proposed to be applied in a transaction between persons other than associated enterprises, in uncontrolled conditions. The application of ALP is generally based on a comparison of the price, margin or profits from particular controlled transaction with the price, margin or profit from comparable transactions between independent enterprises. A comparison of transaction between the associated (related) enterprises (known as controlled transaction) with transaction between independent enterprises (known as uncontrolled transaction) is referred to as ‘comparability analysis’, which is at the heart of the application of the principle of ALP.

Rule 10B of the Rules provides for ‘comparability analysis’ wherein a comparison of a controlled transaction is undertaken with uncontrolled transaction. The controlled and uncontrolled transaction are comparable if none of the differences between the transactions could materially affect the factor viz, price, cost charged, profit arising, etc, being examined in the methodology, or if reasonably accurate adjustments can be made to eliminate the material effects of such differences. In order to establish the degree of actual comparability and then to make appropriate adjustments to establish arm’s length conditions, it is necessary to compare the attributes of the transactions or enterprises that could affect conditions in arm’s length transactions. Some of the main attributes or comparability factors are as under:

• Characteristics of the property or services transferred;

• Functions performed by the parties (taking into account assets used and risk assumed);

• Contractual terms;

• Economic circumstances of the parties;

• Business strategies by the parties, etc.

While transfer pricing is not an exact science and is itself at a nascent and developing stage in India, therefore, there are bound to be controversies on various aspects of transfer pricing provisions. One of the recent controversies has been in respect of one of these attributes of the comparability analysis between the controlled and uncontrolled transaction, i.e. the relevance of ‘turnover filter’ in comparability analysis for determination of ALP.

‘Turnover filter’ in comparability analysis refers to filtration/truncation of the selected comparables vis-à-vis the company, on the basis of turnover, because the difference in turnover may affect the determination of ALP of the transaction. For instance, Company A with a turnover of Rs. 1,000 crore plus could not be considered as a comparable to Company B who has a turnover of Rs. 1 crore, since Company A shall have higher bargaining power, capacity to execute large contracts, risks assumed, skilled staff, etc vis-à-vis Company B who may not be able to undertake similar transactions. So, for determination of ALP of controlled transaction undertaken by Company B, whether ‘turnover filter’ can be applied in comparability analysis of service companies is the issue.

While the Hyderabad, Delhi and Bangalore benches of the Income-tax Appellate Tribunal have taken a stand in favour of the taxpayers allowing ‘Turnover Filter’ in comparability analysis of service companies, the Mumbai bench of the Tribunal has recently taken a contrary view on the subject.

Capgemini India’s case
In Capgemini India Pvt. vs. ACIT (ITA No. 7861/M/2011) (Mum.) dated 28th February 2013, the Appellant had rendered software programming services to its parent Company in US. The Appellant had applied Transactional Net Margin Method (‘TNMM’) as the most appropriate method for benchmarking this international transaction, which was duly accepted by the AO/TPO. Among the various disputes w.r.t. comparability analysis undertaken by the Appellant Company, the AO/TPO had denied the exclusion of comparables viz, Infosys and Wipro, and thereby refused the applicability of ‘turnover filter’.

It was argued before the Tribunal that even though the Appellant Company had considered these companies in its benchmarking exercise, however, the correct and right approach was to exclude such high turnover companies with turnover exceeding Rs. 13,000 crore, whereas the turnover of the Appellant was only around Rs. 558 crore. It was contended that these comparable companies enjoyed economies of scale and better bargaining power vis-à-vis the Appellant Company. Relying on the financials of these comparable companies, it was specifically submitted that the margins of these companies were exceptionally high vis-a-vis the Appellant Company and therefore, these comparables should have been excluded in the benchmarking exercise. Reliance was placed on the following decisions by the Appellant Company to contend that ‘turnover filter’ should be applied and the comparables viz, Infosys and Wipro should be excluded from the benchmarking exercise for want of high turnovers:

• Addl CIT vs. Frost and Sullivan India Pvt. Ltd. (2012) (50 SOT 517)(Mum);

• Dy CIT vs. Deloitte Consulting India (P) Ltd.(2011) (61 DTR 101)(Hyd)(Tri);

• Aginity India Technologies vs. ITO (ITA No. 3856/ Del/2010)(Del)(Tri);

• Genesis Integrating Systems India P. Ltd vs. Dy CIT (2011) (61 DTR 225)(Bang); and

• Brigade Global Services Pvt. Ltd vs. ITO (ITA No. 1494/Hyd/ 2010)(Hyd.)(Tri)

On the other hand, the Department argued that these comparables should not be excluded even though they have exceptionally high turnover and profit. It was argued that economies of scale is not relevant and applicable in case of service companies and the ‘turnover filter’ is relevant only in case of manufacturing companies.

Reliance was placed on the decision of Symantec Software Services Pvt. Ltd (ITA No. 7894/M/2010) [2011-TII-60-Mum-TP], in which the Tribunal had upheld the non-applicability of turnover filter in case of service companies. Further, reliance was also placed on the chart plotted with margin and turnover of the comparables, which concluded that there was no linear relationship between them.

The Tribunal held that turnover filters cannot be applied in case of service companies, since they do not have any high fixed costs and the employees are the only main assets, whose costs are directly related to manpower utilised. Relying on the chart produced by the Department, the Tribunal held that there was no linear relationship between margin and turnover and so the concept of economies of scale does not apply in case of service industry. As regard the contention of the Appellant w.r.t. skilled employees available with the comparables, the Tribunal held that margins of the comparables and the Appellant were not affected on account of such differences and all the companies and comparables had same level of risk as they operated in same field and similar environment. Referring to Rule 10B(2), the Tribunal observed functions performed, asset used and risks assumed [‘FAR’] by the comparable companies should be compared with the Appellant Company in the benchmarking exercise of the international transaction.

As regards the argument of the Appellant Company w.r.t. low bargaining power, it was held by the Tribunal that since the Appellant is a part of multinational group therefore, it cannot be said to have less bargaining power. The Tribunal therefore upheld the contention of the Department, that no turnover filter can be applied in case of service oriented companies.

A similar view has been taken by the Tribunal in the following cases, rejecting the use of turnover filter for comparability analysis of service companies:

•    Vodafone India Services P. Ltd vs. DCIT (ITA No. 7140/M/2012) dated 26th April 2013; and

•    Willis Processing Services India P. Ltd(ITA No. 4547/M/2012);

Genisys Integrating Systems case

In Genisys Integrating Systems India (P) Ltd vs. DCIT (64 DTR 225), the Bangalore Tribunal was opining on the determination of ALP of software development services provided by the Appellant Company to its AEs outside India. TNMM method which was selected as the most appropriate method for determination of ALP was accepted by the AO/ TPO. On the dispute of turnover filter with a range of Rs. 1 crore at the lower end and Rs. 200 crore at the high end, applied during the course of determination of ALP, the Tribunal upheld the following arguments of the Appellant Company:

•    Enterprise level difference is an important facet in determination of ALP. Comparables should have something similar or equivalent and should possess same or almost the same characteristics;

•    A Maruti 800 car cannot be compared to Benz car, even though both are cars only. Unusual pattern, stray cases, wide disparities have to eliminated as they do not satisfy the test of comparability;

•    Companies operating on a large scale benefit from economies of scale, higher risk taking capabilities, robust delivery and business models as opposed to the smaller or medium sizes companies and therefore, size matters;

•    Two companies of dissimilar size therefore, cannot be assumed to earn comparable margins and this impact of difference in size could be removed by a quantitative adjustment to the margins or prices being compared if it is possible to do so reasonably accurately;

•    Reliance was placed on the following decisions, wherein turnover/ quantitative filter was approved for determination of ALP:
–    Dy CIT vs. Quark Systems (P) Ltd (2010)(38 SOT 307)(Chd)(SB);
–    E-Gain Communication (P) Ltd vs. Dy. CIT (2008) (13 DTR 65)(Pune)(Tri);
–    Sony India (P) Ltd vs. Dy CIT (114 ITD 448)(Del);
–    Dy. CIT vs. Indo American Jewellery Ltd. (2010) (40 DTR 386)(Mum)(Tri);
–    Philips Software Centre (P) Ltd vs. Asst. CIT (119 TTJ 721)(Bang.); and
–    Asst. CIT vs. NIT (2011)(57 DTR 334)(Del)(Tri)

•    Further, reliance was placed on the relevant ex-tracts of Para 3.43 of the OECD Transfer Pricing Guidelines, which are as under:

“Size criteria in terms of Sales, Assets or Number of employees. The size of the transaction in absolute value or in proportion to the activities of the parties might affect the relative competitive positions of the buyer and seller and therefore comparability.”

•    NASSCOM also has categorized companies based on turnover, similar to Dun and Bradstreet.

The Tribunal specifically observed that there has to be lower limit and upper limit of range in applying turnover filter, since size matters in business. A big company would be in a position to bargain the price and also attract more customers. It would also have a broad base of skilled employees who are able to give better output. A small company may not have these benefits and therefore, the turnover also would come down reducing profit margin.

The Tribunal therefore approved the use of turn-over filters in comparability analysis of a services company.

A similar view has been taken by the Tribunal, approving the use of turnover filter in comparability analysis of service companies:

•    Adaptec (India) (P) Ltd vs. DCIT (2013)(86 DTR 26)(Hyd.)(Tri);
•    Asst CIT vs. Maersk Global Services Centre (India) P. Ltd. (133 ITD 543)(Mum.);
•    M/s. Patni Telecom Solutions vs. ACIT (1846/ Hyd/2012) dated 25 April 2013;
•    Capital IQ Information Systems vs. Dy. CIT (ITA No. 1961/Hyd/2007);
•    Brigade Global Services (P) Ltd vs. ITO (supra);
•    Triniti Advanced Software Labs (P) Ltd vs. Asst. CIT (2011 TII 92 Tri Hyd-78);
•    Agnity India Technologies (P.) Ltd vs. Asst CIT (supra);
•    Addl CIT vs. Frost and Sullivan India (P) Ltd (supra);
•    Actis Advisors Pvt Ltd vs. DCIT (2012)(20 ITR 138) (Del.)(Tri.);
•    Continuous Computing India (P) Ltd. vs. ITO (2012) (52 SOT 45)(Bang)(URO); and
•    Centillium India P. Ltd vs. DCIT (2012)(20 ITR 69) (Bang)(Tri.)

Observations

On perusal of the contrary decisions discussed above, in all the cases, TNMM was selected and applied as the most appropriate method for bench-marking. TNMM puts more efforts on functional similarities than on product similarities. Functional analysis seeks to identify and compare the eco-nomically significant activities and responsibilities undertaken, assets used and risks assumed by the parties to the transaction. Generally, quantitative and qualitative filters/criteria are used to include or exclude the potential comparables. The choice and application of selection criteria depends on the facts and circumstances of each particular case. Turnover filters are a type of quantitative criteria.

On the touchstone of FAR analysis, the big service companies are generally found providing services to different customers simultaneously, performing additional functions, assuming risks and employing unique intangible assets, unlike small size service companies. Similarly, the goodwill and brands of these companies enjoy premium pricing and due to scale of operations, these companies enjoy economies of scale in lower cost of infrastructural facilities and employees. Employee costs are generally found to be semi-variable in nature, with higher proportion of fixed cost. Further, the big service companies have a capacity and are in a position to execute large service contracts, which may not be possible otherwise for small or medium size service companies. In such a scenario, the bigger companies would also be in a position to have a better bargaining power vis-à-vis other companies.

Economies of scale are the cost advantages that enterprises obtain due to size, with cost per unit of output generally decreasing with increasing scale as fixed costs are spread out over more units of output. Often operational efficiency is also greater with increasing scale, leading to lower variable cost as well. Even though services are different from products, but still they may achieve economies of scale in business operations by using the inputs, viz, process and technology efficiently, which are necessary to render the services. For instance, just as automakers invest in the latest manufacturing processes, service companies can use technology to improve efficiency. A carpet cleaning company may purchase powerful shampooers and vacuums that decrease the time it takes to complete a job by 25 percent, thereby, claiming the cost savings from economies of scale.

Economies of scale should not be confused with the economic notion of returns of scale, which is otherwise sought to be relied by the Department, by proving that there is no linear relationship between margins and turnover and therefore, no economies of scale exist in case of service industry.

Also, the findings of the decision of Symantec Software (supra) which is sought to be relied on by Capgemini India (supra) on the contrary support the applicability of turnover filter, but however, for want of specific facts of the case, it led to opining otherwise against the Appellant Company.

Accordingly, even in case of service oriented companies, if FAR analysis indicates wide disparities in the comparables vis-à-vis taxpayer’s international transaction, then quantitative viz, turnover filter and/or qualitative filters can be applied in comparability analysis for determination of ALP. Therefore, it appears that the ratio of the Mumbai Tribunal decisions requires reconsideration.

A Special Bench of the Income-tax Appellate Tribunal has also been constituted by the Delhi Bench in the case of M/s. Fiesecke and Devirent India Pvt Ltd (in ITA No. 5924/Del/2012) on the issue under consideration with the following questions:

“1. Whether for the purposes of determining the Arm’s length Price in relation to the international transactions, quantitative filter of high/low turnover is to be applied and accordingly, high/low turnover companies vis-à-vis the assessee company are to be excluded from the comparable selected for benchmarking the transaction; and

2.    If the answer to question no. 1 is in affirmative then what should be the parameter, if any, for the exclusion of high/low turnover companies vis-à-vis the assessee company.”